Search Results
Found 2 results
510(k) Data Aggregation
(106 days)
The EksoNR™ is intended to perform ambulatory functions in rehabilitations under the supervision of a trained physical therapist for the following populations:
- · Individuals with acquired brain injury, including traumatic brain injury and stroke (upper extremity motor function of at least 4/5 in at least one arm).
- · Individuals with spinal cord injuries at levels T4 to L5 (upper extremity motor function of at least 4/5 in both arms).
- · Individuals with spinal cord injuries at levels of C7 to T3 (ASIA D with upper extremity motor function of at least 4/5 in both arms).
The therapist must complete a training program prior to use of the devices are not intended for sports or star climbing.
The Ekso is a powered motorized orthosis. It consists of a fitted metal brace that supports the legs, feet, and torso. It is worn via straps on the body, legs, and feet. Battery powered motors drive knee and hip joints. It has an integrated solid torso containing the computer and power supply. It has a hand-held user interface to specify settings and initiate steps. The Ekso is used with a cane, crutch, or walker.
The provided text is a 510(k) Summary for the EksoNR™ exoskeleton, a medical device for rehabilitation. It details the device's indications for use, its technical characteristics, and non-clinical and clinical performance data to support its substantial equivalence to a predicate device.
However, the information provided does not contain details about an AI/algorithm-driven device, nor does it specify acceptance criteria in terms of quantitative performance metrics (like sensitivity, specificity, or AUC) that would be typically found for such a device. The document focuses on showing the device's safety and effectiveness for an expanded patient population in rehabilitation based on clinical studies, rather than the performance of an algorithm.
Therefore, I cannot extract the information required to populate the requested table or answer questions related to AI/algorithm performance, ground truth establishment for a test set, expert adjudication, or MRMC studies.
The document indicates "Software - Verification, Validation, and hazard analysis" was performed, but it does not describe a study involving a test set, ground truth experts, or performance metrics relevant to an AI/algorithm's diagnostic or predictive capabilities. It primarily discusses the exoskeleton's performance in facilitating gait ambulation and safety for patients, not an AI's performance in, for example, image analysis or diagnostic prediction.
To answer your request thoroughly, I would need a document describing an AI/algorithm-driven medical device and its performance evaluation against defined acceptance criteria.
Ask a specific question about this device
(154 days)
The Phoenix orthotically fits to the lower limbs and trunk. The device is intended to enable individuals with spinal cord injury at levels T4 to L5 to perform ambulatory functions in rehabilitation institutions in accordance with the user assessment and training certification program. This device is not intended for sports or stair climbing.
The Phoenix™ is a wearable, powered exoskeleton that assists a trained user to sit, stand, walk, and turn. The Phoenix consists of a pair of motorized leg braces coupled to a torso module, a lithium-ion battery pack, a main controller unit, and a wireless user interface attached to the handle of an assistive device (such as a crutch, walker, or parallel bars), control software, and mobile Android tablet hosting a mobile app. The Phoenix dimensions, such as spine length, torso hip width, femur length, femur bracket, tibia length, tibia bracket, and foot plate length can be adjusted individually. The Phoenix is coupled to the user via soft-good components (i.e. shoulder straps, waist pads, thigh straps, and shin pads), which can be adjusted to accommodate various users' dimensions.
The provided document describes the Phoenix™ powered exoskeleton and its regulatory submission (K183152). It outlines performance data to demonstrate its safety and effectiveness and its substantial equivalence to a predicate device.
Here's a breakdown of the requested information based on the document:
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are not explicitly stated in a quantitative manner as pass/fail thresholds for specific metrics before the study. Instead, the study objectives and the results achieved are presented as evidence that the device meets its intended purpose. The table below synthesizes the implicit acceptance criteria from the study objectives and the reported performance.
