Search Results

The HALYARD* FLUIDSHIELD* N95 Particulate Filter Respirator and Surgical Mask family is intended for use by operating room personnel and other health care workers to protect both the patients and the health care workers from transfer of microorganisms, blood and body fluids, and airborne particulate materials.

Respirator consisting of nonwoven inner facing, filter media(s), a fluid barrier film, and an outer facing. It covers the nose and mouth of the wearer and is held in place with two synthetic elastic headbands, conforming to the curvature of the wearer's nose with a malleable nosepiece

The provided text is a 510(k) Summary for a medical device (HALYARD* FLUIDSHIELD* N95 Particulate Filter Respirator and Surgical Mask family). It describes the device, its intended use, and its performance against pre-defined acceptance criteria through non-clinical testing. It does not involve any clinical studies, expert-based ground truth, or MRMC comparative effectiveness studies as it's a device clearance for a respirator mask, not an AI/software-based medical device.

Therefore, many of the requested elements are "Not Applicable" (NA) for this type of submission.

Here's the information extracted from the text:

1. A table of acceptance criteria and the reported device performance:

Ask a specific question about this device

ProxiDiagnost N90 / Precision CRF is a multi-functional general R/F system. It is suitable for all routine radiography and fluoroscopy exams, including specialist areas like angiography or pediatric work, excluding mammography.

Same as the legally marketed predicate device ProxiDiagnost N90 (K212837, Substantial Equivalent (SE) date on September 21, 2021), the proposed ProxiDiagnost N90 / Precision CRF is a multi-functional general Radiography/ Fluoroscopy (R/F) system. It is suitable for all routine radiography and fluoroscopy exams, including specialist areas like angiography or pediatric work, excluding mammography.

Same as the legally marketed predicate device ProxiDiagnost N90 (K212837, SE date on September 21, 2021), the proposed ProxiDiagnost N90 / Precision CRF is a nearby controlled fluoroscopy system in combination with high-end digital radiography system consisting of a floor-mounted tilt adjustable patient support table and a scan unit consisting of a tube and a flat panel dynamic detector, Pixium FE4343F, for the fluoroscopy examinations. The tabletop can be moved by a motor in the lateral and longitudinal direction and can be tilted at -85° to +90° degrees. The scan unit tilts with the table and can be moved in the longitudinal and lateral direction, relative to the table and to the patient. The system is suitable for routine X-ray examinations and fluoroscopy examinations on patients in standing, seated, or lying positions. Same as the legally marketed predicate device ProxiDiagnost N90 (K212837, SE date on September 21, 2021), the proposed ProxiDiagnost N90 / Precision CRF retrieves images by means of a Cesium Iodide flat panel detector.

Same as the legally marketed predicate device ProxiDiagnost N90 (K212837, SE date on September 21, 2021), the proposed ProxiDiagnost N90 / Precision CRF consists of the Basic unit ("geometry" or "table unit"), Workstation Eleva Workspot (with integrated generator control, hand switch, keyboard, mouse, touch screen and PC), dual screen-monitor, Spot film device (digital camera or flat panel detector). Fixed Detector (Fluoroscopy). X-ray Generator for R/F applications, X-ray tube assembly. The optional components like wireless portable detectors small and large, Bucky tray for wireless portable detectors SkyPlate detector, Ceiling Suspension, Fixed Vertical stand, Ceiling Suspension for monitors, monitor trolley, Remote control for R/F (Radiography-fluoroscopy) viewer, accessories for "Stitching Stand", are also available.

Same as the legally marketed predicate device ProxiDiagnost N90 (K212837, SE date on September 21, 2021), the Eleva software of the proposed ProxiDiagnost N90 / Precision CRF is based on a workstation i.e., Eleva Workspot (computer, keyboard, display, and mouse) that is used by an operator to preset examination data and to generate, process and handle digital x-ray images. The Eleva Software system is decomposed into software components. These components are clustered in three component collections like the image handling focused Backend (BE), the acquisition focused Front-end (FE) and Image Processing (IP). The Eleva software is intended to acquire, process, store, display and export digital fluoroscopy and radiographic images.

The accessories for the proposed ProxiDiagnost N90 / Precision CRF are the same as the predicate device ProxiDiagnost N90 (K212837).

The list of the accessories for the proposed ProxiDiagnost N90 / Precision CRF:

• Footrest ●
• Hand Grips 0

Radiation Protection Accessories

• Flexible Radiation Protection Apron 0

• Front Radiation Protection Apron ●
  Additional Accessories (Optional)

• Monitor Trolley ●

• Monitor Ceiling Suspension

• Parking Frame for Accessories ●

• Shoulder Support ●

• Side bar ●

• Compression Belt ●

• Adjustable Lateral Cassette Holder

• Leg Supports

• o Infusion Bottle Holder

• Arm Support for Catheterization

• Ankle Clamps

• Overhead Hand Grip

• Adult Headrest

• Mattress

• Rotatable Stool for Footrest

• Pediatric Micturition Set

• Stretch Grip for Wall Stand

• Bar Code Scanner o

• Patient Support ●

• o Stitching Ruler

Accessories for the SkyPlate Detector (Optional)

• Mobile Detector Holder ●
• Detector Holder Patient Bed ●
• Portable Panel Protector ●
• Detector Handle ●
• WPD Bags ●
• o Grids for SkyPlate Detector large

The Components for the proposed ProxiDiagnost N90 / Precision CRF are the same as the predicate device ProxiDiagnost N90 (K212837).

The list of the Components for the proposed ProxiDiagnost N90 / Precision CRF:

• Eleva Workspot and RF Viewer ●
• UPS for Eleva Workspot (Optional) ●
• Table ●
• Indication Box ●
• Foot Switch
• Ceiling Suspension Motorized CSM3 (Optional) ●
• Wall Stand (Vertical Stand VS2) (Optional)
• SkyPlate / Portable Detector (Optional) 0

The proposed device complies to 'Guidance for the Submission of 510(k)'s for Solid State Xray Imaging Devices, dated September 1, 2016'. The solid-state imaging components including the detector in the proposed device have the same physical, functional, and operational characteristics as the predicate device (K212837). Also, other image chain components like Xray tube and generator, which are used for exposure characteristics and clinical performance evaluation remains the same. Hence all the features and characteristics potentially influencing image quality of the proposed are in accordance with FDA guidance document. Additionally, image quality testing has been performed on the proposed device for the changes that are affecting the image quality.

