Search Results

The Nova Allegro urine albumin creatinine ratio (UACR) Assay is intended for the quantitative determination of albumin, creatinine, and the albumin/creatinine ratio (UACR) in human urine. The measurement of urine albumin, creatinine, and albumin/creatinine ratio aids in the early diagnosis of nephropathy.

The Nova Allegro Analyzer is intended for in vitro diagnostic use in clinical laboratory and near-patient testing (point-of-care) settings for the quantitative determination of Nova Allegro Assays using Nova Allegro Test Cartridges.

Nova Allegro UACR Assay: The Nova Allegro UACR Assay is a completely automated assay for the determination of albumin and creatinine in human urine. The results are used to calculate the UACR (Urine Albumin to Creatinine Ratio). Nova Allegro UACR Test Cartridges are the key element a user interacts with to determine the albumin and creatinine levels in a human urine sample. The main components of the Test Cartridge are the Capillary, the reaction chamber, and the barcode label.

Nova Allegro Analyzer: The Nova Allegro Analyzer is a compact, point-of-care analyzer that features a clinically important menu of measured and calculated tests. All tests are measured with disposable, ready-to-use cartridges. The analyzer supports multiple wavelengths that are used to measure the assay of interest. The analyzer consists of the following key systems/components that the user interacts with:

• Two analytical bays where the single use test cartridges are analyzed
• Color Touchscreen Display
• Barcode Scanner
• Printer
• Data Export Options
• Ethernet Connection
• USB Port

The provided text is a 510(k) Summary for the Nova Allegro UACR Assay and Analyzer. It details the device's performance testing results. According to the prompt's request, I will extract relevant information to describe the acceptance criteria and the study proving the device meets these criteria.

It's important to note that this document describes a medical device, specifically an in-vitro diagnostic test, not an AI/ML-based device. Therefore, many of the requested fields related to AI/ML specific quality evaluation criteria, such as "number of experts used to establish ground truth," "adjudication method," "MRMC study," "effect size of human readers improving with AI," "standalone performance," and details about training sets/ground truth for training, are not applicable in this context. The study performed is a performance validation of a diagnostic assay, comparing its results to a comparative method (CM), not a reader study.

Here's a breakdown of the requested information based on the provided document:

Device: Nova Allegro UACR Assay, Nova Allegro Analyzer
Indications for Use: The Nova Allegro urine albumin creatinine ratio (UACR) Assay is intended for the quantitative determination of albumin, creatinine, and the albumin/creatinine ratio (UACR) in human urine. The measurement of urine albumin, creatinine, and albumin/creatinine ratio aids in the early diagnosis of nephropathy. The Nova Allegro Analyzer is intended for in vitro diagnostic use in clinical laboratory and near-patient testing (point-of-care) settings for the quantitative determination of Nova Allegro Assays using Nova Allegro Test Cartridges.

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are generally implied by the successful conclusion of each test and the statement that the data "meets the acceptance criteria" or "demonstrated no significant interference." Specific quantitative acceptance criteria are sometimes stated (e.g.,

Ask a specific question about this device

The UC-1800 Automatic Urine Analyzer is automated instrument which is intended for professional, in vitro diagnostic use only.

Depending on the reagent strips being used, the instruments perform semi-quantitative detection of the following analytes in urine: ascorbic acid, microalbumin, leukocytes, creatinine, ketone, urobilinogen, bilirubin, glucose, protein, specific gravity, blood and pH in urine and for qualitative determination of nitrite in urine hydrometer (optional) can determine the color and turbidity of urine. Test results may provide information regarding the status of carbohydrate metabolism, kidney and liver function, acid-base balance and bacteriuria.

The URIT 11FA urine reagent strips provide semi-quantitative tests for ascorbic acid, leukocytes, setone, urobilinogen, bilirubin, glucose, protein, specific gravity, blood and pH in urine and for qualitative determination of nitrite in urine. The URIT 11FA urine reagent strips are for use with the UC-1800 Automatic Urine Analyzer and are for professional, in vitro diagnostic use only. Test results may provide information regarding the status of carbohydrate metabolism, kidney and liver function, acid-base balance and bacteriuria.

The URIT 12FA urine reagent strips provide semi-quantitative tests for microalbumin, leukocytes, creatinine, ketone, urobilinogen, bilirubin, glucose, protein, specific gravity, blood and pH in urine and for qualitative determination of nitrite in urine. The URIT 12FA urine reagent strips are for use with the UC-1800 Automatic Urine Analyzer and are for professional, in vitro diagnostic use only. Test results may provide information regarding the status of carbohydrate metabolism, kidney and liver function, acid-base balance and bacteriuria.

UC-1800 Automatic Urine Analyzer is characterized by fully automated and simple operation. All you need to do is to set test strips and samples, press the START key, and the rest of operations are fully automated with UC-1800, which can measure samples continuously. For each measurement, the instrument automatically performs a series of operation: sample transmitting, sample aspirating, sample dropping, rinsing strip feeding and color identifying, etc. The instrument is used in conjunction with a serial of URIT urine test strips for measuring 15 parameters. Measure results are printed through either built-in printer or external printer.

Urine Reagent Strips is used to determine the components to be measured in urine by dry chemistry method together with urine analyzer. Various components to be tested in the urine can result in changes to the colors of corresponding reagent blocks on the Urine Reagent Strips. The depth of reaction color is proportional to the corresponding component to be tested in the urine. Qualitative and semi-quantitative detection can be conducted to the contents of the corresponding detected components. As a reagent for the determination of multiple components in human urine and the most basic test item for clinical urine routine test), it is suitable for the screening test or auxiliary diagnosis for clinical diagnosis, without the specificity for diseases or indications, and urine dry chemistry test is a screening test and cannot be used as a single diagnostic method.

The provided document describes the URIT UC-1800 Automatic Urine Analyzer and its associated reagent strips (URIT 11FA and 12FA Urine Reagent Strips). The information below summarizes the acceptance criteria and the studies performed to demonstrate the device meets these criteria.

1. Table of Acceptance Criteria and Reported Device Performance

The document presents the performance in terms of "Exact agreement" and "± 1 color block" agreement with expected values or between the proposed device and predicate devices. For most analytes, the criteria seem to be high exact agreement and 100% agreement within ±1 color block. Specific thresholds for acceptance were not explicitly stated as global criteria but are implied by the "Qualified" conclusions for individual tests. The comparison to predicates also uses agreement rates.

Below is a summary of the reported device performance from the "Precision / Reproducibility" section (Tables 8, 9, 10, 11) for repeatability and "Comparison Studies" section (Tables a.1, a.2, a.3) for agreement with predicate devices. Given the extensive number of analytes and concentration levels, key representative results are presented.

a. Repeatability (Within-Run Precision)
Reported as "Exact agreement" and "± 1 color block". All tests for all expected values show 100% agreement for "± 1 color block". Exact agreement varies slightly, but most are 100% or very close.

b. Reproducibility (Between-Run Precision)
Reported as "Exact agreement" and "± 1 color block". All tests for all expected values show 100% agreement for "± 1 color block". Exact agreement varies slightly, but most are 100% or very close.

c. Comparison with Predicate Devices

2. Sample Sizes Used for the Test Set and Data Provenance

• Repeatability (within-run) & Reproducibility (between-run):

  • Sample Size:
    • Repeatability: 60 measurements per concentration level per analyte (20 replicates x 3 instruments).
    • Reproducibility: 120 measurements per concentration level per analyte (20 days x 2 runs/day x 1 time/run in 3 sites, with 1 instrument/site, 3 operators).
  • Data Provenance: The document does not explicitly state the country of origin or if the data was retrospective or prospective. It refers to "negative urines and spiked urines of known concentrations." This suggests controlled laboratory-prepared samples rather than direct patient samples.

• Linearity/Assay Reportable Range:

  • Sample Size: 63 measurements per concentration level per analyte (reference solutions tested 21 times on 3 UC-1800 machines with 3 lots of reagent strips).
  • Data Provenance: Laboratory-prepared reference solutions, not clinical samples.

• Analytical Sensitivity: No sample sizes mentioned, values are stated directly.

• Critical Value: No sample sizes mentioned, values are stated directly.

