(66 days)
The SOZO Pro may be used as an adjunct to existing methods by aiding clinicians who are using Subjective Global Assessment (SGA) tools to assess patients at risk of protein-calorie malnutrition (PCM).
The SOZO Pro may be further used to estimate the following body composition parameters in humans to track clinically relevant body composition parameters over time:
- · Fat mass
- · Fat-free mass
- Total body water
- · Intracellular fluid
- · Extracellular fluid
- Skeletal muscle mass
The following outputs are also presented:
- · Body Mass Index (BMI)
- · Basal metabolic rate (BMR; based on Mifflin St. Jeor's algorithm) displayed in calories per day
• Protein and mineral (also known as 'dry lean mass') represents the content of a body that is not fat or fluid; calculated by subtracting total body water weight from fat-free mass weight.
The SOZO Pro device measures current (1), voltage (V) and phase angle (Phi), and from these values calculates resistance (R).
reactance (Xc), and impedance (Z), which are used to estimate the above body composition parameters. The device/ software will also display the Cole plot, subject height, weight, age and sex.
The SOZO Pro system consists of a connected hand and footplate with built-in stainless-steel electrodes, paired with an Android tablet over Bluetooth connection. An app ("SOZOapp"), supplied with the tablet, controls the functionality of the hardware, and supplies the bioimpedance measurement data to a database ("MySOZO") managed on an external cloud- hosted database. Patient weight is measured with load cells located in the SOZO Pro foot unit or can be hand entered.
Bioimpedance measurements require the patient's weight to be measured by a scale embedded in the base of the system or for the weight of the patient to be entered manually. Following the collection of the patient weight, the patient contacts the SOZO Pro with their bare hands and feet on stainless steel electrodes. The impedance measurement takes about 30 seconds, during which the SOZO Pro® system applies small levels of electrical energy (200µA RMS) to the body across 256 frequencies spaced from 3kHz to 1000kHz and measures the resulting voltage levels.
The provided text is a 510(k) summary for the ImpediMed SOZO Pro device. It describes the device, its indications for use, technological characteristics, and performance data to demonstrate substantial equivalence to a predicate device. However, it does not contain detailed information about specific acceptance criteria or a study proving the device meets those criteria, especially in the context of AI-driven performance or human reader improvement.
The "Performance Data" section primarily focuses on:
- Electrical safety/EMC: Tested according to IEC 60601 standards.
- Software V&V: Tested according to ISO 62304.
- Biocompatibility: Tested according to ISO 10993.
- Functional performance testing: "undertaken using fixed loads and comparing modified SOZO Pro to predicate SOZO measurements to demonstrate that outputs remained consistently accurate and precise." This is high-level and lacks specifics on metrics or thresholds.
- Weight Scale Verification: Performed according to NIST Handbook 44 and EU Directive 2014/31/EU.
Since the provided document does not include the specific details you've requested regarding acceptance criteria and performance study for, for example, an AI component's diagnostic accuracy, human-in-the-loop performance, or detailed ground truth establishment, I cannot populate all sections of your request.
Based on the available information, here is what can be inferred and what is explicitly missing:
1. Table of acceptance criteria and the reported device performance:
| Acceptance Criteria (Generic/Inferred) | Reported Device Performance (as stated or inferred from "no failures") |
|---|---|
| Electrical Safety/EMC | Met IEC 60601 standards; CB certificate granted. |
| Software Verification & Validation | Met acceptance criteria and performed as intended (per ISO 62304). |
| Biocompatibility | Passed ISO 10993 testing; no failures reported. |
| Functional Performance | Outputs remained consistently accurate and precise (compared to predicate, using fixed loads) - Specific metrics or thresholds not provided. |
| Weight Scale Accuracy | Experienced no failures (per NIST Handbook 44/EU Directive). |
2. Sample size used for the test set and the data provenance:
- Sample Size (Test Set): Not specified. The document mentions "fixed loads" for functional testing and a comparison to the predicate device, but no human subject test set size for evaluating body composition parameters is detailed.
- Data Provenance: Not specified. It's likely the testing occurred during device development/modification.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Not applicable/Not specified. The testing described focuses on engineering/functional validation (electrical, software, biocompatibility, scale accuracy, and comparison to predicate measurements using fixed loads), not on establishing diagnostic ground truth from human data for body composition parameters. The device provides "estimates" and tracks parameters over time, suggesting its performance is validated against its own internal consistency or engineering standards rather than external "expert" ground truth for individual body composition values in a clinical diagnostic sense based on this document.
4. Adjudication method for the test set:
- Not applicable/Not specified.
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 a Bioimpedance Spectroscopy (BIS) device for body composition analysis, not an AI-assisted diagnostic imaging device requiring human reader interpretation or MRMC studies. The document does not describe any AI component that would require human-in-the-loop evaluation for diagnostic improvement.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- The "Performance Data" section discusses "functional performance testing" using "fixed loads" and comparison to the predicate device, which would be a form of standalone performance evaluation for the device's measurement capabilities.
- The software V&V confirms the software meets acceptance criteria and performs as intended.
- However, specific standalone accuracy metrics (e.g., against a gold standard for body composition in human subjects) are not provided in this summary.
7. The type of ground truth used:
- For electrical, software, biocompatibility, and scale testing: Engineering standards, established regulatory requirements, and fixed measurement loads act as "ground truth."
- For "functional performance testing" of body composition outputs: The "ground truth" used was a comparison to the predicate SOZO measurements and "fixed loads" to ensure "consistently accurate and precise" outputs. There is no mention of a clinical gold standard (e.g., DEXA, whole-body MRI, dilution methods) used as ground truth for the body composition parameters in human subjects for this specific K-summary documentation.
8. The sample size for the training set:
- Not applicable/Not specified. This document describes a medical device, not an AI model that undergoes "training" in the machine learning sense with a distinct training dataset. The device measures bioimpedance and calculates parameters using established physiological models and algorithms.
9. How the ground truth for the training set was established:
- Not applicable. As above, this is not an AI model that undergoes training on a dataset with external ground truth. The device relies on biophysical principles and algorithms for its calculations.
§ 870.2770 Impedance plethysmograph.
(a)
Identification. An impedance plethysmograph is a device used to estimate peripheral blood flow by measuring electrical impedance changes in a region of the body such as the arms and legs.(b)
Classification. Class II (special controls). The device, when it is a body composition analyzer which is not intended to diagnose or treat any medical condition, is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 870.9.