The information provided is about a regulatory submission (510(k)) to the FDA for a medical device called ProxiDiagnost N90 / Precision CRF. This type of submission aims to demonstrate that a new device is "substantially equivalent" to an already legally marketed predicate device. Therefore, the "study" referred to is primarily focused on demonstrating this equivalence rather than a traditional clinical study proving novel efficacy.

Here's an analysis 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 document describes the acceptance criteria in terms of compliance with various FDA-recognized standards and guidance documents, and the reported device performance is that it "Passes" these criteria. The safety and effectiveness of the device are deemed "Equivalent" to the predicate device.

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

The document describes non-clinical performance testing and mentions "non-clinical performance test data" and "tests performed on the proposed ProxiDiagnost N90 / Precision CRF". This type of submission relies on engineering and design verification/validation, not patient data in the sense of a clinical trial. Therefore, there is no information on a "test set" sample size in terms of patient data or data provenance (country of origin, retrospective/prospective). The testing was conducted on manufactured devices and their components.

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

Again, for this type of non-clinical submission, there is no mention of experts establishing a "ground truth" for a test set in the clinical image interpretation sense. The "ground truth" for non-clinical performance refers to the device's adherence to engineering specifications and regulatory standards, which are verified through various tests and reports.

4. Adjudication Method:

Given that this is not a study involving human reader interpretation of images for diagnosis, there is no adjudication method described.

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

No, a MRMC comparative effectiveness study was not done. The submission explicitly states: "There is no clinical data submitted in this 510(k) premarket notification." The purpose is to demonstrate substantial equivalence to a predicate device, primarily through non-clinical performance and a comparison of technological characteristics.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study:

Based on the information, this is an X-ray system, not an AI algorithm for image interpretation. The "Eleva software" is described as used by an operator to "preset examination data and to generate, process and handle digital x-ray images." While software verification testing was done, it pertains to the functionality of the system's control and image processing, not a standalone diagnostic algorithm. Therefore, a standalone AI algorithm study was not performed.

7. Type of Ground Truth Used:

For the non-clinical performance testing, the "ground truth" is effectively the engineering specifications and the requirements outlined in the FDA-recognized consensus standards and guidance documents. For example, the "System Verification testing" aims to confirm that the "system conforms to the system requirements," which serve as the ground truth for that specific test.

8. Sample Size for the Training Set:

Not applicable. As this is not an AI diagnostic algorithm, there is no "training set" in the context of machine learning. The device is a traditional X-ray imaging system with integrated software for image acquisition and processing.

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

Not applicable for the same reasons as point 8.

Ask a specific question about this device

The N9+ (OTC) device, combined with a smartphone application, is a multiple-sensor device that is intended to measure, record, and transmit the recorded data of auscultation sound of human body temperature, oxygen saturation (SpO2) and pulse rate, and images of an examined body part.

N9+ (OTC) device is intended for use by adult lay users in a non-clinical environment. Nonagon application enables counseling with the physician and transmission of the information over an IP network.

The oxygen saturation (SpO2) sensor of the device is for Rx only.

The N9+ (OTC) (except for the oxygen saturation function), is a home use device. The device is comprised of a handheld unit that interfaces with a smartphone, a wireless otoscope kit comprised of the otoscope camera, a reusable tongue depressor and 2 reusable ear specula (adult and toddlers) and a dedicated smartphone software application that runs on the user's personal smartphone.

The system uses the smartphone's camera to record images of the patient's skin.

An API is defined to enable healthcare providers to communicate with the device. The pulse oximeter will be a prescription device in the US market.

The Device enables user's examination at home with or without the quidance of a remote physician. It also enables guidance and data examination by a remote physician over an IP network. The Smartphone Application controls the Handheld Unit and wireless otoscope functionality (besides recording images of the patient's skin) and processes and displays the collected data.

The provided text describes the regulatory clearance of the N9+ (OTC) device and details some performance data, but it does not contain the specific information requested in points 1, 2, 3, 4, 5, 6, 7, 8, and 9 regarding the acceptance criteria and the study that proves the device meets those criteria for the AI/algorithm component.

The document discusses general performance data related to biocompatibility, software validation, electrical safety, EMC, wireless quality of service, and self-selection testing. It does not provide performance metrics (like accuracy, sensitivity, specificity, AUC) for the device's functional capabilities (e.g., measuring auscultation sounds, temperature, SpO2, or image analysis).

Therefore, I cannot populate the requested table or answer the specific questions about acceptance criteria for device performance as an AI/algorithm.

The document does mention software validation was conducted and the software level of concern is moderate, but it does not specify the performance metrics or studies directly related to the accuracy or effectiveness of the data collection or image acquisition beyond general compliance.

Information provided in the document that somewhat relates to the request:

• Software Validation: "Software verification and validation testing were conducted, and documentation was provided as recommended by FDA's Guidance for Industry and FDA Staff, "Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices". The N9+ (OTC) software level of concern is moderate." (This confirms software validation was done, but gives no specifics on acceptance criteria or performance metrics).
• Self-Selection Testing: "The study enrolled 31 men and women of different ages and literacy levels. All the participants made a correct purchasing decision. The results of this self-selection study show that consumers are able to decide whether the N9+ (OTC) is right for them based on the information communicated on the outer package (box)." (This is about user understanding of the product, not the device's functional performance.)