• Analytical Specificity (Interference, pH, Color, Specific Gravity):

  • Sample Size: For interference studies, samples were tested 5 times on 2 UC-1800 machines with 2 batch numbers of 11FA and 12FA reagent strips. This totals 20 measurements per condition (5 tests x 2 machines x 2 strips).
  • Data Provenance: Laboratory-prepared urine samples (negative samples prepared and spiked with interfering substances or adjusted for pH/color/SG).

• Comparison Studies with Predicate Devices (Clinical Samples):

  • Uritest-500B (K082811): 1000 clinical urine samples.
  • Mission® U120 Ultra Urine Analyzer (K142391): 979 clinical urine samples.
  • AUTION MAX AX-4030 Urinalysis System (K093098):
    • Color: 1365 clinical urine samples.
    • Turbidity: 1000 clinical urine samples.
  • Data Provenance: The document explicitly states "clinical urine samples were collected." The country of origin is not specified but given the submitter's address (China), it is highly likely that these clinical samples were collected in China and were retrospective as the study compares against existing predicate devices.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

The document does not mention the use of experts to establish a "ground truth" for the test set in the traditional sense of medical image interpretation or clinical diagnosis. For chemical analyzers, the "ground truth" (or reference standard) is typically established by:

• Known concentrations: For repeatability, reproducibility, linearity, and analytical sensitivity, samples are prepared with known concentrations of the analytes.
• Reference methods: For the analytical specificity and comparison studies, the predicate devices themselves or established reference methods (e.g., 2,6-Dichlorophenolindophenol Titration Method for Ascorbic Acid, Lange method for Ketone, etc. as specified in Table 35 "Traceability") serve as the reference for comparison.

Therefore, the concept of "number of experts" and their "qualifications" for ground truth establishment, as typically applied in AI/ML performance evaluation (e.g., for image interpretation), is not directly applicable here. The ground truth is analytical and based on laboratory standards and established measurement techniques.

4. Adjudication Method for the Test Set

Not applicable in the context of this analytical device. As explained above, the "ground truth" refers to known concentrations or results from predicate/reference methods. There isn't a subjective interpretation by multiple experts that would require an adjudication method like 2+1 or 3+1.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

Not applicable. This device is an automated urine analyzer, not an AI/ML-driven diagnostic aid for human readers. It performs measurements to provide semi-quantitative results for various urine analytes. Therefore, there is no human-in-the-loop performance or comparison of human reader improvement with or without AI assistance.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done

Yes, the studies presented are all standalone performance evaluations of the device (UC-1800 Automatic Urine Analyzer with URIT 11FA/12FA Urine Reagent Strips) without human intervention in the measurement process. The device operates automatically to detect and report analyte levels. The comparison studies demonstrate its performance against existing predicate (standalone) analyzers.

7. The Type of Ground Truth Used

The ground truth used in the studies includes:

• Known Concentrations: For analytical performance studies such as precision (repeatability and reproducibility), linearity, analytical sensitivity, and analytical specificity (interference testing), samples were prepared with known, precisely measured concentrations of the target analytes or interfering substances.
• Predicate Device Measurements: For method comparison studies, the results obtained from the established predicate devices (Uritest-500B Urine Analyzer, Mission® U120 Ultra Urine Analyzer, and AUTION MAX AX-4030 Urinalysis System) served as the reference standard for comparison. These predicate devices also operate based on defined analytical principles.
• Reference Methods: Table 35 details the "Reference Method" used for traceability for each analyte (e.g., 2,6-Dichlorophenolindophenol Titration Method for Ascorbic Acid, Glucose Oxidase Method for Glucose, Acidometer measurement for pH, etc.). This indicates fundamental analytical standards are the basis for the stated detection ranges and performance.

8. The Sample Size for the Training Set

The document does not explicitly mention a "training set" in the context of machine learning. The UC-1800 Automatic Urine Analyzer uses reflectance photometry and other physical principles (refractometer for specific gravity, light-scattering for turbidity, light-transmission for color) to generate results, not an AI algorithm that learns from a dataset in the conventional sense. The "training" of such a system typically involves calibrating optical sensors and algorithms to known standards, which is part of the engineering design and quality control processes.

If interpretation of "training set" refers to the data used for the initial development and calibration of the device's measurement algorithms:

• No specific sample size for a "training set" for an AI model is reported because the device does not employ machine learning that requires a distinct "training set" and "test set" in the typical AI/ML development lifecycle.
• The system is calibrated using URIT urine control materials and calibration test strips (Table 9, "Calibration").

9. How the Ground Truth for the Training Set Was Established

As noted above, the device does not use an AI/ML model with a "training set" in the common understanding. The system's operational parameters and calibration are established using:

• Reference materials and calibrators: The device is calibrated using "URIT urine control materials and calibration test strips" (Table 9, "Calibration"). These control materials and calibration strips would have their values established using highly accurate reference methods or certified reference materials, ensuring traceability to scientific standards (as indicated in Table 35 "Traceability").
• Chemical principles: The underlying "ground truth" for the device's internal algorithms (i.e., how they convert optical signals to analyte concentrations) is based on established chemical reactions and physical measurement principles (e.g., reflectance photometry, refractometry, light scattering, dye-binding, enzymatic reactions). The "ground truth" for developing and fine-tuning these algorithms would be derived from rigorous scientific validation against these known chemical and physical properties.

Ask a specific question about this device

The Minuteful - kidney test is an in-vitro diagnostic, home-use urine analysis test system for the semi-quantitative measurement of albumin and creatinine in urine, as well as the presentation of their ratio, the albumin-creatinine ratio (ACR). The system consists of a smartphone application, proprietary Color-Board and an ACR Reagent Strip. The system is available for prescription-use only and is intended for people at risk of kidney disease. Results are used in conjunction with clinical evaluation as an aid in the assessment of kidney health.

The Minuteful - kidney test is comprised of a kit and a smartphone application. It is intended for the semi-quantitative measurement of albumin and creatinine in urine, as well as the presentation of their ratio, the albumin-creatinine ratio (ACR). The Minuteful - kidney test is intended for prescription-use only, as a home-use device to aid in the assessment of kidney health. The results can be used together with clinical evaluation to guide patient care. The device's kit includes a urine receptacle, an ACR Reagent Strip, an absorbing (i.e. "blotting") pad, a proprietary Color-Board and a user manual. The device also consists of an easy-to-use smartphone application, image recognition algorithms, and a physician compendium. The software component of the Minuteful - kidney test consists of both an application (app) and a backend server. Both components encompass different computer vision and machine learning algorithmic components, performing the image analysis activities. The app instructs the user how to accurately administer the test. The Image Validation Transfer System (IVTS) component of the Minuteful - kidney test enables its usage across a wide range of smartphone types and operating systems, essentially making the test platform agnostic.

The provided text describes the Minuteful-kidney test (K222921) and its substantial equivalence to a predicate device (K210069). However, it specifically states that "The rest of the analytical performance studies are still relevant for the modified version of the Minuteful - kidney test, and their summary is available in the predicate device documentation (K210069)." This means the detailed acceptance criteria and the comprehensive study demonstrating the device meets those criteria are not present in this document but are referenced as being in the predicate device's documentation.

Therefore, I can report on the studies performed for K222921 to assert its substantial equivalence, but I cannot provide a table of acceptance criteria and reported device performance from this document for the overall device functionality as those details are in K210069. Nor can I provide information regarding sample sizes for test sets, expert qualifications, adjudication methods, MRMC studies, standalone performance, or ground truth details for K222921's overall performance since those are tied to the K210069 submission.

The studies described in K222921 (the current device) are focused on demonstrating that changes made to the device in K222921 do not negatively impact performance, thus maintaining substantial equivalence to its predicate.

Here's an analysis based solely on the provided text for K222921, noting the limitations:

Acceptance Criteria and Device Performance (Limited to K222921 changes):

Since the comprehensive performance data is referenced in K210069, the "acceptance criteria" discussed here are implicitly related to demonstrating that the modifications in K222921 (e.g., multilingual support, software enhancements) do not degrade the performance previously established for K210069. The studies conducted for K222921 focused on the robustness of the Image Validation Transfer System (IVTS) and the analytical limits of detection.

