Trial Outcomes & Findings for Detection and Classification of Diabetic Retinopathy From Posterior Pole Images With A Deep Learning Model (NCT NCT04805541)

NCT ID: NCT04805541

Last Updated: 2024-07-15

Results Overview

Sensitivity for detecting more than mild diabetic retinopathy

• Recruitment status: COMPLETED
• Target enrollment: 900 participants
• Primary outcome timeframe: 1 visit (1 day)
• Results posted on: 2024-07-15

Participant Flow

Participants were reffered to by the Endocrionlogy Department at Akdeniz University Hosptital for an ophthalmic examitation to the opthalmology department to screen for diabetic retinopathy, where they were invited to participate in the study.

Participants were evaluated for media opacity and other retinal diseases that might prevent good retinal photography.

Participant milestones

Measure	Screening Population Prospective Study Participants
Overall Study STARTED	900
Overall Study COMPLETED	865
Overall Study NOT COMPLETED	35

Reasons for withdrawal

Measure	Screening Population Prospective Study Participants
Overall Study Unable to be retinally photographed with good quality.	35

Baseline Characteristics

Race and Ethnicity were not collected from any participant.

Baseline characteristics by cohort

Measure	Screening Population n=900 Participants Prospective Study Participants
Age, Continuous	58.33 years STANDARD_DEVIATION 11.16 • n=900 Participants
Sex: Female, Male Sex · Female	442 Participants n=900 Participants
Sex: Female, Male Sex · Male	458 Participants n=900 Participants
Region of Enrollment Turkey	900 participants n=900 Participants

PRIMARY outcome

Timeframe: 1 visit (1 day)

Sensitivity for detecting more than mild diabetic retinopathy

Outcome measures

Measure	Prospective Study Participants (Optomed) n=865 Participants Participants Photographed Using Optomed Aurora	Prospective Study Participants (Canon) n=704 Participants Participants Photographed Using Canon CR2	Prospective Study Participants (Topcon) n=585 Participants Participants Photographed Using Topcon NW400
Sensitivity(mtmDR)	90.48 percentage of true positives Interval 83.95 to 94.98	95.65 percentage of true positives Interval 90.15 to 98.57	95.19 percentage of true positives Interval 89.13 to 98.42

PRIMARY outcome

Timeframe: 1 visit (1 day)

Sensitivity for detecting vision threatening diabetic retinopathy

Outcome measures

Measure	Prospective Study Participants (Optomed) n=865 Participants Participants Photographed Using Optomed Aurora	Prospective Study Participants (Canon) n=704 Participants Participants Photographed Using Canon CR2	Prospective Study Participants (Topcon) n=585 Participants Participants Photographed Using Topcon NW400
Sensitivity(vtDR)	95.12 percentage of true positives Interval 87.98 to 98.65	96.00 percentage of true positives Interval 88.75 to 99.17	96.34 percentage of true positives Interval 94.56 to 97.67

PRIMARY outcome

Timeframe: 1 visit (1 day)

Specificity for detecting more than mild diabetic retinopathy

Outcome measures

Measure	Prospective Study Participants (Optomed) n=865 Participants Participants Photographed Using Optomed Aurora	Prospective Study Participants (Canon) n=704 Participants Participants Photographed Using Canon CR2	Prospective Study Participants (Topcon) n=585 Participants Participants Photographed Using Topcon NW400
Specificity(mtmDR)	97.21 percentage of true negatives Interval 95.73 to 98.29	95.92 percentage of true negatives Interval 93.99 to 97.37	95.93 percentage of true negatives Interval 93.85 to 97.47

PRIMARY outcome

Timeframe: 1 visit (1 day)

Specificity for detecting vision threatening diabetic retinopathy

Outcome measures

Measure	Prospective Study Participants (Optomed) n=865 Participants Participants Photographed Using Optomed Aurora	Prospective Study Participants (Canon) n=704 Participants Participants Photographed Using Canon CR2	Prospective Study Participants (Topcon) n=585 Participants Participants Photographed Using Topcon NW400
Specificity(vtDR)	98.82 percentage of true negatives Interval 97.77 to 99.45	96.34 percentage of true negatives Interval 94.56 to 97.67	95.93 percentage of true negatives Interval 93.86 to 97.47

SECONDARY outcome

Timeframe: 1 Day

percentage of patients with image quality sufficient for EyeCheckup to produce an output to patients with image quality sufficient for the clinical reference standard to produce an output

Outcome measures

Measure	Prospective Study Participants (Optomed) n=865 Participants Participants Photographed Using Optomed Aurora	Prospective Study Participants (Canon) n=704 Participants Participants Photographed Using Canon CR2	Prospective Study Participants (Topcon) n=585 Participants Participants Photographed Using Topcon NW400
Diagnosability	96.57 percentage of participants Interval 96.45 to 98.58	100 percentage of participants Interval 99.47 to 100.0	100 percentage of participants Interval 99.37 to 100.0

SECONDARY outcome

Timeframe: 1 Day

Percentage of participants for whom Non-Dilated Images were sufficient to use EyeCheckup

Outcome measures

Measure	Prospective Study Participants (Optomed) n=865 Participants Participants Photographed Using Optomed Aurora	Prospective Study Participants (Canon) n=704 Participants Participants Photographed Using Canon CR2	Prospective Study Participants (Topcon) n=585 Participants Participants Photographed Using Topcon NW400
Non-Dilated EyeCheckup Usage	62.24 percentage of patients imaged undilated Interval 58.88 to 65.52	68.90 percentage of patients imaged undilated Interval 65.32 to 72.29	78.46 percentage of patients imaged undilated Interval 74.9 to 81.72

Adverse Events

Prospective Study Participants

Serious events: 0 serious events

Other events: 0 other events

Deaths: 0 deaths

Serious adverse events

Adverse event data not reported

Other adverse events

Adverse event data not reported

Additional Information

Prof.Dr. Ahmet Burak Bilgin

Akdeniz University

Phone: 5465473336

Email: gozdrburakbilgin@gmail.com

Results disclosure agreements

• Principal investigator is a sponsor employee
• Publication restrictions are in place