The document does NOT include:

1. A table of acceptance criteria and reported device performance.
2. Sample size used for the test set and data provenance for an AI/algorithm.
3. Number of experts used to establish ground truth or their qualifications for an AI/algorithm.
4. Adjudication method for the test set for an AI/algorithm.
5. Information on a multi-reader multi-case (MRMC) comparative effectiveness study or effect size of human reader improvement with AI.
6. Standalone performance of an algorithm.
7. Type of ground truth used for an AI/algorithm.
8. Sample size for the training set for an AI/algorithm.
9. How the ground truth for the training set was established for an AI/algorithm.

Ask a specific question about this device

Consona N9, Consona N9 Pro, Consona N9 Super, Consona Nova, Consona N9S, Consona NI, Consona NT, Nuewa N9, Recho N9, Consona N9 Exp, Consona N9 Elite, Resona N9, Consona N9T, Consona N8, Consona N8 Pro, Consona N8 Super, Consona OR, Consona N8S, Consona OI, Consona OT, Consona N8 Exp, Consona N8 Elite, Consona N8T, Consona N7, Consona N7 Pro, Consona N7 Super, Consona VR, Consona N7S, Consona VI, Consona VT, Consona N7 Exp, Consona N7 Elite, Consona N7T Diagnostic Ultrasound System is applicable for adults, pregnant women, pediatric patients and neonates.

It is intended for use in fetal, abdominal, Intra-operative, small organ(breast, thyroid, testes), neonatal cephalic.adult cephalic.trans-rectal, trans-vaginal, musculo-skeletal(conventional), musculo-skeletal(superficial), Thoracic Pleural, cardiac adult, cardiac pediatric, Trans-esoph., peripheral vessel and urology exams.

Modes of operation include: B, M, PWD, Color Doppler, Amplitude Doppler, Combined mode(B+M, PW+B, Color+B, Power+B, PW+Color+B, Power+PW+B), Tissue Harmonic Imaging, Smart 3D, 4D(Real-time 3D), iScape View, TDI, Color M, Strain Elastography, STE, STQ, Contrast imaging (Contrast agent for LVO), Contrast imaging (Contrast agent for Liver).

This device is a general purpose diagnostic ultrasound system intended for use by qualified and trained healthcare professionals for ultrasound imaging,measurement, display and analysis of the human body and fluid, which is intended to be used in a hospital or medical clinic.

Consona N9, Consona N9 Pro, Consona N9 Super, Consona Nova, Consona N9S, Consona NI, Consona NT, Nuewa N9, Recho N9, Consona N9 Exp, Consona N9 Elite, Resona N9, Consona N9T, Consona N8, Consona N8 Pro, Consona N8 Super, Consona OR, Consona N8S, Consona OI, Consona OT, Consona N8 Exp, Consona N8 Elite, Consona N8T, Consona N7, Consona N7 Pro, Consona N7 Super, Consona VR, Consona N7S, Consona VI, Consona VT, Consona N7 Exp, Consona N7 Elite, Consona N7T Diagnostic Ultrasound System is a general purpose, mobile, software controlled, ultrasonic diagnostic system. Its function is to acquire and display ultrasound images in Modes of operation include : B, M, PWD, CWD, Color Doppler, Amplitude Doppler, Combined mode(B+M, PW+B, Color+B, Power+B, PW+Color+B, Power+PW+B), Tissue Harmonic Imaging, Smart 3D, 4D(Real-time 3D), iScape View, TDI, Color M, Strain Elastography, STE, STQ, Contrast imaging (Contrast agent for LVO), Contrast imaging (Contrast agent for Liver). Subject device can also measure anatomical structures and offer analysis packages to provide information based on which the competent health care professionals can make the diagnosis.

Here's an analysis of the provided text regarding acceptance criteria and the study that proves the device meets them:

1. Table of Acceptance Criteria and Reported Device Performance

The provided document does not explicitly state quantitative acceptance criteria or reported device performance metrics for the Consona N9/N8/N7 series Diagnostic Ultrasound System. The submission is a 510(k) premarket notification for substantial equivalence, not a performance study against specific acceptance criteria.

The document primarily focuses on demonstrating that the subject device is substantially equivalent to legally marketed predicate devices (primarily the Mindray MX7, K212900) in terms of intended use, imaging modes, features, functions, and technological characteristics.

Instead of a table of specific performance metrics, the document highlights:

• Indications for Use: The subject device's indications are comparable to the predicate.
• Patient Contact Materials: Same as predicate or tested under ISO 10993-1.
• Acoustic Power Levels: Below FDA limits and same as the predicate (MX7).
• Electrical and Physical Safety Standards: Complies with FDA recognized standards, same as the predicate (MX7).
• Features and Functions: Same as predicate devices, with new options being either improvements or enhancements of cleared functions, with no new intended uses or clinical risks introduced.

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

The document explicitly states: "Not applicable. The subject of this submission... does not require clinical studies to support substantial equivalence."

Therefore, there is no test set, no sample size, and no data provenance (e.g., country of origin, retrospective/prospective) related to clinical performance for this submission. The evaluation is based on non-clinical tests and comparison to predicate devices.

3. Number of Experts Used to Establish Ground Truth and Qualifications

As there were no clinical studies conducted and no test set, this information is not applicable. Ground truth for clinical performance was not established by experts for this particular submission.

4. Adjudication Method for the Test Set

Not applicable, as no clinical test set was used for performance evaluation.

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

No MRMC study was done, as the document states that clinical studies were not required and were not performed. Therefore, there's no reported effect size of human readers improving with or without AI assistance.

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

No standalone performance study of an algorithm was conducted or reported, as this is a general-purpose diagnostic ultrasound system, not an AI-driven diagnostic algorithm. The new features like "Smart Calc" are described as semi-automatic tools that assist the user, implying a human-in-the-loop design, but no performance study specific to this feature is detailed.

7. Type of Ground Truth Used

Given the absence of clinical studies, there is no mention of ground truth established through expert consensus, pathology, or outcomes data. The basis for substantial equivalence relies on technical specifications, compliance with safety standards, and functional comparison to predicate devices, rather than clinical performance data against a defined ground truth.