Study Details (for K222921 specific enhancements):

1. Sample size used for the test set and the data provenance:

   • Limit of Detection (LoD): The document does not specify the sample size for the LoD study for K222921. It mentions the study was designed and executed according to CLSI document EP17-A2.
   • Illumination, Physical, Shadow, Blurriness, Misplaced Urine Stick, Dirty Color-Board Studies: The document refers to "Tested smartphones" and "different conditions," but specific numerical sample sizes (e.g., number of images, tests, or smartphones) are not provided. The data provenance is implied to be laboratory-controlled since these are experimental conditions, but no explicit country of origin or retrospective/prospective nature is stated for these new studies. The overall device is intended for home use, so these validations mimic adverse home conditions.

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

   • This information is not provided in the K222921 document. These types of analytical studies typically rely on reference methods or scientifically established standards rather than expert consensus. For the clinical performance, the document refers to the predicate device K210069, where such details would likely be found if applicable.

3. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

   • Not applicable/Not provided for these specific analytical studies. The assessment of whether performance was "not impacted" would likely come from statistical analysis against pre-defined thresholds.

4. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

   • No MRMC or comparative effectiveness study involving human readers with and without AI assistance is mentioned in the K222921 submission. This device is an in-vitro diagnostic home-use test system where the smartphone app performs the measurement, rather than assisting a human in interpreting diagnostic images. Thus, the concept of "human readers improve with AI" in a traditional MRMC sense does not directly apply to this device's function.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

   • The core of the described studies (LoD, Illumination, Physical, Shadow, Blurriness, Misplaced Urine Stick, Dirty Color-Board) are indeed standalone performance tests of the device's algorithmic capability to accurately read the test strip under various challenging conditions encountered in a home setting. The device is described as having "image recognition algorithms" and performing "image analysis activities."

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

   • For the analytical studies described (LoD, Illumination, etc.), the ground truth would be based on controlled experimental conditions and potentially quantitative reference measurements (e.g., known concentrations for LoD, controlled lighting conditions). The document does not specify the exact methods for establishing this ground truth but implies scientific rigor (e.g., "in accordance with guidance provided by the Clinical and Laboratory Standards Institute (CLSI) document EP17-A2"). For the clinical performance aspects, the document refers to K210069.

7. The sample size for the training set:

   • Not provided in the K222921 document. Training set details would typically be part of the initial K210069 submission for the machine learning algorithms.

8. How the ground truth for the training set was established:

   • Not provided in the K222921 document. This information would be found in the K210069 submission, likely involving laboratory-controlled tests with known analyte concentrations and reference methods for accurate measurement.

Ask a specific question about this device

The AUTION MAX AX-4060 Urinalysis System (AUTION MAX) is comprised of the AUTION MAX AX-4060 automated urine analyzer and AUTION Sticks 9EB multi-parameter test strips.

The AUTION MAX AX-4060 urine analyzer, when used with AUTION Sticks 9EB test strips is a fully automated urinalysis system intended for the in vitro qualitative or semi-quantitative measurement of the following analytes: glucose, protein, bilirubin, urobilinogen, pH, blood, ketones, nitrite, leukocytes, turbidity, and color. The test results of these parameters can be used in the evaluation of kidney, urinary, liver and other metabolic disorders. This system is intended to be used by trained operators in clinical laboratories.

AUTION Sticks 9EB test strips are test strips for the in vitro qualitative or semi-quantitative measurement of the following analytes: glucose, protein, bilirubin, urobilinogen, pH, blood, ketones, nitrite, and leukocytes with the AUTION MAX AX-4060 urine analyzer. The test results of these parameters can be used in the evaluation of kidney, urinary, liver and other metabolic disorders.

Special conditions for use statements: Prescription use only. AUTION Sticks 9EB test strips are not to be read visually.

The AUTION MAX AX-4060 Urinalysis System is a fully automated urine analyzer that provides a semi-quantitative or qualitative measurement for glucose, protein, bilirubin, urobilinogen, pH, blood, ketones, nitrites, leukocytes, specific gravity, turbidity, and color tone. The AUTION MAX AX-4060 Urinalysis System consists of AUTION MAX AX-4060 urine analyzer and AUTION Sticks 9EB test strips.

The AUTION Sticks 9EB test strips consist of a plastic strip containing 9 dry chemistry reagent pads impregnated with chemical substances for the determination of the above analytes in urine. These substances if present in urine leads to a chemical reaction that results in a color change, which is measured by the AUTION MAX AX-4060 urine analyzer based on spectrophotometry. In addition, three additional parameters, specific gravity, turbidity, and color tone are directly measured based on reflectance refractometry, transmitted and scattered light measurements.

Here's a breakdown of the acceptance criteria and the study that proves the device meets them, based on the provided FDA 510(k) summary:

Device: AUTION MAX AX-4060 Urinalysis System
Predicate Device: AUTION MAX AX-4030 Fully Automated Urinalysis System

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are not explicitly stated as distinct numerical targets in the document. Instead, the study aims to demonstrate substantial equivalence to the predicate device. Therefore, the "acceptance criteria" are implicitly understood as matching or performing comparably to the predicate device's established performance, particularly in terms of agreement percentage. The reported device performance is presented as the agreement between the proposed device (AUTION MAX AX-4060) and the predicate device (AUTION MAX AX-4030).

Key Performance Metric: Percentage of Exact Match and Percentage within +/- 1 Color Block Match compared to the predicate device.

(Note: The +/- 1 CB Match indicates results falling within one color block difference from the predicate, which is often considered acceptable for semi-quantitative tests.)

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size for Test Set (Method Comparison): A total of 1374 samples were used. This included:
  • 1333 natural patient samples
  • 41 spiked samples
• Data Provenance: The study was conducted at two (2) clinical laboratories. The document does not specify the country of origin, but given the FDA submission, it's likely conducted in or in accordance with standards for the U.S. and/or Japan (where the manufacturer is located). It is a prospective comparison study.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

This study does not involve human experts establishing a "ground truth" for each sample in the typical sense of diagnostic imaging or pathology. Instead, the "ground truth" for the method comparison study is the result obtained from the legally marketed predicate device (AUTION MAX AX-4030 Automated Urinalysis System). The study demonstrates the correlation and agreement of the new device's readings with the predicate device's readings. No information is provided about experts interpreting results aside from the automated systems.

4. Adjudication Method for the Test Set

Since the "ground truth" is established by the predicate device's results and the comparison is between two automated systems, there was no human adjudication process (e.g., 2+1, 3+1 consensus) described for the test set.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No. A Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not performed. This study evaluates the performance of an automated urinalysis system against a predicate automated system, not the improvement of human readers with AI assistance.

6. Standalone Performance (Algorithm Only without Human-in-the-Loop Performance)

Yes. The primary study detailed, especially the Method Comparison (Section 14), assesses the standalone performance of the AUTION MAX AX-4060 Urinalysis System. It compares the results generated by the new automated system directly against those generated by the predicate automated system, without human intervention in the result interpretation or decision-making process during the comparison phase. The device itself is an automated system intended to be used by trained operators in clinical laboratories, but the performance data presented is for the automated analyzer's output.

7. Type of Ground Truth Used

The ground truth for the method comparison study was the results obtained from the legally marketed predicate device (AUTION MAX AX-4030 Automated Urinalysis System). For the initial precision study (Tables 3-5), the "Expected Result" for quality controls served as the reference.

8. Sample Size for the Training Set

The document is a 510(k) summary for a medical device (urinalysis system), not a typically "AI-driven" software device that undergoes explicit training. The device determines analytes based on chemical reactions and spectrophotometry (reflectance refractometry, transmitted and scattered light measurements). Therefore, there is no explicit "training set" in the context of machine learning algorithms usually discussed. The system is likely calibrated and validated during its development, but this is a different process than training an AI model.

9. How the Ground Truth for the Training Set Was Established

As noted above, there is no explicit "training set" in the machine learning sense for this device. The chemical principles and measurement methods are well-established for urinalysis. Any calibration or internal development would have relied on known concentrations and reference methods for each analyte to ensure accurate colorimetric or other measurements. The basis of the technology relates to established chemical reactions on the test pads (e.g., Glucose oxidase reaction for glucose, Protein-error reaction for protein, etc.).

Ask a specific question about this device

The Minuteful - kidney test is an in-vitro diagnostic, home-use urine analysis test system for the semi-quantitative measurement of albumin and creatinine in urine, as well as the presentation of their ratio, the albumin-creatinine ratio (ACR). The system consists of a smartphone application, proprietary Color-Board and an ACR Reagent Strip. The system is available for prescription-use only and is intended for people at risk of kidney disease. Results are intended to be used in conjunction with clinical evaluation as an aid in the assessment of kidney health.

