(123 days)
K-D Balance is an objective balance assessment tool and is intended to be used for individuals for whom a balance measurement is a desired outcome. K-D Balance is intended for use to assess balance performance. Individual suitability for assessment must be judged on a case by case basis, by a qualified individual including those certified or licensed in their state to prescribe and use balance devices including but not limited to athletic trainers and coaches, physical therapists, nurses, physicians, and other health care providers. Patients with could affect balance include, but are not limited to, vestibular dysfunction, nausea, headaches, orthopedic injury, inner ear infection or dysfunction, neurological conditions, head injury, medication side effects, dehydration and fatigue. K-D Balance can be used wherever an iOS mobile operating device is available. K-D Balance may only be used by health care professionals.
K-D Balance is a mobile balance test to be used on the 6th generation iPhone or iPod touch with iOS versions 9.0 or later. K-D Balance measures balance using motion sensors already built-in to the iPhone or iPod touch devices. K-D Balance utilizes a proprietary algorithm that calculates a balance score based on the analysis of subject movements detected by the device's internal accelerometer. Balance is measured during a series of stances with the mobile device secured to the subject with a device holder. Health care professionals follow the step-by-step voice guidance that runs after launching the app during the entirety of the test to ensure consistency in obtaining balance measures. K-D Balance does not function as a diagnostic test, and K-D Balance results should be interpreted by healthcare professionals only.
Here's a breakdown of the acceptance criteria and the study details for the K-D Balance device, based on the provided FDA 510(k) summary:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly state pre-defined acceptance criteria with specific numerical thresholds for performance metrics. Instead, the primary acceptance criterion appears to be statistical agreement with the predicate device (Sway Balance™) for balance outcomes in both normal and balance deficit groups, across various stances.
| Acceptance Criteria (Inferred) | Reported Device Performance (K-D Balance) |
|---|---|
| Statistical agreement with predicate device (Sway Balance) for balance outcomes in normal subjects. | K-D Balance measures were in statistical agreement with Sway Balance outcomes for the normal group for the double leg, tandem right, and tandem left leg stances, according to linear regression analysis and Deming Regression Model. This supports substantial equivalence. |
| Statistical agreement with predicate device (Sway Balance) for balance outcomes in subjects with balance deficits. | K-D Balance measures were in statistical agreement with Sway Balance outcomes for the balance deficit group (including subjects with brain injury) for the double leg, tandem right, and tandem left leg stances, according to linear regression analysis and Deming Regression Model. Poorer balance scores were observed for subjects with balance deficits. This supports substantial equivalence. |
| Conformance to product specification (general). | K-D Balance met specification. |
| Sensitivity of the software to access data from the accelerometer. | Bench testing analyzed the sensitivity of the software to access data from the ST Microelectronics MEMS Accelerometer. Sensitivity scores with K-D Balance were comparable to the predicate. |
| Safety (no adverse effects or complications during or after testing). | No adverse effects or complications were encountered during or after performance testing. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: 79 subjects (25 males, 54 females; mean age: 38.63 ± 12.88, range: 18-65).
- 34 subjects had brain injury or other neurological conditions (balance deficits).
- 45 subjects were non-injured (normal group).
- Data Provenance: The study was conducted by the University of Texas, Southwestern Medical Center, Neurology and Sports Medicine Department. The data is prospective as testing was performed for the purpose of this device evaluation. The country of origin is the USA.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
This study does not appear to use human experts to establish "ground truth" in the traditional sense of diagnostic interpretation. Instead, the "ground truth" for balance performance is based on the objective measurements derived from the predicate device (Sway Balance™) and the clinical categorization of subjects (injured vs. non-injured).
The document mentions that the study was conducted by the University of Texas, Southwestern Medical Center, Neurology and Sports Medicine Department, implying medical professionals at that institution were involved in classifying subjects (e.g., identifying those with brain injury or neurological conditions). However, specific numbers or qualifications of individual experts for this categorization are not provided.
4. Adjudication Method for the Test Set
Not applicable. This was a comparative performance study against a predicate device and clinical categories, not an expert-driven adjudication of AI output. The "ground truth" for comparison was the predicate device's output and the subject's clinical status (injured/non-injured).
5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was done
No, an MRMC comparative effectiveness study was not explicitly described. This study focused on the performance of the K-D Balance device itself in comparison to a predicate device, not on how the K-D Balance assists human readers.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
Yes, a standalone performance study was done. The performance testing directly evaluated the K-D Balance application's ability to measure balance and its statistical agreement with the predicate device (Sway Balance™). The balance assessment is performed by the application's algorithm, and the output is a score. Human involvement is in administering the test and securing the device, but the balance measurement itself is algorithmic.
7. The Type of Ground Truth Used
The ground truth for this comparison study was:
- Objective measurements from a legally marketed predicate device: The balance scores generated by the Sway Balance™ application.
- Clinical categorization: Subjects were categorized as "normal" or "balance deficit" (which included those with brain injury or other neurological conditions). Poorer balance scores were expected and observed in the deficit group, confirming the clinical relevance of the measurements.
8. The Sample Size for the Training Set
The document does not provide information regarding a specific training set size or how the K-D Balance algorithm was developed (i.e., whether it used a training set in machine learning terms). The description focuses on its proprietary algorithm and its use of motion sensors.
9. How the Ground Truth for the Training Set Was Established
As no training set is described, the method for establishing ground truth for a training set is not available in the provided text. The evaluation is focused on the device's performance against a predicate and clinical categories rather than a machine learning model's training process.
N/A