8. Sample Size for the Training Set

Not applicable. This submission concerns a diagnostic ultrasound system and does not involve an AI model with a distinct "training set" in the conventional sense of machine learning algorithms requiring large datasets for training. The "new options" like Smart Calc are described as improvements or enhancements but not as standalone AI algorithms whose performance requires training and validation datasets.

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

Not applicable, as there was no training set for an AI model.

Ask a specific question about this device

ProxiDiagnost N90 is a multi-functional general R/F system. It is suitable for all routine radiography and fluoroscopy exams, including specialist areas like angiography or pediatric work, excluding mammography.

The ProxiDiagnost N90 is a multi-functional general Radiography/ Fluoroscopy (R/F) system. It is suitable for all routine radiography and fluoroscopy exams, including specialist areas like angiography or pediatric work, excluding mammography.

The ProxiDiagnost N90 is a nearby controlled fluoroscopy system in combination with high-end digital radiography system consisting of a floor-mounted tilt adjustable patient support table and a scan unit consisting of a tube and a flat panel dynamic detector, Pixium FE4343F, for the fluoroscopy examinations. The tabletop can be moved by a motor in the lateral and longitudinal direction and can be tilted by -85° to +90° degrees. The scan unit tilts with the table and can be moved in the longitudinal and lateral direction, relative to the table and to the patient. The system is suitable for routine X-ray examinations and fluoroscopy examinations on patients in standing, seated or lying positions. The ProxiDiagnost N90 retrieves images by means of a Cesium Iodide flat panel detector.

Philips fluoroscopy systems (standard configuration) consist of the Basic unit ("geometry" or "table unit"), Workstation Eleva Workspot (with integrated generator control, hand switch, keyboard, mouse, touch screen and PC), dual screen-monitor, Spot film device (digital camera or flat panel detector), Fixed Detector (Fluoroscopy), X-ray Generator for R/F applications, X-ray tube assembly. The optional components like wireless portable detectors small and large, Bucky tray for wireless portable detectors (SkyPlate) detector, Ceiling Suspension, Fixed Vertical stand, Ceiling Suspension for monitors, monitor trolley, Remote control for R/F (Radiography-fluoroscopy) viewer, accessories for “Stitching Stand", are also available.

The Eleva software of the proposed ProxiDiagnost N90 is based on a workstation i.e., Eleva Workspot (computer, keyboard, display, and mouse) that is used by an operator to preset examination data and to generate, process and handle digital x-ray images. The Eleva Software system is decomposed into software components. These components are clustered in three component collections like the image handling focused Back-end (BE), the acquisition focused Front-end (FE) and Image Processing (IP). The Eleva software is intended to acquire, process, store, display and export digital fluoroscopy and radiographic images.

The ProxiDiagnost N90 uses the same workflow from the currently marketed and predicate device, ProxiDiagnost N90 (K173433) with only the following modifications:

• Inclusion of Extended reviewing options (like the optional reference monitor & remote control),
• Inclusion of some image processing features
• Updates to Operating system and Eleva application Software to include state-of-art operating system and incorporate the changes
• Replacement of the ceiling suspension with that of reference device, DigitalDiagnost C90 (K202564)
• Updates to improve serviceability
• Option for upgradability of Predicate device (K173433) to include the above changes

The provided text is a 510(k) summary for the Philips ProxiDiagnost N90, an X-ray system. It primarily focuses on demonstrating substantial equivalence to a predicate device, rather than providing a detailed study proving the device meets specific acceptance criteria through an AI/human comparative effectiveness study or a standalone algorithm performance study.

The document does not describe acceptance criteria for an AI algorithm or a study proving an AI algorithm meets those criteria. Instead, it outlines the changes from a predicate device (K173433) and refers to the performance of other reference devices (K203087 and K202564) as justification for the modifications. The "acceptance criteria" discussed are in the context of device safety and effectiveness testing for a conventional medical device (X-ray system), aligning with recognized standards and guidance documents (e.g., IEC 60601 series, ISO 14971).

Therefore, I cannot provide the requested information regarding AI acceptance criteria and performance study details from the given text.

However, I can extract information about the overall device acceptance criteria and testing methodology as described for this X-ray system:

Overall Device Acceptance Criteria (Implied by Testing):

The acceptance criteria for the ProxiDiagnost N90 are implicitly demonstrated through adherence to various international standards and FDA guidance documents related to X-ray systems, electrical safety, electromagnetic compatibility, radiation protection, usability, software lifecycle processes, risk management, and biological evaluation. The testing performed is aimed at ensuring the device's safety and effectiveness compared to its predicate and reference devices, despite the noted modifications.

Study Proving the Device Meets Acceptance Criteria (as described in the document):

The "study" described is a series of non-clinical performance tests and verifications rather than a comparative clinical study with human readers or standalone AI performance.

Information Extracted from the Document (to the extent possible given the context):

1. A table of acceptance criteria and the reported device performance:

   The document does not present a table of specific quantitative performance acceptance criteria for an AI algorithm or human reading performance. Instead, it states that "Tests were performed on the proposed ProxiDiagnost N90 according to the following FDA recognized standards and guidance documents." The reported "performance" is that these tests support the device being "safe and effective" and "substantially equivalent" to the predicate.

   Acceptance Criterion Category (Implied)	Reported Device Performance/Verification Method
   General Safety & Performance	- Compliance with ANSI AAMI ES60601-1:2005/(R)2012 And A1:2012 (Medical electrical equipment - Part 1: General requirements for basic safety and essential performance).