The Minuteful - kidney test is comprised of a kit and a smartphone application. It is intended for the semi-quantitative measurement of albumin and creatinine in urine, as well as the presentation of their ratio, the albumin-creatinine ratio (ACR). The Minuteful - kidney test is intended for prescription-use only, as a home-use device to aid in the assessment of kidney health. The results can be used together with clinical evaluation to guide patient care. The device is provided as a kit that comprises a urine receptacle, an ACR Reagent Strip, an absorbing (i.e. "blotting") pad, a proprietary Color-Board and a user manual. The device also consists of an easy-to-use smartphone application, image recognition algorithms, and a physician compendium. The software component of the Minuteful - kidney test consists of both an application (app) and a backend server. Both components encompass different computer vision and machine learning algorithmic components, performing the image analysis activities. The app instructs the user how to accurately administer the test. The Image Validation Transfer System (IVTS) component of the Minuteful - kidney test enables its usage across a wide range of smartphone types and operating systems, essentially making the test platform agnostic.

Here's a breakdown of the acceptance criteria and study information for the Healthy.io Minuteful - kidney test (K210069), based on the provided document:

Acceptance Criteria and Device Performance for Minuteful - kidney test (K210069)

1. Table of Acceptance Criteria and Reported Device Performance

Note: The document implies acceptance criteria by reporting performance results against the predicate device that demonstrate substantial equivalence.

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size: More than 450 subjects were recruited for the clinical trials.
• Data Provenance: The document does not explicitly state the country of origin. The study was a prospective clinical trial, as subjects were "recruited" and tasks were "completed" within the context of the study.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

The ground truth was established by comparing the Minuteful - kidney test results to those obtained from a professional user performing the test on the URISCAN Optima Urine Analyzer (predicate device).

• Number of Experts: Not explicitly stated as a count of individual professionals, but referred to as "a professional user." It's implied that this
  professional operated the predicate device.
• Qualifications: "Professional user" suggests trained laboratory or healthcare personnel familiar with operating the URISCAN Optima Urine Analyzer and interpreting its results. Specific credentials (e.g., medical technologist, clinical laboratory scientist, years of experience) are not provided.

4. Adjudication Method for the Test Set

The adjudication method involved a 2-part comparison:

1. A lay user (subject in the clinical trial) performed the test using the Minuteful - kidney test app.
2. A professional user (operating the predicate device, URISCAN Optima Urine Analyzer) then performed the test on the same urine sample.

The professional user was blinded to the results of the lay user until after they had completed their test. This can be considered a form of adjudication where the predicate device's result, as read by a professional, serves as the comparison benchmark. There was no explicit multi-expert consensus or 2+1/3+1 method described for establishing a single "ground truth" independent of the comparison devices; rather, the predicate device's output was the reference.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done

No, an MRMC comparative effectiveness study was not explicitly done in the sense of comparing multiple human readers with and without AI assistance on the same cases. The study compared a lay user with the AI-powered device to a professional user with a predicate device. It was a method comparison study to show substantial equivalence, not a study evaluating human reader improvement with AI assistance.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

The clinical performance study involved a lay user interacting with the smartphone application and performing the test steps guided by the app. While the app uses "image recognition algorithms" and "machine learning algorithmic components," it is not a purely standalone algorithm-only performance assessment in the sense of a laboratory-based algorithm evaluation without human interaction for image capture and strip preparation. However, the analytical performance testing (Precision, Interference, Limit of Detection, Linearity, Stability) would represent the closest to "standalone" algorithm performance testing, as these evaluate the device's technical capabilities in a controlled environment. The linearity study, showing 100% exact match for every level of albumin, creatinine, and ACR, is a strong indicator of the core algorithm's accuracy at different concentrations.

7. The Type of Ground Truth Used

The ground truth was established by comparison to a legally marketed predicate device (URISCAN Optima Urine Analyzer) operated by a professional user. This is a form of reference standard comparison where the predicate device's output serves as the truth.

8. The Sample Size for the Training Set

The document does not provide information regarding the sample size for the training set used for the device's algorithms.

9. How the Ground Truth for the Training Set Was Established

The document does not provide information on how the ground truth was established for the training set of the device's algorithms. It only describes the ground truth for the clinical performance (test set) comparison.

Ask a specific question about this device

The ACR | LAB Urine Analysis Test System is comprised of a smartphone application, a proprietary Color-Board, and ACR Reagent Strips. It is intended for the semi-quantitative detection of albumin and creatinine in urine, as well as the presentation of their ratio. The ACR | LAB Urine Analysis Test System is intended for in-vitro diagnostic use by a healthcare professional in a point of care setting. These results may be used in conjunction with clinical evaluation as an aid in the diagnosis for kidney function.

The ACR | LAB Urine Analysis Test System is comprised of a smartphone application, a proprietary Color-Board and ACR Reagent Strips. It is intended for the semi-quantitative detection of albumin and creatinine in urine, as well as the presentation of their ratio. The device is provided as a kit that is comprised of a canister of 100 FDA-cleared urine test strips (ACON Laboratories Inc. Mission Urinalysis Reagent Strips (Microalbumin/Creatinine) K150330), 10 Color-Boards, and a User Manual. The ACR | LAB Urine Analysis Test System also consists of a smartphone application for use on iPhone 7 device (iOS 12), and an image recognition algorithm running on the Backend. The software component of the ACR | LAB consists of both an application (App) and a Backend server (Backend). The App instructs the professional user how to accurately perform the test. The App conducts a series of boundary condition analyses, and if the scan is approved, sends the information to the Backend for complete analysis and results classification. Once analyzed, the results are securely transmitted to a patient Electronic Medical Record for review by a healthcare professional. The patients do not have access to the results at any point during the testing process.

Here's a breakdown of the acceptance criteria and the study that proves the device meets them, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The core acceptance criteria are based on the agreement between the ACR | LAB Urine Analysis Test System and the predicate device (ACON Laboratories' Mission U120 Ultra Urine Analyzer). The study aimed for high percentages of exact match and ±1 color block match.

| Metric (Agreement with Predicate Device) | Acceptance Criteria (Implicit from "high levels of accuracy") | Reported Device Performance (ACR | LAB) |
| :--------------------------------------- | :----------------------------------------------------------- | :-------------------------------------- |
| Albumin | High Exact Match % | 89% Exact Match |
| Albumin | High ±1 Color Block Match % | 100% ±1 Color Block Match |
| Creatinine | High Exact Match % | 84% Exact Match |
| Creatinine | High ±1 Color Block Match % | 100% ±1 Color Block Match |
| Albumin-Creatinine Ratio | High Exact Match % | 93% Exact Match |
| Albumin-Creatinine Ratio | High ±1 Color Block Match % | 100% ±1 Color Block Match |

Note: The document explicitly states that the primary acceptance criteria for the method comparison study were the percent of exact match and ±1 color block match. While specific numerical targets for "high levels of accuracy" are not given as explicit "acceptance criteria," the reported performance exceeding predicate device agreement in these metrics is implicit evidence of meeting those criteria.

2. Sample Size Used for the Test Set and Data Provenance

• Sample Size:
  • Native Urine Samples: 375 subjects
  • Contrived Samples: 60 samples
  • Total Samples for Clinical Performance: 435 samples (375 native + 60 contrived)
• Data Provenance: The study evaluated native urine samples from 375 subjects as well as 60 contrived samples at three U.S. clinical sites. This indicates the data is prospective (newly collected for the study) and from the United States.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

The document refers to the predicate device (ACON Laboratories' Mission U120 Ultra Urine Analyzer) as the "ground truth" or reference for comparison.

• Number of "Experts" (for ground truth): The ground truth was established by readings from the predicate device (ACON Laboratories' Mission U120 Ultra Urine Analyzer). Two separate lab technicians measured each urine sample, one using the iPhone 7 device (ACR | LAB) and the second using the predicate device (U120 Ultra).
• Qualifications of "Experts": The document states "Two separate lab technicians were responsible for measuring each urine sample." Their specific qualifications (e.g., years of experience, certifications) are not explicitly mentioned, but they are identified as "lab technicians."

