OPHTAI-EVAL: OphtAI Diagnostic Performance Validation for Automated Screening of Eye Diseases

Sponsor

Evolucare Technologies (Industry)

Overall Status

Not yet recruiting

CT.gov ID

NCT05752045

Collaborator

Slb Pharma (Other), Assistance Publique - Hôpitaux de Paris (Other), BPIfrance (Other)

1,389

Enrollment

Locations

Arm

Anticipated Duration (Months)

173.6

Patients Per Site

14.4

Patients Per Site Per Month

Study Details

Study Description

Brief Summary

Evolucare OphtAI is a medical device offering automated, artificial intelligence powered, screening capabilities for Diabetic Retinopathy, Diabetic Macular Edema, Glaucoma, ARM and AMD, whose performances will by tested through the OphtAI-EVAL.

Condition or Disease	Intervention/Treatment	Phase
Diabetic Retinopathy Diabetic Macular Edema Age-Related Macular Degeneration Age-related Maculopathy Glaucoma	Device: Eye Fundus Double Capture for Eye Diseases Screening with OphtAI Medical Device	N/A

Detailed Description

OphtAI-Eval: is a prospective, multicentre, post-marketing clinical follow-up study (SCAC) of diagnostic validation (comparative vs gold standard).

It aims to: validate the diagnostic performance of the OphtAI software for the automated screening of diabetic retinopathy, diabetic macular edema, glaucoma, ARM and AMD.

Evolucare OphtAI is a medical imaging console for ophthalmology, interfaced with Evolucare Imaging.

It allows the detection, by statistical learning algorithms, of the following ocular pathologies using photographs of the retina:

Diabetic retinopathy (DR) (including gradation),
Diabetic macular edema (DME)
Age-related macular degeneration (AMD)
Age-related maculopathy (ARM, early form of AMD),
Glaucoma.

Evolucare OphtAI, is available on the French market since March 2019.

Study Design

Study Type:

Interventional

Anticipated Enrollment :

1389 participants

Allocation:

N/A

Intervention Model:

Single Group Assignment

Intervention Model Description:

Imaging and clinical data will be collected from patients to be reviewed by an expert reader group providing ground truth, and the medical device will be tested against this ground truthImaging and clinical data will be collected from patients to be reviewed by an expert reader group providing ground truth, and the medical device will be tested against this ground truth

Masking:

None (Open Label)

Masking Description:

Ground truth and Results being obtained in the end of recruitment and imaging collection, there are no effects on participant or care provider. Also, experts won't be aware of AI results and vice versa.

Primary Purpose:

Screening

Official Title:

Validation of OphtAI Software Diagnostic Performance for Automated Screening of Diabetic Retinopathy, Diabetic Macular Edema, Glaucoma, ARM and ARMD: a Multicentre Study

Anticipated Study Start Date :

Mar 1, 2023

Anticipated Primary Completion Date :

Oct 1, 2023

Anticipated Study Completion Date :

Mar 1, 2024

Arms and Interventions

Arm	Intervention/Treatment
Other: Diabetic patient group assessed for multiple eye diseases Each patient eye disease status will be assessed by expert readers so to provide ground truth against which algorithms performances will be assessed	Device: Eye Fundus Double Capture for Eye Diseases Screening with OphtAI Medical Device Double ophthalmological imaging capture of eye fundus with different fundus camera to screen patient for various eye diseases.

Arm

Intervention/Treatment

Other: Diabetic patient group assessed for multiple eye diseases

Each patient eye disease status will be assessed by expert readers so to provide ground truth against which algorithms performances will be assessed

Device: Eye Fundus Double Capture for Eye Diseases Screening with OphtAI Medical Device

Double ophthalmological imaging capture of eye fundus with different fundus camera to screen patient for various eye diseases.

Outcome Measures

Primary Outcome Measures

Referable Diabetic Retinopathy screening performance [1 year]
Referable Diabetic Retinopathy screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for RDR-EOphtha and RDR-OphDiaT_full algorithms

Secondary Outcome Measures

Diabetic retinopathy grading performance [1 year]
Diabetic retinopathy grading performance (OphtAI software vs Expert reader) for each grading sub algorithms (sensitivity/specificity) and globally (accuracy/agreement ratio)
Diabetic retinopathy grading + RDR-EOphtha algorithms combination performance for RDR detection [1 year]
Referable Diabetic Retinopathy screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for RDR-EOphtha and RD grading algorithms combination
Diabetic retinopathy grading + RDR-OphDiaT_full algorithms combination performance for RDR detection [1 year]
Referable Diabetic Retinopathy screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for RDR-OphDiaT_full and RD grading algorithms combination
Diabetic retinopathy grading + RDR-EOphtha +DME algorithms combination performance for RDR detection [1 year]
Referable Diabetic Retinopathy screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for RDR-EOphtha, RD grading and DME algorithms combination
Diabetic retinopathy grading + RDR-OphDiaT_full +DME algorithms combination performance for RDR detection [1 year]
Referable Diabetic Retinopathy screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for RDR-OphDiaT_full , RD grading and DME algorithms combination
Diabetic Macular Edema algorithm performance for DME detection [1 year]
Diabetic Macular Edema screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for DME algorithm
Age-Related Macular Degeneration algorithm performance for AMD detection [1 year]
AMD screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for AMD algorithm, for all forms and for either atrophic or neovascular AMD
Age-Related Maculopathy algorithm ("drusen") performance for ARM detection [1 year]
ARM screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for ARM algorithm, for all forms and for either macular or peripheric drusen, or for either hard/soft drusen
Glaucoma algorithms performance for Glaucoma suspicion [1 year]
Glaucoma screening sensitivity/specificity evaluation (OphtAI software vs Expert reader) for Glaucoma algorithms
Quality algorithm performance for quality assessment [1 year]
Bad quality images sensitivity/specificity evaluation (OphtAI software vs Expert reader) for quality assessment algorithms
Laterality determination algorithm performance for laterality assessment [1 year]
Laterality accuracy of OphtAI software vs Expert reader for laterality determination algorithm

Eligibility Criteria

Criteria

Ages Eligible for Study:

18 Years and Older

Sexes Eligible for Study:

All

Accepts Healthy Volunteers:

Doesn't not accept healthy volunteers as diabetic patients are needed, still they may be free of any eye diseases.

Inclusion Criteria:

The characteristics required for a subject to take part in the research are

Male or female over 18,
Type 1 or 2 diabetic,
Presenting for screening for diabetic retinopathy,
Beneficiary of a social security scheme,
For whom written consent has been obtained for participation in the protocol.

Exclusion Criteria:

The following characteristics do not allow the subject to take part in the research:

Patient with known DR, more severe than "minimal", including having been treated,
Any other condition that, in the opinion of the health professionals, may interfere with their ability to complete the study or may present a significant risk,
Presence of social, medical and/or psychological factors that may compromise the patient's adherence to the protocol,
Simultaneously participating in another clinical research protocol or having recently participated in another research study for which the exclusion period would not be completed.

Patients who participate in this research will not be able to participate in another research at the same time. However, there is no exclusion period at the end of this research for participation in any other study.

Contacts and Locations

Locations

	Site	City	State	Country	Postal Code
1	CHU Brest	Brest		France
2	Retinodiab Bourgogne	Dijon		France
3	Clinique Honoré Cave	Montauban		France
4	CHU Nantes	Nantes		France	44000
5	Retinodiab Franche-Comté	Roppe		France	90380
6	Diabète Occitanie	Toulouse		France	31059
7	CHU Toulouse	Toulouse		France
8	OPHDIAT	Paris	Île De France	France	75000