• Compliance with IEC 60601-1-2 Edition 4.0 2014-02 (Electromagnetic disturbances - Requirements and tests).
• Compliance with IEC 60601-1-3 Edition 2.1 2013-04 (Radiation protection in diagnostic X-ray equipment).
• Compliance with IEC 60601-1-6 Edition 3.1 2013-10 (Usability).
• Compliance with IEC 60601-2-54 Edition 1.1 2015-04 (Particular requirements for the basic safety and essential performance of X-ray equipment for radiography and radioscopy).
• Compliance with ANSI AAMI ISO 14971: 2007/(R)2010 (Medical devices-Application of risk management to medical devices).
• Compliance with ISO 10993-1, Fifth edition 2018-08 (Biological evaluation of medical devices).
• System and software verification testing was performed for all modifications to demonstrate safety and effectiveness. |
  | New Features Performance | - Extended Reviewing Options: System Verification for Bluetooth remote control and additional reference monitor (test protocol identical to CombiDiagnost R90 K203087).
• Image Processing Features:
  • Digital Subtraction Angiography: Sub-system (Eleva software) & System Verification (test protocol identical to CombiDiagnost R90 K203087).
  • Predefined annotations: Sub-system (Eleva software) & System Verification (test protocol identical to CombiDiagnost R90 K203087).
  • Bone Suppression: Sub-system (Eleva software) & System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • UNIQUE 2 (radiography): Sub-system (Eleva software) & System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • Intuitive User Interface for Processing Parameters: Sub-system Verification (Eleva software) (test protocol identical to DigitalDiagnost C90 K202564).
  • Deviation and Target Exposure Indices: Sub-system (Eleva software) & System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • Update of optional Skyflow feature: Sub-system (Eleva software) & System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • Access to and Export of Original Image Data: System Verification.
  • Improved OBSA: Sub-system (Eleva software) & System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • View Selection for Changed X-Ray Generation Data Sets: Sub-system Verification (Eleva software) (test protocol identical to DigitalDiagnost C90 K202564).
  • Avoid Ghosting in Verification Images of Portable Detectors: System Verification (test protocol identical to DigitalDiagnost C90 K202564).
• Software Updates:
  • Operating system upgrade to Microsoft Windows 10: Sub-system (Eleva software) & System Verification (test protocol identical to CombiDiagnost R90 K203087).
  • Upgrade of Eleva Application software to increment 42: All relevant software functions tested at system and subsystem level (referencing tests for change #1, 2, 4 and 5).
• Ceiling Suspension & Service Features:
  • Tube head control: System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • Collimator: System Verification (test protocol identical to DigitalDiagnost C90 K202564).
  • Monitoring and Firmware Updates for Field Service: System Verification (test protocol identical to CombiDiagnost R90 K203087).
  • Remote access for the field service Engineer: Sub-system Verification (Eleva software).
  • Service Diagnostic: System Verification (test protocol identical to CombiDiagnost R90 K203087 & DigitalDiagnost C90 K202564).
  • Hardware upgrades (Alpha drive Upgradeability): System Verification (test protocol identical to DigitalDiagnost C90 K202564). |
    | Upgradeability of Predicate Device | All relevant Software functions are tested at system and subsystem level (referencing tests for change #1, 2, 3 and 5 a,b,c). |

2. Sample sizes used for the test set and the data provenance:

   • Test Set Sample Size: Not specified in terms of number of patient cases or images. The testing appears to be primarily system-level, software-level, and component-level verification, rather than evaluation on a diagnostic image dataset.
   • Data Provenance: The document explicitly states "There is no clinical data submitted in this 510(k) premarket notification." Therefore, there is no information on country of origin or retrospective/prospective nature of data for clinical evaluation, as none was performed.

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

   Not applicable. No clinical data or expert-established ground truth for diagnostic image interpretation was used or provided in this 510(k) submission.

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

   Not applicable, as no clinical image evaluation requiring adjudication was described.

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 MRMC study was performed or described. This submission is for an X-ray system, not an AI-powered diagnostic algorithm requiring such a study for its clearance.

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

   No standalone algorithm performance study was performed or described, as this submission is for an X-ray system, not a standalone AI diagnostic algorithm.

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

   Not applicable. No clinical ground truth (expert consensus, pathology, outcomes data) was used in this 510(k) submission, as it explicitly states "There is no clinical data submitted." The testing relies on engineering and regulatory compliance standards.

8. The sample size for the training set:

   Not applicable. The document describes an X-ray imaging system, not an AI model that would require a training set.

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

   Not applicable, as there is no training set for an AI model discussed.

Ask a specific question about this device

The N95 Particulate Respirator (MF01) is single-use, disposable device, provided non-sterile, and are intended to be worn by operating room personnel or other healthcare workers to protect and operating room personnel from the transfer of microorganisms, body fluids, and particulate material.

The N95 particulate respirator (Model MF-01) is NIOSH certified (TC-84A-7697), pouched-shaped respirators when worn. The flat-folded respirator is composed of four layers-inner layer, filter (2 layers) and outer layer. The four layers are combined via ultrasonically welded. The respirator also contains a nose-piece enclosed in a binding tape welding the top edge to conform to the contours of the face. The respirator contains straps to secure the respirator in place on the wearer. The respirator is offered in white. The respirator is provided non-sterile and for single use only.

This document is a 510(k) Premarket Notification for a Master-Frank N95 Particulate Respirator. It focuses on demonstrating substantial equivalence to a predicate device, rather than presenting a study of diagnostic AI. Thus, much of the requested information (like sample sizes for test/training sets, expert qualifications, adjudication methods, MRMC studies, standalone algorithm performance, and ground truth types for AI models) is not applicable or present in this document.

However, I can extract the acceptance criteria and reported device performance from the "Non-Clinical Performance Data" and "Performance testing" sections, which are focused on the physical performance and biocompatibility of the respirator.

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample size used for the test set and the data provenance

The document does not specify exact sample sizes for each of the biocompatibility and physical performance tests. It refers to established standards (e.g., ISO 10993, ASTM F1862, 16 CFR 1610, NIOSH certification), which would have their own specified sample sizes. The data provenance is from tests conducted on the Master-Frank N95 Particulate Respirator (Model MF-01) by accredited labs for NIOSH certification and other standards. The document doesn't specify the country of origin of the testing data itself, but the manufacturer is based in China. The data is prospective as it pertains to the device being submitted for clearance.