4. Adjudication Method for the Test Set

The adjudication method appears to be none in the traditional sense of multiple human experts reviewing and deciding. Instead, the study directly compared the results of the ACR | LAB device against the results obtained from the predicate device (ACON Laboratories' Mission U120 Ultra Urine Analyzer). Each sample was tested once by the ACR | LAB and once by the predicate device, and the agreement between these two measurements was assessed.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size of Human Improvement

• Was an MRMC study done? No, a traditional MRMC comparative effectiveness study was not done. This study is focused on the performance of a clinical diagnostic device, where consistency with a reference device is key, rather than an AI-assisted interpretation by multiple human readers.
• Effect size of human readers improve with AI vs without AI assistance: This information is not applicable/not provided, as the study design was a direct comparison of the new device to a predicate device, not an assessment of human reader performance with and without AI assistance. The ACR | LAB system itself includes the smartphone app and image recognition algorithm as central components of its operation, so human interaction is inherent, but not a separate "with vs. without AI assistance" arm for human readers.

6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done

The device description indicates that the "ACR | LAB Urine Analysis Test System is comprised of a smartphone application, a proprietary Color-Board, and ACR Reagent Strips." It also mentions that "The App instructs the professional user how to accurately perform the test. The App conducts a series of boundary condition analyses, and if the scan is approved, sends the information to the Backend for complete analysis and results classification."

This suggests that the device does not operate purely standalone (algorithm-only without human-in-the-loop). A healthcare professional is involved in:
* Performing the physical test (dipping the strip).
* Operating the smartphone application.
* Placing the strip on the Color-Board for scanning.

The algorithm on the Backend performs the complete analysis and classification, but this is initiated and guided by the human user through the app. Therefore, it's a human-in-the-loop system, and no standalone algorithm-only performance is documented separately.

7. The Type of Ground Truth Used

The ground truth for the clinical performance study was established by comparison to a legally marketed predicate device (ACON Laboratories' Mission U120 Ultra Urine Analyzer). The aim was to demonstrate substantial equivalence, meaning the new device's results should align closely with those of the established predicate.

8. The Sample Size for the Training Set

The document does not specify the sample size used for the training set for the image recognition algorithm. It focuses on the validation studies.

9. How the Ground Truth for the Training Set Was Established

The document does not explicitly describe how the ground truth for the training set was established. It broadly mentions the software validation and hazard analysis but doesn't detail the data labeling process for the algorithm's training. It is common for such systems to be trained on a large dataset of images with corresponding known (e.g., laboratory-confirmed) values for albumin and creatinine, but this specific information is not provided here.

Ask a specific question about this device

The cobas u 601 urinalysis test system is comprised of the cobas u 601 urine analyzer and the cobas u pack.

The cobas u 601 urine analyzer when used with the cobas u pack is a fully automated urinalysis system intended for the in vitro qualitative or semi- quantitative determination of urine analytes, nitrite, protein, glucose, ketones, urobilinogen, bilirubin, color and erythrocytes, as well as clarity. These measurements are useful in the evaluation of renal, urinary, hepatic and metabolic disorders. This system is intended to be used by trained operators in clinical laboratories.

The cobas u pack is a cassette loaded with cobas u 601 test strips for the in vitro qualitative or semi-quantitative determination of pH, leukocytes, nitrite, protein, glucose, ketones, urobilinogen, bilirubin, color and erythrocytes in urine with the cobas u 601 urine analyzer. These measurements are useful in the evaluation of renal, urinary, hepatic and metabolic disorders.

The cobas u 601 Urinalysis Test System consists of the following components:

• . cobas u 601 urine analyzer
• . cobas u pack

The cobas u 601 urine analyzer is a fully automated urine analysis system. It is optimized for high throughput workloads in the professional environment. The cobas u 601 urine analyzer performs a maximum theoretical throughput of up to 240 samples per hour.

The cobas u 601 analyzer consists of several major components:

• Rack transport system
• Liquid handling system
• Test strip cassette compartment
• Automated test strip processing area
• Photometer which is a 4 wavelength reflectance measuring unit based on a Complementary Metal Oxide Semiconductor chip used in digital cameras (CMOS sensor)
• Physical Measurement Cell (PMC): flow cell connected to an optical detector ●
• Touch Screen
• Inbuilt Computer

The functions of the cobas u 601 urine analyzer include:

• Sample loading and transport ●
• . Sample identification
• Robotic pipetting of samples onto test pads on test strips
• Robotic aspiration of samples into the PMC
• . Controlled incubation
• . Photometric measurement of test strips
• Optical determination in the PMC
• Automatic disposal of used test strips ●
• . Result readout
• Result memory
• Optional formats for data output including electronic result communication

The operating system will be a Microsoft Windows for embedded devices. The system will use a Postgres/SQL database.

The cobas u 601 urine analyzer is designed to be inter-connected mechanically and electronically with another urine sediment analyzer (cobas u 701) in order to create a urine work area (cobas® 6500).

The cobas u pack is a cassette containing 400 tests strips. The cobas u 601 analyzer will use the cobas u pack to dispense single test strips for each sample.

Each test strip has ten individual test pads that are used to test for different substances or characteristics. The test strips are analyzed automatically through the analyzer. One test strip is used per sample. When a strip is dispensed for use by the cobas u 601, an aliquot of the urine sample is pipetted onto each of the test pads. The resulting color changes are measured photometrically.

The test strip in the cobas u pack cassette ("cassette test strip") is a multi-parameter urine analysis test strip, with test pads for blood (Erythrocytes), Leukocytes, Nitrite, Proteins, Glucose, Ketones, Bilirubin, Urobilinogen, Color and pH.

Here's an analysis of the provided text, focusing on the acceptance criteria and the study proving the device meets them:

Device: cobas u 601 urinalysis test system
Predicate Devices: cobas u 411 (for pH, leukocytes, nitrite, protein, glucose, ketones, urobilinogen, bilirubin, and erythrocytes), Urisys 2400 (for specific gravity, color, and clarity).

Based on the provided 510(k) Summary, the term "acceptance criteria" is not explicitly defined as a single, overarching set of numerical thresholds for all performance metrics. Instead, the document describes the studies performed and their results, implying that demonstrating acceptable performance within clinical ranges and in comparison to predicate devices constitutes meeting the "acceptance criteria" for substantial equivalence.

For each study, the "acceptance criteria" are implied by the reported results meeting the necessary performance for a diagnostic device, particularly demonstrating "exact agreement" or "agreement ± 1 block" within clinically relevant ranges and comparable to the predicate devices.

The information is extracted from the "NON-CLINICAL PERFORMANCE EVALUATION" and "CLINICAL PERFORMANCE DATA" sections.

1. Table of Acceptance Criteria (Implied) and Reported Device Performance

Since explicit numerical acceptance criteria for each test in a pass/fail format are not directly stated as "acceptance criteria," the table below presents the implied acceptance criteria (what the study aimed to demonstrate as acceptable performance) and the reported device performance as found in the document. The primary method for showing acceptance is often "exact agreement" or "agreement ± 1 block" with the predicate device/reference.

2. Sample Sizes and Data Provenance

• Test Set Sample Sizes:

  • Analytical Sensitivity: Multiple samples tested for each analyte, each sample measured 20 times on each of 3 instruments using 3 reagent test strip lots. (e.g., 20 measurements x 3 instruments x 3 lots = 180 total per sample condition per analyte if all combinations were tested). Specific number of "samples" (unique spiked concentrations) not explicitly stated for each analyte.
  • Drug and Endogenous Interferences: Urine pools (negative/normal and first positive range) tested at 2 concentrations of interferents. Multiple replicates measured. Number of replicates not specified for each condition.
  • Color Interference: Final test solutions (for each parameter) tested in a 10-fold determination. (e.g., 10 measurements per condition).
  • Stability:
    • Real-time: Defined set of samples (native urine, artificial urine, low/high spiked urine) measured with n=10 determinations at each time point (0, 3, 13, 16 months).
    • On-board: 400 tests over 15 days from a single cassette (using native and artificial urine samples).
  • Precision (Repeatability): Controls measured in 2 runs, 21 determinations each, producing n=42 results per control.
  • Precision (Intermediate Precision): Controls measured in 21 days with 2 runs per day and duplicate measurements per control, producing n=84 results per control.
  • Method Comparison (cobas u 411): "fresh samples" used to cover claimed ranges. Specific total number of samples for comparison is not explicitly stated.
  • Method Comparison (Urisys 2400):
    • Color: 478 total samples.
    • Clarity: 1364 total samples.
  • Sample Carryover: Not specified, but involved testing low/negative and high concentration samples.

