(140 days)
The CureSight™ system is indicated for improvement in visual acuity in amblyopia patients, aged 4 - years, associated with anisometropia and/or with mild strabismus, having received treatment instructions (frequency and duration) as prescribed by a trained eye-care professional. CureSight™ is intended for both previously treated and untreated patients and is intended to be used as an adjunct to full-time refractive correction, such as glasses, which should also be worn under the anaglyph glasses during CureSight™ is intended for prescription use only, in an at-home environment.
The CureSight™ system is an eye-tracking-based system aimed for improving visual acuity and stereo acuity under dichoptic conditions. The technology is based on real-time eye tracking and separation of the visual stimuli presented on a monitor into two separate digital channels, one for each eye. Using this dichoptic method, any streamed video content can be tailored individually per eye and then presented simultaneously to each eye.
During the treatment, dichoptic anaglyph (red-blue) glasses that are part of the CureSight system are to be worn over the habitual spectacle correction.
The CureSight™ system comprises the following components/modules:
- CureSight-CS100 device (console and anaglyph glasses) ●
- CureSight Web-App/Portal
CureSight™ system is aimed for improving visual acuity and stereo acuity under dichoptic conditions, using digital content in pediatric patients (age 4 to
Acceptance Criteria and Device Performance for NovaSight CureSight-CS100
1. Table of Acceptance Criteria and Reported Device Performance
The primary effectiveness endpoint for the study was non-inferiority in the improvement of amblyopic eye distance visual acuity (AEDVA) compared to standard patching treatment.
| Acceptance Criterion | Reported Device Performance (CureSight System) | Met? |
|---|---|---|
| Improvement in amblyopic eye distance VA (AEDVA) from baseline to 16 weeks is not inferior to that of the control (patching) group within a margin of -0.10 logMAR. | Improvement in AEDVA: 2.63 lines (95% CI [2.24, 3.03] lines) | |
| Difference between groups (CureSight - Patching): 0.34 line (90% CI [-0.08, 0.76]) | ||
| Lower bound of the 90% CI for the difference (-0.08 logMAR) is greater than the pre-specified non-inferiority margin (-0.1 logMAR). | Yes |
Secondary Effectiveness Endpoints:
| Secondary Endpoint | Reported Device Performance (CureSight System) |
|---|---|
| Change from baseline in stereo acuity score (Randot preschool test) to week 16 in arcseconds (treatment group). | Median improvement: 0.40 log arcseconds (Range: -0.65 to 1.77, P 7 years on a study-approved device displaying single surrounded optotypes." This implies trained and qualified personnel skilled in these standardized visual acuity assessment methods, typically ophthalmologists, optometrists, or trained ophthalmic technicians, were responsible for these measurements. No specific years of experience or board certifications are provided for these evaluators. |
4. Adjudication Method for Test Set
The document does not describe an adjudication method involving multiple experts reviewing cases to establish ground truth for the visual acuity measurements. The visual acuity data was collected prospectively through direct measurement using standardized protocols (Lea symbol per ATS VA protocol and E-ETDRS VA protocol).
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not explicitly described for the device's performance. The study was a clinical trial comparing the CureSight device to a traditional treatment (patching) in human subjects ("evaluator-masked"), not a study comparing human readers with and without AI assistance in interpreting diagnostic images. Thus, no effect size of human readers improving with AI vs. without AI assistance is applicable or reported.
6. Standalone (Algorithm Only) Performance
Based on the provided text, a standalone (algorithm only without human-in-the-loop performance) study was not conducted or reported for the effectiveness of the device in improving visual acuity. The CureSight device is a treatment system that subjects use, and its performance is measured by the change in their visual acuity over time, not by an algorithm's diagnostic output. The software validation mentioned ("Software verification and validation testing") refers to the functional performance and safety of the software within the device, not its standalone diagnostic accuracy.
7. Type of Ground Truth Used for the Test Set
The ground truth for the clinical study (test set) was based on direct clinical measurements of visual acuity (Amblyopic Eye Distance Visual Acuity - AEDVA, Binocular Distance Visual Acuity - DVA) and stereo acuity using standardized, clinically accepted tests (Lea symbol per ATS VA protocol, E-ETDRS VA protocol, and Randot preschool test). This can be categorized as clinical outcome measures or objective clinical assessments.
8. Sample Size for the Training Set
The document does not provide information about a separate training set or its sample size. The description of "Performance Data" focuses exclusively on the pivotal clinical study (test set) and bench testing. The device is a digital therapy device, and while it uses "real-time software algorithm[s]," the text does not detail an AI/ML model that would typically have a distinct training set for diagnostic or predictive purposes related to the primary effectiveness endpoint.
9. How the Ground Truth for the Training Set Was Established
As no specific training set for an AI/ML model in the context of diagnostic or predictive performance is described, the method for establishing its ground truth is not provided. The "software algorithm" mentioned dynamically processes visual information based on eye tracking, but there's no indication that this algorithm was "trained" on a dataset with external ground truth labels in the conventional sense of AI validation studies.
N/A