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

Not applicable in the context of this device clearance. The "ground truth" for these tests is defined by the objective measurement criteria within the specified international standards and NIOSH certification protocols. These tests are conducted by trained technicians in certified laboratories, following stringent procedures rather than subjective expert interpretation.

4. Adjudication method for the test set

Not applicable. The tests performed are objective, quantitative measurements following standardized protocols (e.g., measuring filter efficiency, fluid penetration, flame spread). There is no "adjudication" in the sense of reconciling differing expert opinions. The results are either "pass" or "fail" against defined thresholds.

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 document is for the clearance of an N95 particulate respirator, not an AI-powered diagnostic device. Therefore, no MRMC study involving human readers or AI assistance was performed or reported.

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

Not applicable. This device is a physical respirator, not an algorithm.

7. The type of ground truth used

The "ground truth" for the performance claims of the N95 respirator is based on objective, standardized physical and biocompatibility testing criteria. Examples include:

• Physical Measurements: Filtration efficiency (% of particles filtered), differential pressure (airflow resistance), fluid penetration (absence of synthetic blood penetration at a given pressure), and flame spread time.
• Biocompatibility Assessments: Observational and analytical results from in vitro cytotoxicity assays, irritation tests (e.g., skin irritation), and sensitization tests (e.g., skin sensitization) performed according to ISO 10993 standards.

These are verifiable, objective measures against international standards and regulatory requirements.

8. The sample size for the training set

Not applicable. There is no training set as this is not an AI/machine learning device. The device is a physical product designed and manufactured to meet specific performance standards.

9. How the ground truth for the training set was established

Not applicable. As there is no training set for an AI model, the concept of establishing ground truth for a training set does not apply here. The device's performance is established through direct physical and biological testing against predefined standards.

Ask a specific question about this device

ProxiDiagnost N90 is a multi-functional general R/F system. It is suitable for all routine radiography and fluoroscopy exams, including specialist areas like angiography or pediatric work, excluding mammography.

The ProxiDiagnost N90 is a multi-functional nearby controlled fluoroscopy system in combination with a high-end digital radiography system consisting of a floor-mounted tiltadjustable patient support table and a scan unit consisting of a tube and a flat panel dynamic detector, Pixium FE4343F, for the fluoroscopy examinations. The tabletop can be moved by a motor in the lateral and longitudinal direction and can be tilted -85° to +90° degrees. The scan unit tilts with the table and can be moved in the longitudinal and lateral direction, relative to the table and to the patient. The fully integrated system is provided with a x-ray tube(s) with collimator and high resolution displays. The ProxiDiagnost N90 is configured with a Philips x-ray generator and a flat panel dynamic detector, Pixium FE4343F, components of the Philips radiography/fluoroscopy Image Chain. As additional options, the ProxiDiagnost N90 can be used as a digital radiography system consisting of a mounted tube in a ceiling suspension together with the portable or fixed detector in the vertical stand.

Based on the provided text, the ProxiDiagnost N90 is evaluated for substantial equivalence to its predicate device, EasyDiagnost Eleva (K031535). The document outlines non-clinical data and makes a case against the need for a clinical study.

Here's an analysis of the requested information:

1. Table of acceptance criteria and the reported device performance:

The document broadly states that "The test results demonstrate that the proposed ProxiDiagnost N90 meets the acceptance criteria and is adequate for its intended use." However, specific numerical acceptance criteria or performance metrics for the ProxiDiagnost N90 are not detailed in a table. The comparison provided focuses on demonstrating technological similarities to the predicate device rather than presenting specific performance thresholds for the new device.

The "Discussion" column in the "Summary of technological characteristics" table implicitly suggests that if a characteristic is "Similar" or "Equivalent" to the predicate, it meets acceptance criteria for substantial equivalence.

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective):

The document explicitly states: "The proposed ProxiDiagnost N90 did not require a clinical study since substantial equivalence to the currently marketed and predicate device was demonstrated with the following attributes: Design features, Indication for use, Fundamental scientific technology, Non-clinical performance testing including validation; and Safety and effectiveness."

Therefore, there is no test set sample size from a clinical study for the ProxiDiagnost N90 itself. The evaluation relied on non-clinical verification and validation tests against recognized consensus standards and a comparison to the predicate and reference devices.

No information on data provenance for a clinical test set is available, as no clinical study was conducted.

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

Since no clinical study was performed and no test set with clinical data was used for evaluation, there were no experts used to establish ground truth for a test set in this submission.

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

As no clinical test set was used for evaluation of the ProxiDiagnost N90, no adjudication method was applied for a test set.

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:

There is no mention of a multi-reader multi-case (MRMC) comparative effectiveness study being conducted for the ProxiDiagnost N90. The device is a general R/F system, not an AI-assisted diagnostic tool, so such a study would not be relevant in this context.

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

The device is an imaging system, not an algorithm in the AI sense. Its "performance" would relate to image quality and system functionality, which was assessed through non-clinical performance testing (verification and validation tests according to standards like IEC 60601-1, IEC 60601-1-2, IEC 60601-1-3, IEC 60601-2-54, IEC 62220-1). These tests are typically "standalone" in that they evaluate the device's technical specifications and functionality independent of direct human interpretation studies.

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

For the non-clinical performance and validation tests, the "ground truth" would be established by the specified engineering tolerances, recognized international standards (e.g., IEC standards for image quality, safety, and performance), and the performance characteristics of the predicate and reference devices. For instance, the MTF and DQE values for the detector are reported and implicitly "compared" to cleared components, meaning the "ground truth" for the acceptance of these values is their alignment with previously cleared and accepted digital imagers.

8. The sample size for the training set:

As no clinical study was performed and the device is an imaging system (not an AI model requiring a training set), there is no training set sample size.

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

Since no training set was used, ground truth for a training set was not established.

Ask a specific question about this device

The N95 Particulate Respirator and Surgical Mask, Models TN01-12 are single-use, disposable devices, provided non-sterile, and are intended to be worn by operating room personnel or other healthcare workers to protect both the patient and operating room personnel from the transfer of microorganisms, body fluids, and particulate material.