• Data Provenance: The document does not explicitly state the country of origin for the data or whether the studies were retrospective or prospective. Given it's a 510(k) submission for an in vitro diagnostic device, these are typically prospective laboratory studies conducted by the manufacturer, often at their R&D facilities or contracted clinical sites.

3. Number of Experts and Qualifications for Ground Truth

• The document does not specify the number of experts used to establish ground truth for the test sets.
• It also does not specify the qualifications of these experts.
• For urinalysis strips, ground truth is typically established by reference methods such as quantitative chemical assays or microscopic examination, rather than relying solely on human expert consensus on visual interpretation of the strips themselves. The comparison is made against a "reference system" which implies an objective and validated method.

4. Adjudication Method for the Test Set

• The document does not mention any adjudication method (e.g., 2+1, 3+1) for the test sets. For objective chemical measurements like those performed by this device, human adjudication of "ground truth" is typically less relevant than the use of quantitative reference methods.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

• A MRMC comparative effectiveness study was not conducted. This device is an automated urinalysis system, and its performance is evaluated against established analytical methods and predicate devices, not by comparing human reader performance with and without AI assistance. The "AI" component is implicit in the automated analysis of reflectance, which is a core function of the machine's software, but it's not presented as an AI-assistive tool for human readers.

6. Standalone (Algorithm Only) Performance

• Yes, the performance data presented (e.g., Analytical Sensitivity, Accuracy/Method Comparison studies) represents the standalone performance of the cobas u 601 urinalysis test system. It is an automated device designed to perform urinalysis without human interpretation of the test strip results; the human role is in operating the system and interpreting the numerical/qualitative results provided by the machine.

7. Type of Ground Truth Used

• The ground truth used for these studies generally aligns with:
  • Reference System/Predicate Device Comparison: For the method comparison studies, the cobas u 411 and Urisys 2400 systems served as "reference systems" against which the new device's qualitative and semi-quantitative results were compared.
  • Spiked Samples/Defined Concentrations: For sensitivity, interference, stability, and precision studies, the ground truth was established by preparing urine samples with precisely known concentrations of analytes or interfering substances ("spiking the negative urine pool with the appropriate agent," "known concentrations").
  • Control Materials: For precision studies, standardized control materials with known values were used.

8. The Sample Size for the Training Set

• The document does not provide information regarding the sample size for a "training set." This type of device, based on reflectance photometry and chemical reactions, typically relies on predetermined algorithms derived from extensive analytical characterization of the strip chemistry and optical properties, rather than "training" an AI model in the conventional machine learning sense using a large, distinct "training set" of patient data. The development process would involve calibration and algorithm refinement using controlled samples, but not necessarily a "training set" as defined in AI/ML contexts with expert-labeled ground truth for each case.

9. How the Ground Truth for the Training Set Was Established

• Since there's no explicit mention of a "training set" in the context of an AI/ML model for this device, the question of how ground truth was established for it is not applicable in the provided document. The "training" in this context would refer to the calibration and algorithm development process, which relies on the principles of analytical chemistry and physics inherent to reflectance photometry to accurately read the color changes on the test strips. This would involve precise chemical and optical characterization using known standards and samples.

Ask a specific question about this device

The DIP | U.S. Urine Analysis Test System consists of a smartphone application, a proprietary Color-Board, and Urinalysis Reagent Strips. It is intended for the semi-quantitative detection of the following analytes in urine: Glucose. Specific Gravity, Blood, pH and Protein, as well as the qualitative detection of Nitrite,

The DIP | U.S. Urine Analysis Test System is intended for prescription home-use only, with results provided directly to the physician. The results can be used to guide patient management and care, and aid in the diagnosis and monitoring of metabolic or systemic diseases that affect kidney function and endocrine disorders. Physician interpretation of the results should be made in conjunction with the patient's other clinical information to determine if further confirmatory tests or consultations are necessary. Patients do not have access to the results at any point in the process.

The DIP | U.S. Urine Analysis Test System consists of a smartphone application, a proprietary Color-Board, and Urinalysis Reagent Strips. It is intended for the semi-quantitative detection of the following analytes in urine: Glucose, Specific Gravity, Blood, pH and Protein, as well as the qualitative detection of Nitrite.

The DIP | U.S. Urine Analysis Test System is intended for prescription home-use only, with results provided directly to the physician. The results can be used to guide patient management and care, and aid in the diagnosis and monitoring of metabolic or systemic diseases that affect kidney function and endocrine disorders. Physician interpretation of the results should be made in conjunction with the patient's other clinical information to determine if further confirmatory tests or consultations are necessary. Patients do not have access to the results at any point in the process.

The device is provided as a kit that comprises a urine receptacle, an FDA-cleared urine test strip (ACON Mission Urinalysis Reagent Strips, 510K number K061559), a Color-Board, and a User Manual. The DIP | U.S. Urine Analysis Test System also consists of a smartphone application for use with a LG Nexus 5 device (running operating system Lollipop 5.0), and an image recognition algorithm running on the back-end.

The software component of the DIP | U.S. Urine Analysis Test System consists of both an application and a back-end server. The App instructs the patient how to accurately administer the test and conducts a number of algorithm processes. Once analyzed, the DIP | U.S. Urine Analysis Test System's software securely transmits the clinical results directly to the patient's Electronic Medical Records for review by the physician. As stated above, the patients do not have access to the results at any point during the testing process.

The provided text describes the acceptance criteria and study proving the device meets those criteria for the Healthy.io DIP | U.S. Urine Analysis Test System.

Here's a breakdown of the requested information:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria are primarily defined by the "percent exact match" and "±1 color block match" compared to the predicate device (ACON Mission U500 Urinalysis System). The exact numerical acceptance thresholds are not explicitly stated as strict percentages for all analytes, but the narrative implies "high-levels of accuracy" and specific target ranges are mentioned for certain analytes.

Additional Performance Metrics (from Analytical Performance Testing):

• Repeatability: 99.3% exact match
• Reproducibility: 98.5% exact match
• Linearity: At least 89.4% exact match and 100% ±1 color block accuracy.
• Illumination Study: 99.5% exact match
• Boundary Study: 99.5% exact match

2. Sample Sizes Used for the Test Set and Data Provenance

Two method comparison studies were conducted for the test set:

• Study 1:

  • Sample Size: 429 subjects, 500 total samples (including spiked samples). Only 284 results from LG Nexus 5 smartphones were used for performance data.
  • Data Provenance: Two U.S. clinical sites. The studies involved lay-users in a simulated home-use environment. This indicates prospective data collection for the purpose of this validation.

• Study 2:

  • Sample Size: 250 subjects, 289 total samples (including spiked samples).
  • Data Provenance: One U.S.-based clinic. Similar to Study 1, this appears to be prospective data collection in a simulated home-use setting.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

The ground truth for the comparison studies was established by "a laboratory professional using the predicate (ACON Mission U500 Urinalysis System)" measuring aliquots of the same samples. The specific number of laboratory professionals or their detailed qualifications (e.g., years of experience, specific certifications) are not explicitly stated in the provided text.

4. Adjudication Method for the Test Set

The text describes comparing the DIP system's results to those from a laboratory professional using the predicate device. It does not mention any formal adjudication method (e.g., 2+1, 3+1 consensus) for discrepancies between the device and the predicate. The predicate device's readings appear to be treated as the reference standard (ground truth).

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and the Effect Size of How Much Human Readers Improve with AI vs. Without AI Assistance

This was not a MRMC comparative effectiveness study in the typical sense of evaluating human reader improvement with AI assistance. The study evaluates the performance of the device itself as used by a lay-user, compared to a predicate device operated by a laboratory professional. There is no mention of human readers interpreting images with and without AI assistance to measure improvement.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

The performance data presented, particularly the "Method Comparison Study," represents the device's performance with a human-in-the-loop (lay-user) interacting with the smartphone application and Color-Board. The device itself (including its image recognition algorithm on the back-end) performs the analysis of the test strip after the user interacts with it. The usability results (e.g., 99% success rate for lay users completing the test) address the human-in-the-loop aspect.

