Retinal Deep PhenotypingTM by A Novel Mydriatic Hyperspectral Retinal Camera (MHRC) And Analysis by Deep Learning
Overview
- Phase
- Not Applicable
- Intervention
- Not specified
- Conditions
- Healthy
- Sponsor
- Optina Diagnostics Inc.
- Locations
- 1
- Primary Endpoint
- At least one (1) Deep Learning (DL) model that characterizes the retina.
- Status
- Withdrawn
- Last Updated
- 2 years ago
Overview
Brief Summary
This study aims to collect a baseline dataset of MHRC retinal scans that will be used for the pre-training of deep learning models from the hyperspectral retinal image phenotypic features that may form the basis for multiple future classification applications.
As an exploratory study, there are no endpoints per se, however, the following sub-objective will be evaluated for determining the success of this study:
- Collection and characterization of MHRC retinal images from at least 250 participants that score at least 80 on the real-time Quality Index (included in the MHRC software).
- Development of at least one (1) DL model of the retina. Models may be used for the development of novel classifier tests and potential use in a clinical setting.
Detailed Description
This is an exploratory, observational, cross-sectional, multi-site study designed to collect a baseline dataset of MHRC retinal image scans for use in Deep Learning models. Subjects will be recruited from eye clinics where patients will undergo mydriasis (pupil dilation) as part of their clinical visit. Eligible participants will be provided information about the study and delegated site personnel will assist with the Informed Consent process. If a participant provides their Informed Consent, they will be enrolled in the study and undergo a single retinal imaging session with the Optina MHRC device, on one or both eyes.
Investigators
Eligibility Criteria
Inclusion Criteria
- •Adults 18 years and older who will undergo mydriasis (pupil dilation) as part of their eye clinic visit.
Exclusion Criteria
- •Contraindications for mydriatic fundus imaging.
- •Pupil dilation contraindicated (due to a pathology, or presence of 3 quadrants with Van Herick grading of 0 or 1 without iridotomy).
- •Inadequate pupil dilatation (\< 6mm diameter) preventing uniform illumination of the retina with the MHRC.
- •Refractive error outside the range of -15 D to +15 D.
- •Deficient visual fixation (inability to fixate for at least 2 s)
- •Inability of obtaining at least 3 images of satisfactory quality with the MHRC per the Optina Diagnostics quality index software and /or per the eye specialists' evaluation.
Outcomes
Primary Outcomes
At least one (1) Deep Learning (DL) model that characterizes the retina.
Time Frame: 12 months
Review of any safety events (AE, SAE, UADEs) that occur throughout the study.
Time Frame: 12 months
As an exploratory study, there are no acceptance criteria for these study outcomes, rather the study results will be summarized in a Final Clinical Study report that will document the study outcomes and any learnings/models that may be applied to new development project using the MHRC H5 images