| Acceptance Criterion (Implied from Study Objective) | Reported Device Performance (Result) |
|---|---|
| Safe and effective for intended use (SCI T4-L5 non-ambulatory to poorly ambulatory to stand up and walk under various conditions) | Safety: "Adverse Events (AE) reported during the study included minor instances of bruising. The causes attributed to these reported incidents were related to improper fitting or improper padding... There were no Unanticipated Adverse Events (UAE)." "The clinical study concluded that the Phoenix device is safe and effective for its intended use..." |
| Effectiveness: All detailed performance metrics below support effectiveness. | |
| Participants with SCI T4-L5 can safely complete transitional movements (stand up, turn, sit down) and walk using Phoenix with minimal contact assistance or Functional Independence Measure (FIM) | Transitional Movements (TUG Test): 39 out of 40 subjects completed with minimal contact assistance (FIM score of 4 or higher); 1 subject completed with moderate contact assistance (FIM score of 3). |
| Level of Assistance (WISC-II): Averaged mean scores of 8.60 (±2.19) for the final assessment. | |
| Level of Assistance (FIM): "FIM scores as noted previously support that subjects were capable of managing all scenarios presented..." | |
| Participants with SCI T4-L5 are able to achieve walking during the 10 Meter Walk Test (10MWT) and 6 Minute Walk Test (6MWT). | 10MWT: All participants (40/40) were able to complete the 10MWT. Mean FIM was 4.6 (±0.50). Average completion time was 61.9 seconds (±34.64), with a mean speed of 0.12 m/s (±0.06). |
| 6MWT: Mean FIM was 4.37 (±0.49), "indicating an acceptable level of functional independence." | |
| User exertion for basic level-ground walking (Modified Borg Rating of Perceived Exertion) | Averaged results of 3.3 for indoor level-ground walking at the end of sessions, corresponding to an exertion level just above "moderate." |
2. Sample size used for the test set and the data provenance
- Sample Size: 40 subjects.
- Data Provenance: The document does not specify the country of origin of the data. It states, "The study was performed in compliance with Good Clinical Practices (GCP) with subjects enrolled in an IRB approved study that were consented for participation according to the intended use of the device, defined inclusion criteria, and defined exclusion criteria..." This implies a prospective clinical study.
3. Number of experts used to establish the ground truth for the test set and the 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 diagnostic accuracy studies (e.g., radiologists interpreting images). Instead, the study evaluates the functional performance and safety of the exoskeleton in human subjects. The outcome measures (FIM, WISC-II, 10MWT, 6MWT, TUG Test, Modified Borg) are objective measures or standardized assessments commonly used in rehabilitation, performed by trained clinicians/investigators within the study protocol. The "ground truth" here is the direct, observed performance of the subjects using the device.
4. Adjudication method (e.g., 2+1, 3+1, none) for the test set
The concept of an adjudication method (like 2+1 for conflicting interpretations) is not applicable here as this is a functional performance and safety study, not a diagnostic study requiring interpretation of outcomes by multiple experts. The clinical outcomes were directly measured or observed by the study staff.
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, a multi-reader, multi-case (MRMC) comparative effectiveness study was not done. This study is a clinical trial assessing the performance of a medical device (exoskeleton), not an AI-powered diagnostic tool. Therefore, the concept of human readers improving with or without AI assistance is irrelevant to this document.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done
No, a standalone algorithm-only performance study was not done. The Phoenix™ is a physical medical device (exoskeleton) intended for use by individuals with spinal cord injury in rehabilitation settings. Its entire purpose involves human-in-the-loop performance (the user wearing and operating the device under clinician supervision).
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
For this functional performance study, the "ground truth" consists of observed clinical outcomes data and standardized functional assessment scores collected directly from the subjects using the Phoenix™ device in a controlled clinical setting. These include:
- Functional Independence Measure (FIM) scores
- WISC-II scores (likely a typo for WISCI II - Walking Index for Spinal Cord Injury)
- 10 Meter Walk Test (10MWT) results (time, speed)
- 6 Minute Walk Test (6MWT) results (implied distance/FIM score)
- Timed Up-and-Go (TUG) Test results
- Modified Borg Rating of Perceived Exertion scores
- Adverse Event (AE) reporting.
8. The sample size for the training set
The document mentions "training with the device" for the 40 subjects enrolled in the clinical study as part of their 20 sessions. This training is for the actual users (patients) and clinicians to learn how to operate the device, not an algorithm's training data.
The document does not refer to a "training set" in the context of machine learning or AI algorithms. The Phoenix™ is a powered exoskeleton, not an AI diagnostic tool that requires a separate training data set for model development.
9. How the ground truth for the training set was established
As there is no mention of a "training set" for an algorithm in this context, this question is not applicable to the provided document. The device itself is "trained" in terms of its parameters and control mechanisms during its engineering and development phase, but this is distinct from "ground truth for a training set" in AI/ML performance studies.
Ask a specific question about this device
Page 1 of 1