However, the "Analytical Performance Testing" (Precision, Interference, Limit of Detection, Linearity, Illumination, Boundary Studies) implicitly evaluates the algorithm's performance under controlled conditions with pre-determined reagent values or spiked samples, which can be seen as a form of standalone evaluation of the core analytical capability. For example, the Illumination Study and Boundary Study evaluate the device's (and thus the algorithm's) ability to measure accurately under varying external conditions. In these studies, the device measured against "predetermined reagent values," which serves as the ground truth for evaluating the algorithm's accuracy under those specific conditions.

7. The Type of Ground Truth Used

The primary ground truth used for the method comparison studies (clinical validation) was the readings obtained from the predicate device (ACON Mission U500 Urinalysis System) by a laboratory professional.

For the analytical performance studies (Precision, Interference, Limit of Detection, Linearity, Illumination, Boundary), the ground truth was based on validated spiked urine solutions at known concentrations or predetermined reagent values.

8. The Sample Size for the Training Set

The provided text does not explicitly state the sample size used for the training set for the image recognition algorithm. It focuses on the validation studies.

9. How the Ground Truth for the Training Set Was Established

The text does not describe how the ground truth for the training set was established. It only mentions that the device includes an "image recognition algorithm running on the back-end." It is standard practice for such algorithms to be trained on large datasets with established ground truth, but the details of this process are not provided in this specific document.

Ask a specific question about this device

The inui In-Home Urine Analysis Test System consists of the inui In-Home Urine Analysis Device and the inui Urine Analysis Mobile Application. The inui In-Home Urine Analysis Test System is intended for detecting the following parameters in urine: Protein, Glucose, Leukocytes, Nitrites, and Ketones. The test results provide information regarding the status of Urinary Tract Infections (UT), proteinuria, glucosuria, and ketonuria. These results can be used as an aid for monitoring kidney functions and metabolic disorders (e.g. diabetes mellitus), and can be used in the screening for Urinary Tract Infections (UTI).

The inui In-Home Urine Analysis Device is intended for use as a prescription home use device.

The inui In-Home Urine Analysis Test System consists of inui In-Home Urine Analysis Device and the inui Urine Analysis Mobile Application. It is an in vitro diagnostic device comprised of a disposable test paddle, a urine collection cup, a gray background sheet, and a mobile application ("inui App"). The plastic "paddle" contains multiple chemistry test pads (CTPs) and 1 quick response (QR) barcode. CTPs contain pre-dried chemicals that react to specific substances in a urine sample. A color reaction occurs on the CTP based on the amount of substance in the urine sample. The color reaction is captured as an image using the mobile device camera and should only be read using the inui App. The results are reported as specified by the measurement range for each test. The QR Code is used for quality control purposes to track lot number, expiration date, and paddle use.

The inui In-Home Urine Analysis System reports semi-quantitative or qualitative results for each test parameters.

Here's a breakdown of the acceptance criteria and the study details for the inui In-Home Urine Analysis Test System, based on the provided FDA 510(k) summary:

Acceptance Criteria and Reported Device Performance

General Acceptance Criteria from Analytical Performance:

• Precision and Reproducibility: Proportion of agreement for each level and analyte tested to be above 95% to the expected values.
• Limit of Detection (LOD) and Linearity: Established for each level and analyte.
• Specificity and Interference Testing: Identification of interfering substances and their minimum concentration levels causing false positives/negatives. The device should produce an error message for substances introducing colors outside the analyte's normal range.

Clinical Performance Acceptance Criteria:

• Clinical Accuracy (Method Comparison): Lay-user clinical test results should be comparable to those of professional users using the predicate device. Specific agreement percentages are implicitly set by the reported data in Table 5.
• Usability: Lay users should be able to follow procedural steps to obtain successful test results with a specified performance percentage.
• Lay-user Reproducibility: Lay-users should obtain the same results after repeated testing with a high reproducibility percentage.

1. Table of Acceptance Criteria and the Reported Device Performance

• Protein: Albumin (>3000 mg/dL), Bilirubin (>10 mg/dL), Hemoglobin (>100 mg/dL), Chlorhexidine (>40 mg/dL), Riboflavin (>5 mg/dL), Hypochlorite (>375 mg/dL), Specific Gravity (>1.025).
• Leukocyte: Human Leukocyte Esterase (>0.025 U/mL), Hemoglobin (>150 mg/dL), Albumin HSA (>3000 mg/dL), Bilirubin (>10 mg/dL), Sodium Chloride (>324 mg/dL), Hypochlorite (>375 mg/dL), Chlorhexidine (>60 mg/dL), Microbial Peroxidase (>0.65%), Riboflavin (>5 mg/dL), Sodium Acetate (>600 mg/dL), Sodium Bicarbonate (>630 mg/dL), Specific Gravity (>1.020), Urobilinogen (>4 mg/dL).
• Nitrite: Sodium Nitrite (>5 mg/dL), Human Hemoglobin (>100 mg/dL), Hypochlorite (>375 mg/dL), Human Leukocyte (>0.0375 U/mL), Urobilinogen (>4 mg/dl), Sodium Bicarbonate (>945mg/dl), Sodium Acetate (>900 mg/dL), Hypochlorite (>375 mg/dL), Sodium Nitrite (>7.5 mg/dL).
• Ketone: Lithium Acetoacetate (>40 mg/dL), Hypochlorite (>375 mg/dL), Sodium Nitrite (>7.5 mg/dL).
• Glucose: D-(+)-Glucose (>500 mg/dL), Hypochlorite (>375 mg/dL), Bilirubin (invalid results at all conc), Lithium Acetoacetate (>80 mg/dL), Sodium Chloride (>486 mg/dL). (Pages 8-9) |
  | Clinical Performance | | |
  | Method Comparison (Lay User vs. Predicate) | Lay user results comparable to professional users with predicate device; high agreement percentages. An explicit threshold for each analyte/level is implied by the successful submission. | Achieved; "clinical test results that are comparable to those of a professional users".
  Protein: Negative (83% exact, 100% within 1 block), Trace (75% exact, 100% within 1 block), Moderate (90% exact, 100% within 1 block), Large (100% exact, 100% within 1 block).
  Glucose: Negative (100% exact, 100% within 1 block), Low (74% exact, 100% within 1 block), Moderate to Large (100% exact, 100% within 1 block).
  Leukocyte: Negative (100% exact), Positive (97% exact).
  Nitrite: Negative (96% exact), Positive (100% exact).
  Ketone: Negative/Trace (99% exact, 100% within 1 block), Small (67% exact, 100% within 1 block), Moderate to Large (100% exact, 100% within 1 block). (Table 5, Page 10) |
  | Usability | Lay users can follow all procedural steps to obtain a successful test result, with a high overall performance percentage. | Achieved: Overall performance of 88.9% (1st paddle), 86.7% (2nd paddle), and 91.1% (3rd paddle). (Page 10) |
  | Lay-user Reproducibility | High reproducibility (e.g., 100%) for lay-user repeated testing. | Achieved: "reproducibility of the inui Device in the hands of the lay user was 100%." (Page 10) |

2. Sample Size Used for the Test Set and Data Provenance

• Test Set Sample Size:
  • Clinical Accuracy (Method Comparison): Two studies were conducted using 190 and 91 lay users, respectively.
  • Usability Study: 45 lay users.
  • Lay-User Reproducibility Study: 10 lay users.
• Data Provenance: The document does not explicitly state the country of origin. It indicates that the clinical evaluation followed an "Investigational Review Board approved clinical protocol," suggesting prospective data collection.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

• The ground truth for the clinical accuracy study was established by comparing lay-user results from the inui device to results obtained from the predicate device, Siemens Multistix® 10SG, "by professional users."
• The document does not specify the exact number of professional users or their specific qualifications (e.g., "radiologist with 10 years of experience"). It just refers to them as "professional users."

4. Adjudication Method for the Test Set

• The document describes a direct comparison method where lay-user results from the inui device are compared against results from the predicate device read by professional users.
• It does not mention an explicit adjudication method like "2+1" or "3+1" to resolve discrepancies between multiple professional readers or to establish a consensus ground truth beyond the professional reading of the predicate device. The professional reading of the predicate device serves as the reference.

5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance

• No, an MRMC comparative effectiveness study involving AI assistance for human readers was not explicitly described.
• The study focuses on the standalone performance of the inui system (which incorporates an app for image capture and analysis) when used by lay-users, against a predicate device read by professional users.
• The system itself is the "AI assistance" in the sense that the mobile app performs the color analysis traditionally done visually. The comparison is between the inui system's performance by lay users and the predicate device's performance by professional users, not human readers with vs. without AI assistance.

