Search Results

Computer and video system used to quantify and graphically display human movement patterns and techniques for uses such as assessment and training of limb or body motion in gait analysis, prosthetic design, pre/post rehabilitation evaluation, physical therapy, and the like.

The system uses off-the-shelf video cameras, sensors, and computers to collect, quantify, and document human movement in two-dimensional or three-dimensional space. The proprietary elements of the system are Peak's software applications, which run on Microsoft Windows, and an interface box, which can mix and synchronize multiple digital and analog inputs when active data capture is used. The system can effect the capture of real-time video images of motion along with associated sensor information, then sequence the coordinate and sensor information for recording on the computer. Alternatively, the system can use video recordings to provide the motion information. Motion and sensor data are recorded, analyzed, and displayed or reported from the computer.

A typical study involves recording the locations of small reflective markers on subjects as movement occurs. This provides instant coordinate data. Coordinate data may also be gathered from digitized video sequences. Analog sensors, such as force platforms or EMG systems, may also be part of the instrumentation during movement. Their information can be synchronized, processed, and displayed along with the coordinate information. The system can calculate standard 2-D or 3-D kinematic and kinetic parameters using specific calculation modules. A report generation package displays and reports the data: in raw or processed form. step-by-step, or sequentially.

The Peak Motus system may include software modules, such as

• • Advanced video or 3D-optical-motion-capture module
• • 2D and/or 3D kinematic calculation and display software
• • Analog acquisition module
• • 3D pan and tilt module
• • KineCalc mathematical analysis module
• • Peak Motus Gait Analysis Template

Its hardware comprises a Microsoft Windows computer workstation with proprietary synchronizing and processing boards, and a variety of equipment setups, including

• • Purchased cameras, VCRs, lights, calibration frames, high-speed video
• • Video, optical-capture, and analog-acquisition boards
• • Event synchronization unit
• Analog EMG and force-platform systems and analog-to-digital interface unit (accepts up to 64 sensors)

Movement tracking can be done with or without markers in virtually any collection environment—indoors, outdoors, or even under water. For most medical applications, the tracking is done indoors in a controlled laboratory environment with markers.

The Peak Motus System has standard data reduction algorithms for editing, filtering and calculating linear and anqular displacements, velocities, and accelerations. Specialized templates are available for custom calculations: for example, the Peak Motus 3D Gait Analysis Template provides industry-standard gait-analysis calculations such as body segment angles. joint forces, moments, and powers.

The systems' report and display package can print standard reports or can display data on the computer screen through a multimedia-display tool. Parametric data created with the system may be displayed as line graphs, x-y plots, bar graphs, stick figures (with vector overlays), text, and data tables.

Acceptance Criteria and Study for Peak Motus Motion Measurement System (K030714)

Based on the provided K030714 submission, the Peak Motus Motion Measurement System is not a traditional artificial intelligence (AI) device and therefore, many of the typical acceptance criteria and study components associated with AI/ML-based medical devices are not applicable or explicitly detailed in the document. This submission pre-dates widespread AI integration in medical devices and focuses on demonstrating substantial equivalence to a predicate device.

1. Table of Acceptance Criteria and Reported Device Performance

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

The document does not specify a distinct "test set" in the context of an AI/ML algorithm evaluation. The effectiveness of the system was demonstrated through:

• "Extensive in-house testing." No sample size, number of cases, or data provenance (country, retrospective/prospective) for this in-house testing is provided.
• "In-house and clinical testing by independent third parties." Again, no specifics regarding sample size, cases, or data provenance are detailed for this testing.

Given the device's nature (motion measurement system, not AI/diagnostic), "testing" likely refers to validation of measurement accuracy, repeatability, and functionality against known standards or expected human movements, rather than a diagnostic performance evaluation on a labeled dataset.

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

This information is not provided as the device isn't being evaluated against a diagnostic ground truth in the way an AI diagnostic algorithm would be. The "ground truth" for a motion measurement system would typically be the actual motion, measured by highly accurate reference systems or predefined physical movements, or an assessment of the output's concordance with physiological understanding. The document does not specify expert involvement for establishing such ground truth.

4. Adjudication Method for the Test Set

This information is not applicable/provided. Adjudication methods (e.g., 2+1, 3+1) are typically used in studies involving expert readers for diagnostic tasks where there might be inter-reader variability. For a motion measurement system, the "correctness" of a measurement would be against a physical standard or expert interpretation of the measured parameters, not a consensus on a diagnostic label.

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

No, an MRMC comparative effectiveness study was not explicitly described or indicated. This type of study is relevant for evaluating the impact of an AI algorithm on human reader performance (e.g., radiologists interpreting images). The Peak Motus system is a measurement tool, not an AI diagnostic aid that assists human readers in interpreting data they are already analyzing.

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

The device itself is a "standalone" system in that the software processes inputs (video, sensor data) and outputs measurements and displays. The effectiveness claims ("Extensive in-house testing has demonstrated the effectiveness of this system.") would imply a form of standalone performance evaluation, but the specifics (what metrics, against what reference) are not detailed. It's not "standalone" in the AI sense of performing a diagnostic task without human intervention, but rather a tool that generates data for human interpretation.

7. The Type of Ground Truth Used

The document does not explicitly state the "type of ground truth" used for quantitative performance evaluation. For a motion measurement system, potential ground truths could include:

• Reference measurements: Comparison against highly accurate, calibrated measurement systems (e.g., optical motion capture systems known for their precision).
• Physical standards: Measuring known distances or movements of mechanical jigs.
• Physiological plausibility/expert consensus on normal movement patterns: While not explicitly stated as "ground truth" for numerical validation, the "clinical testing by independent third parties" likely involved experts assessing the output for its clinical utility and physiological accuracy.

The summary emphasizes the system's ability to "collect, quantify, and document human movement" and "calculate standard 2-D or 3-D kinematic and kinetic parameters." The "effectiveness" therefore implies that these calculations and measurements are accurate and reliable for their stated purpose.

8. The Sample Size for the Training Set

This information is not applicable and not provided. The Peak Motus Motion Measurement System is described as a "mature product" leveraging proprietary software applications and standard data reduction algorithms. It is not presented as an AI/ML system that undergoes a "training phase" on a dataset in the modern sense. Its algorithms are likely deterministic and based on biomechanical principles and established physics.

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

This information is not applicable and not provided for the same reasons as point 8. The algorithms are not "trained" on a labeled dataset to learn patterns; they are designed and programmed based on scientific principles to perform specific calculations and measurements.

Ask a specific question about this device