The N95 Particulate Respirator and Surgical Mask, Models #TN01-11 and #TN01-12 are NIOSH certified (TC 84A-3348), pouchedshaped respirators when worn. The flat-folded masks are composed of four layers of materials consisting of polypropylene and polyethylene (inner layer), polypropylene meltblown (two filter layers), and polypropylene spunbond (outer layer). The masks contain a conformable nose clamp enclosed in a binding tape welding the top edge to conform to the contours of the face. In addition, the masks contain an ultrasonically welded, polyurethane elastic headband not made with natural rubber latex, to secure the masks in place on the wearer. The masks are offered in orange or white.

The provided document describes the acceptance criteria and performance of the N95 Particulate Respirator and Surgical Mask, Models TN01-11 and TN01-12, in comparison to a predicate device.

Here's a breakdown of the requested information:

1. A table of acceptance criteria and the reported device performance

2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

The document does not explicitly state the specific sample sizes used for each performance test (e.g., number of masks tested for fluid resistance or flammability). It simply states that the models meet the requirements of the standards. The provenance of the data is not specified beyond indicating it was a submission to the FDA. The tests are described as performance testing and non-clinical tests, which are typically prospective.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g., radiologist with 10 years of experience)

This information is not applicable to the type of device and testing described. The "ground truth" for N95 masks is established through standardized laboratory performance tests according to recognized standards (e.g., ASTM, NIOSH, ISO) and not through expert consensus or interpretation of medical images.

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

This information is not applicable. Adjudication methods like 2+1 are typically used in studies where human readers interpret medical data (e.g., radiology studies) and their findings need to be reconciled. For performance testing of a physical device like an N95 mask, objective laboratory measurements are taken against predefined criteria.

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

This information is not applicable. The device is a physical N95 mask, not an AI-powered diagnostic tool. Therefore, MRMC studies and "human readers improving with AI" are irrelevant to this document.

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

This information is not applicable. The device is a physical N95 mask; there is no algorithm or AI component involved. The testing described is for the physical performance of the mask.

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

The ground truth for the performance of the N95 mask is based on objective measurements against established performance standards set by regulatory bodies and testing organizations. These include:

• ASTM F1862 for Fluid Resistance
• NIOSH Certification (TC 84A-3348) for Particulate Filtration Efficiency, Bacterial Filtration Efficiency, and Differential Pressure
• 16 CFR 1610 for Flammability
• ISO 10993 series for Biocompatibility (cytotoxicity, irritation, sensitization)

8. The sample size for the training set

This information is not applicable. There is no "training set" in the context of performance testing for a physical medical device like an N95 mask. Training sets are relevant for machine learning or AI models.

9. How the ground truth for the training set was established

This information is not applicable, as there is no training set for this type of device and study.

Ask a specific question about this device

The SpectraShield 9500 Surgical Mask is a single use, disposable, surgical mask, tested for continuous use up to 8 hours, embedded with a zeolite carrier containing a silver-copper agent on the outer layer and is not an antimicrobial drug. SpectraShield 9500 kills 99.99% of test bacteria after one hour of contact with the surface of the respirator, and inactivates 99.99% of test influenza viruses after 5 minutes of contact with the surface of the mask. In vitro (laboratory) tests have demonstrated 99.99% kill on the surface of the respirator when tested in vitro against single isolates of the following test bacteria: Streptococus pyogenes, MRSA (Methicillin Resistant Staphylococcus aureus), and Haemophilus influenzae, under tested contact conditions.

In vitro tests have demonstrated 99.99% inactivation on the surface of the mask when tested against the following seasonal, pandemic, avian, swine influenza viruses: Influenza A subtypes: H1N1 (A/ California/04/2009, A/Brisbane/59/2007, A/Wisconsin/10/1998, A/Puerto Rico/8/1934), H3N2 (A/Wisconsin/67/2005, A/Brisbane/10/2007, A/Hong Kong/8/1968, A/Victoria/3/75), H2N2 ( A/2/Japan/305/1957); Avian Flu (bird flu) subtypes: Avian Influenza A (H5N1), Avian Influenza A (H9N2), Swine Flu subtype: H1N1 (A/Swine/1976/1931), Equine Flu subtypes: H3N8 (A/Equine/2/Miami/1963), Influenza B strains: B/Florida/4/2006, B/Brisbane/60/2008, B/ Lee/40.

No clinical studies have been conducted comparing the ability of an untreated surgical mask and the SpectraShield model 9500 surgical mask to protect the wearer from infection, and the antibacterial/anti-influenza virus treatment cannot affect pathogens that are inhaled around the edges of the respirator.

The SpectraShield 9500 Surgical Mask is a single use device, intended for occupational use to protect against microorganisms, body fluids, and particulate material.

The SpectraShield model 9500 Surgical Mask is a molded shape surgical mask composed of 4 layers of material, molded to form the mask. A 2-ply meltblown polypropylene middle layer is sandwiched by inner and outer layers of 100% polyester nonwoven fabric. The inner and outside layers of polyester nonwoven fabric include fibers that have been embedded with an antibacterial/antiviral agent to provide antibacterial and antiviral performance. The mask has 2 elastic straps (not made with natural latex) and an aluminum nose strip.

The provided document is a 510(k) Summary for the SpectraShield Model 9500 Surgical Mask, describing its characteristics and comparing it to predicate devices to demonstrate substantial equivalence. It does not describe an AI medical device. Therefore, it is not possible to answer the request by extracting information about acceptance criteria for an AI device.

However, I can extract the acceptance criteria and performance data for the surgical mask as presented in the document, which are based on standard tests for such devices.

Here's the information regarding the SpectraShield Model 9500 Surgical Mask, formatted as requested for an "AI device," but adapted to fit the context of a physical medical device (surgical mask):

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample size used for the test set and the data provenance

The document does not specify exact sample sizes for each test. However, it indicates that the tests performed are in vitro (laboratory) tests. Data provenance is not explicitly stated beyond "in vitro (laboratory) tests," implying laboratory settings and not human clinical data or geographical origin. The study is a pre-market notification (510k) using performance data to demonstrate substantial equivalence to predicate devices, not a clinical study on patient outcomes.

