Search Results

The Neurolutions IpsiHand Upper Extremity Rehabilitation System is indicated for use in chronic stroke patients (≥ 6 months post-stroke) age 18 or older undergoing stroke rehabilitation, to facilitate muscle re-education and for maintaining or increasing range of motion in the upper extremity.

The Neurolutions IpsiHand Upper Extremity Rehabilitation System (a.k.a. Neurolutions System or IpsiHand System) detects goal-oriented brain activity using non-invasive EEG electrodes to allow a stroke patient to perform therapeutic exercises that they would otherwise not be able to perform, due to their impairment. The Neurolutions System consists of the Neurolutions Handpiece, a Tablet computer, and a Biometric Headset. Using a standard Windows Tablet as the patient interface, the System translates brain signals from the Biometric Headset into movement of the motor-driven Handpiece worn over the patient's hand and wrist. The motion of the Handpiece, in turn, opens and closes the patient's impaired hand. The combined action of these System components allows the stroke patient to perform physical therapy exercises that they would otherwise not be able to perform, due to their impairment. The Biometric Headset, which contains dry electroencephalographic (EEG) sensors, receives electrical signals from the motor or pre-motor cortex, predominantly of the unaffected hemisphere of the patient's brain, and in doing so, detects the patient's intentions to move their affected hand. These intentions to move are translated into motor movements of the Handpiece using software that drives a linear actuator contained inside of the Handpiece. The Handpiece opens and closes the hand using a 3-finger pincer grip mechanism designed to emulate the movement of grasping an object using the two forefingers and thumb in a grasping motion (one degree of freedom). The device functions as powered exercise equipment for the patient's hand using three separate modes: the main mode of therapy is referred to as the brain-computer interface (BCI) or 'thought' mode in which the patient's hand is opened or closed by the powered orthosis based on brain signals received from the Biometric headset; the second mode is referred to as a 'volitional' mode in which a patient actively opens and closes their hand with the System enabling independent range of motion; the third is a continuous passive motion (CPM) mode in which the System simply moves the patient's hand passively through a comfortable range of motion in a repetitive fashion. The Neurolutions System is designed for use in clinic or home settings as part of prescribed rehabilitation therapy.

Acceptance Criteria and Device Performance for Neurolutions IpsiHand Upper Extremity Rehabilitation System

The Neurolutions IpsiHand Upper Extremity Rehabilitation System is an EEG-driven upper extremity powered exerciser indicated for use in chronic stroke rehabilitation. The device's safety and effectiveness were evaluated through a series of clinical investigations and non-clinical studies.

1. Table of Acceptance Criteria and Reported Device Performance

The regulatory information outlines several special controls which act as acceptance criteria. The summaries of non-clinical and clinical studies demonstrate how these criteria were met.

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

The primary clinical evidence comes from three studies (QRS-0008, ORS-0012, ORS-0013) that served as the "test set" for the device's clinical performance.

• QRS-0008: 10 chronic hemiparetic stroke survivors.
• ORS-0012: 24 subjects progressed to Phase 2 (the therapeutic phase), with 17 completing the study protocol and 12 completing the 6-month durability visit.
• ORS-0013 (Preliminary): 13 chronic hemiparetic stroke survivors, with 12 completing 12 weeks of therapy.

Data Provenance: All three studies were:

• Prospective and Non-Randomized.
• Self-controlled: Patients served as their own controls, with baseline measurements compared to post-treatment measurements.
• Country of Origin: Not explicitly stated for all sites, but ORS-0013 was conducted at Washington University IRB in St. Louis, MO, indicating a US origin. The others are also implied to be within the US given the FDA submission context.

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

The ground truth in these studies was based on established clinical assessments of motor function and impairment. These assessments (e.g., ARAT, UEFM, Motricity Index, grip strength) are performed by trained clinical specialists. The document does not specify the exact number of experts or their specific qualifications (e.g., years of experience as a neurologist or physical therapist), beyond referring to "clinical specialist" in ORS-0013. However, the use of recognized, standardized clinical outcome measures implies that these assessments were performed by qualified personnel.

4. Adjudication Method for the Test Set

The studies used retrospective self-comparison (baseline vs. post-treatment for each patient) as the primary method to assess improvement. There is no mention of an independent expert adjudication panel for individual patient outcomes from the clinical assessment tools. The interpretation of the aggregated data and the overall clinical meaningfulness was part of the sponsor's analysis and FDA's review.

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

No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not explicitly mentioned or performed in the context of human readers improving with AI vs. without AI assistance. The Neurolutions System is a direct brain-computer interface therapy device, not an AI-assisted diagnostic tool that aids human interpretation of medical images or data. The "AI" here refers to the BCI algorithm that translates brain signals into device movement.

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

Yes, in a way. The core function of the device is the "brain-computer interface (BCI) or 'thought' mode" where "the patient's hand is opened or closed by the powered orthosis based on brain signals received from the Biometric headset." This represents the device (algorithm and hardware) performing its intended function based solely on the patient's brain signals, without direct human intervention in each movement. The clinical studies (especially Phase 1 of QRS-0008, ORS-0012, and ORS-0013) inherently demonstrated the standalone performance of the BCI in detecting intentions and driving the handpiece. The "human-in-the-loop" aspect is the patient's brain activity providing the input.

7. The Type of Ground Truth Used

The ground truth used was primarily clinical assessment scores from recognized, validated scales performed by clinical specialists. These include:

• Action Research Arm Test (ARAT)
• Fugl-Meyer Upper Extremity (UEFM)
• Canadian Occupational Performance Measure (COPM)
• Motricity Index
• Grip Strength
• Active Range of Motion (AROM)
• Modified Ashworth Scale

These are considered objective measures of functional motor outcomes and impairment.

8. The Sample Size for the Training Set

The document does not explicitly delineate a "training set" in the conventional machine learning sense for the BCI algorithm. The development of the BCI algorithm itself would have involved data collection for signal processing and pattern recognition. However, the regulatory submission focuses on the clinical studies (QRS-0008, ORS-0012, ORS-0013) that served as the validation and effectiveness studies using human participants. If "training set" refers to the initial development of the BCI's core functionality, that information is not provided here. The three clinical studies mentioned served to demonstrate the device's efficacy in real-world use rather than to train the underlying BCI algorithm.

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

As noted above, specific details about a "training set" for the BCI algorithm in the context of this submission are not provided. However, for such BCI systems, the ground truth for training would typically involve:

• Intentional Movement Observation: Patients would perform specific movements (e.g., attempt to open/close hand) while EEG signals are simultaneously recorded. The intent to move or the actual movement (if physically possible) would serve as the "ground truth label" for the corresponding brain signals.
• Motor Imagery Tasks: Patients would imagine performing movements, and associated brain patterns would be recorded and labeled.
• Expert Observational Data: Clinical specialists observing and documenting patient movement or attempted movement in conjunction with EEG data.

Without specific details, it is inferred that the BCI algorithm was developed using standard BCI training methodologies where intentional or imagined movements are correlated with EEG patterns.

Ask a specific question about this device