6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Yes, in spirit, the core of the inui device's performance is essentially standalone algorithm performance within the context of a "lay-user-in-the-loop" setting.
• The "inui App" performs the electronic color analysis via the mobile phone camera. The "lay user reproducibility study" and the "clinical accuracy study" demonstrate the device's (including the algorithm's) performance when operated by lay users. The comparison in the clinical accuracy study is effectively between the inui device's algorithm output (managed by lay users) and a predicate device's visual interpretation by professionals.

7. The Type of Ground Truth Used

• The ground truth for the clinical accuracy (method comparison) study was established by comparison to a legally marketed predicate device (Siemens Multistix® 10SG device) assessed by professional users. This is a form of "reference method" ground truth.
• For the analytical performance studies (Precision, LOD, Specificity), the ground truth was based on expected values and established protocols in accordance with CLSI guidelines.

8. The Sample Size for the Training Set

• The document is a 510(k) summary, which focuses on validation data. It does not provide information on the sample size for the training set used to develop the inui device's mobile application algorithm.

9. How the Ground Truth for the Training Set Was Established

• The document does not provide details on how the ground truth for the training set was established, as it pertains to the development process of the device which is typically not included in a 510(k) summary.

Ask a specific question about this device

This device is intended for the in vitro measurement of the following in urine: Leukocyte, Nitrite, Urobilinogen, Protein, pH, Blood, Specific gravity, Ketone, Bilirubin, Glucose. These strips are intended for prescription, in vitro diagnostic use only and they are visually read.

DUS 2GP reagent strips provide qualitative and semiquantitative measurements for protein, and glucose in urine specimens. Test results may provide information regarding the status of carbohydrate metabolism and kidney function.

DUS 5 reagent strips provide qualitative and semiquantitative measurements for leukocytes, nitrite, blood , protein, and glucose in urine specimens. These measurements are used to aid in the diagnosis of metabolic disorders, kidney function anomalies and urinary tract infections.

DUS 10 reagent strips provide qualitative and semiquantitative measurements for specific gravity, pH, leukocytes, nitrite, protein, glucose, ketone, urobilinogen, bilirubin and blood in urine speciments are used to aid in the diagnosis of metabolic disorders, kidney function anomalies, urinary tract infections and liver function.

The DUS Series are urine test strips with different reagent pads for the determination of specific gravity, pH, leukocytes, nitrite, protein, glucose, ketone, urobilinogen, bilirubin and blood affixed onto plastic strips. of which leukocyte, nitrite, urobilinogen, protein, pH, blood, specific gravity, ketone, bilirubin and glucose reagent pads are affixed onto the plastic strips. The reagent pads react with analytes in the urine giving a visible color. Results are confirmed by comparison of the test strip with the color chart on the container. For each color result for each analyte, a semiquantitative value is available on the box label (e.g. bilirubin results include negative, 1, 2, and 4 mg/dL) and the associated qualitative result (e.g. bilirubin results include negative. +. ++, +++).

Here's an analysis of the provided document, outlining the acceptance criteria and study details for the DUS Reagent Strips:

Acceptance Criteria and Device Performance

The provided document does not explicitly state pre-defined acceptance criteria for the "Exact Agreement" or "Agreement within +/- one block (%)" in the method comparison study. However, the study results, which consistently show high percentages (generally in the high 90s and 100%) for both metrics across all analytes, implicitly represent the device meeting an expected high level of agreement with the predicate device.

For the linearity/assay reportable range, the acceptance criteria are implied by the reported "% Exact match." A high percentage (generally 97.7% to 100%) indicates acceptance.

For the detection limit study, the acceptance criteria are stated as "Sensitivity was defined as the cutoff for which ≥95% of the contrived pooled measurements were trace or the first positive result."

The interference study implicitly accepts the device if interference is either not observed at clinically relevant concentrations or if observed interferences are adequately addressed in the labeling.

Here's a table summarizing the reported device performance, where the values themselves act as the demonstration that the implied or stated acceptance criteria (high agreement, 95% detection, or non-interference/labeling for interference) were met.

Table of Reported Device Performance

Study Details

1. Sample size used for the test set and the data provenance:

   • Method Comparison Study: Total of 867 samples.
     • Provenance: Fresh urine samples obtained at three medical facilities. The country of origin is not explicitly stated but implied to be South Korea, given the submitter's address. The data is prospective as samples were "processed within 4 hours."
   • Precision/Reproducibility: Two levels of commercially available urine-based control solutions.
     • Sample Size: 90 replicates for within-run (10 tests from 3 lots at 3 sites) and 90 replicates for within-day (1 test a day from 3 lots, at 3 sites for 10 days) for each level.
     • Provenance: Commercially available control solutions.
   • Linearity/Assay Reportable Range:
     • Sample Size: 90 replicates per concentration level (10 replicates with each of 3 lots of test strips).
     • Provenance: Samples created by spiking known concentrations of standard materials or by serial dilution of a high concentration urine sample with negative urine.
   • Detection Limit:
     • Sample Size: 90 replicates for each concentration (each sample concentration analyzed 30 times using 3 reagent strip lots).
     • Provenance: Negative urine spiked with standard materials.
   • Analytical Specificity:
     • Sample Size: 3 replicates using 3 lots of DUS 10 test strips for each concentration level of interfering substance.
     • Provenance: Urine sample pools prepared at 3 analyte concentrations (negative, low, high positive) spiked with potential interfering substances.

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

   • Precision, Linearity, Detection Limit, Analytical Specificity: Ground truth was established by the known concentrations of control solutions or spiked samples. Testing was performed by 3 medical technicians as reported in the linearity study and precision study, but their specific qualifications (e.g., years of experience) are not stated beyond being "medical technicians."
   • Method Comparison Study: The ground truth for the method comparison study was established by comparing the DUS 10 test strips results against a predicate device (Multistix 10SG). The testing was performed by three medical technicians at each of the clinical sites. Similar to above, their specific qualifications are not detailed beyond "medical technicians."

3. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

   • The document does not describe an explicit adjudication method (like 2+1 or 3+1) involving multiple human readers or experts resolving discrepancies for the test set in the traditional sense of image-based AI studies.
   • The method comparison study compares the new device's readings to the predicate device's readings, and the percentage agreement is calculated. The process for resolving discrepancies between the new device and the predicate device is not detailed, nor is there a mention of an expert panel reviewing cases.
   • For other studies (precision, linearity, detection limit), the ground truth is analytically determined by control concentrations.

4. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

   • No, a multi-reader, multi-case (MRMC) comparative effectiveness study as typically understood for AI-assisted diagnostic tools (i.e., human readers with and without AI assistance) was not explicitly performed or described in this document.
   • The studies focus on the performance characteristics of the device itself (the reagent strips) and its comparison to a predicate device, which is also a reagent strip for visual reading. The device's use is "visually read" by operators, but the study design is not one of AI assistance to human readers.

5. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:

   • This device is a reagent strip that is visually read. It is not an AI algorithm. Therefore, the concept of "standalone (algorithm only without human-in-the-loop performance)" does not directly apply in the context of an AI device. The performance data presented (precision, linearity, detection limits, analytical specificity, method comparison) are essentially standalone performance characteristics of the physical reagent strip, with human visual interpretation being the intended mode of operation for reading the results. The method comparison specifically assesses this human-read performance against a predicate device.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

   • Method Comparison Study: The ground truth for comparative analysis was the results obtained from the predicate device (Multistix 10SG).
   • Precision, Linearity, Detection Limit, Analytical Specificity: The ground truth was analytically determined concentrations derived from commercial control solutions or spiked urine samples with known analyte levels.

7. The sample size for the training set:

   • This document is for a traditional in-vitro diagnostic (IVD) reagent strip that is visually read, not an AI/ML device. Therefore, there is no "training set" in the context of machine learning model development. The assays are based on chemical reactions, and the performance is inherent to the chemical formulation and strip manufacturing.

8. How the ground truth for the training set was established:

   • As there is no AI/ML component or "training set" in the context of an algorithm, this question is not applicable. The chemical reactions on the strips are designed to react to specific analytes, and their performance is validated through the studies described.

Ask a specific question about this device