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

This is not applicable as the tests described are in vitro laboratory tests measuring physical and microbiological properties of the mask, not evaluations requiring expert interpretation of medical images or patient data to establish ground truth.

4. Adjudication method for the test set

This is not applicable as the tests are objective laboratory measurements, not subjective evaluations requiring adjudication.

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

This is not applicable. The device is a surgical mask, not an AI diagnostic or assistive tool.

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

This is not applicable. The device is a surgical mask, not an AI algorithm. The performance data presented are for the mask itself in laboratory settings.

7. The type of ground truth used

For the physical and biological tests (fluid resistance, filtration, pressure, flammability, cytotoxicity, skin irritation, systemic toxicity, physico-chemical, gas/leach-off), the "ground truth" is defined by the objective measurement adhering to established ASTM, Mil, CFR, ISO, and USP standards.

For bacterial and viral reduction, the "ground truth" is the quantitative reduction in viable microorganisms as measured in the in vitro laboratory tests.

8. The sample size for the training set

This is not applicable. The device is a physical product (surgical mask), not an AI model that requires a training set.

9. How the ground truth for the training set was established

This is not applicable, as there is no AI training set involved. The product's characteristics are inherent to its design and manufacturing.

Ask a specific question about this device

Pasture F550S, Pasture F550CS, Pasture A520S, Pasture A520CS, Pasture E520S, and Pasture E520CS is a NIOSH certified N95 respirator, and is indicated to be used by the healthcare personnel during procedures to protect both the patient and the healthcare personnel from the transfer of microorganisms, body fluids, and particulate material.

Pasture F550S and Pasture F550CS respirators are disposable duck bill shaped N95 respirators with NIOSH certification number TC-84A-7504. They contain 5 layers composed of polypropylene, with a nose cushion, a synthetic elastic loop or strap, and a nosepiece which is the combination of zinc wires and embedded polyester inside of layers. F550CS is the same respirator as the F550S, but F550CS contains an adjustable buckle on the headband.

Pasture A520S and Pasture A520CS respirators are cup shaped N95 respirators with NIOSH certification number TC-84A-7454. They contain 4 layers composed of polypropylene with a synthetic elastic loop or a strap. A520CS is the same respirator as the A520S, but A520CS contains an adjustable buckle on the headband.

Pasture E520S and Pasture E520CS respirators are cup shaped N95 respirators with NIOSH certification number TC-84A-7453. They contain 4 layers composed of polypropylene with a nose cushion, synthetic elastic loop or strap, and an aluminum nosepiece. E520CS is the same respirator as E520S, but E520CS contains an adjustable buckle on the headband.

This document describes the premarket notification for several models of Pasture Surgical N95 Respirators (F550S, F550CS, A520S, A520CS, E520S, E520CS). The provided information is insufficient to answer all parts of the request as it pertains to medical device AI/ML performance studies. However, I can extract the relevant information from the document regarding the acceptance criteria and performance of these N95 respirators. The questions asked are typically for AI/ML devices, not for surgical respirators.

Here's an analysis based on the provided text, adapted to the context of a physical medical device (N95 respirator) and its regulatory submission.

1. A table of acceptance criteria and the reported device performance

The acceptance criteria for these N95 respirators are based on various performance standards and regulatory requirements for surgical masks and N95 respirators. The document primarily demonstrates substantial equivalence to a predicate device (WILLSON ONE-FIT HEALTHCARE PARTICLE RESPIRATOR AND SURGICAL MASK, HC-NB095) through performance testing.

The following questions are not applicable to the provided document, as it concerns a physical medical device (N95 respirator) and not an AI/ML algorithm or software as a medical device (SaMD). The document does not describe a study involving readers, ground truth establishment by experts, or training/test sets in the context of an AI/ML device.

2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

• Not Applicable. This document describes a 510(k) premarket notification for physical N95 respirators, not an AI/ML device or its test set. Performance testing is typically conducted on a sample of manufactured devices according to established test methods for physical properties and biocompatibility. The specific sample sizes for each test (e.g., fluid resistance, flammability) are not detailed in this summary but would be specified in the full test reports referenced by the NIOSH certification and ASTM standards. The provenance of the data would be from the laboratories that performed these physical tests.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

• Not Applicable. Ground truth in the context of expert consensus is relevant for diagnostic AI/ML algorithms. For N95 respirators, the "ground truth" is determined by physical and chemical test results against a defined standard (e.g., NIOSH standards, ASTM F1862). Expert qualifications would pertain to the technicians conducting these standardized tests, ensuring they follow protocols, not to clinical interpretation of data.

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

• Not Applicable. Adjudication methods like 2+1 or 3+1 are used in AI/ML performance studies to resolve discrepancies in expert interpretations to establish a robust ground truth. For N95 respirators, performance is objectively measured by standardized tests, not by expert consensus on clinical findings.

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. MRMC studies are designed to assess the impact of AI assistance on human reader performance, typically in diagnostic imaging. This document does not pertain to an AI device.

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

• Not Applicable. This question relates to the performance of an AI algorithm in isolation. The product is a physical N95 respirator.

7. The type of ground truth used (expert concensus, pathology, outcomes data, etc)

• For physical devices like N95 respirators, the "ground truth" for performance is defined by adherence to established industry standards and regulatory requirements (e.g., NIOSH N95 certification, ASTM F1862 for fluid resistance, 16 CFR 1610 for flammability, ISO 10993 for biocompatibility). The tests objectively measure specific characteristics.

8. The sample size for the training set

• Not Applicable. Training sets are used for machine learning algorithms. This document describes a hardware device.

9. How the ground truth for the training set was established

• Not Applicable. As above, training sets are not relevant to this type of device.

Ask a specific question about this device