Explainable Ocular Fundus Diseases Report Generation System

Explainable Multimodal Deep Neural Networks for Identifying Ocular Fundus Diseases and Report Generation

To establish a deep learning system of various ocular fundus disease analytics based on the results of multimodal examination images. The system can analyze multimodal ocular fundus images, make diagnoses and generate corresponding reports.

Study Overview

• Status: Recruiting
• Conditions: Image, Body; Ophthalmological Disorder
• Intervention / Treatment: Diagnostic test: Various modalities of ocular fundus imaging

Detailed Description

The ocular fundus is the only part of the human body that can directly see the blood vessel microcirculation and nerve tissue. Through various imaging tests, including Color Fundus Photograph (CFP), Optical Coherence Tomography (OCT), Fluorescein Fundus Angiography (FFA) and Indocyanine Green Angiography (ICGA), etc., it is possible to statically overview or dynamically observe the retina and choroid, the condition of blood vessels and nerves, and comprehensive diagnosis of the disease. The screening, interpreting and accurate diagnosis of ocular fundus diseases are crucial for disease prevention, control and precise treatment. However, due to the variety of fundus examination methods, and the complexity and professionalism of the examination, there is a lack of fundus specialists who have sufficient clinical experience and knowledge to interpret fundus examinations. With the continuous development of artificial intelligence (AI) in diagnosing fundus diseases, various modalities of imaging examination methods are gradually applied to the development of fundus disease diagnosis systems. Moreover, medical images often come with corresponding reports, which are mostly generated by clinicians' or radiologists' experience.

Here, we are establishing a fundus disease diagnosis and report-generating system based on cross-modal ocular fundus imaging examinations, and fundus lesions were visualized at the same time. Multi-center data verification will also be conducted. The results of the research will assist in fundus lesions diagnosis and imaging reports generation. We hope this could popularize more complex fundus imaging examination methods to society, and help improve the early diagnosis and treatment of fundus lesions that cause blindness.

• Study Type: Observational
• Enrollment (Estimated): 15000

Contacts and Locations

• Study Contact: Name: Yingfeng Zheng, M.D. Ph.D; Phone Number: +8613922286455; Email: zhyfeng@mail.sysu.edu.cn
• Study Contact Backup: Name: Wenjia Cai, M.D. Ph.D; Phone Number: +8615017593912; Email: caiwenjia@gzzoc.com

Study Locations

• China
  • Guangdong
    • Guangzhou, Guangdong, China, 510000
      • Recruiting
      • Zhognshan Ophthalmic Center, Sun Yat-sen University
      • Contact: Yingfeng Zheng, M.D, Ph.D; Phone Number: +8613922286455; Email: zhyfeng@mail.sysu.edu.cn

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: Yes

• Sampling Method: Probability Sample

Study Population

Ocular fundus images were collected from different health care institutes all over China and from other countries.

Description

Inclusion Criteria:

• The quality of multimodal ocular fundus disease examination images and corresponding reports should be clinically acceptable.

Exclusion Criteria:

• Reports with key information missing.
• Images with severe image resolution reductions, blur or artifacts were excluded from further analysis.

Study Plan

How is the study designed?

Number of groups / cohorts

3

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Training set; Multimodal ocular fundus images and corresponding reports collected from multiple screening sites in China.
Internal Validation set; Records separated from the training set.
External Test set; Multimodal ocular fundus images and corresponding reports collected from multi-centers in China and around the world.	Diagnostic test: Various modalities of ocular fundus imaging; Through various modalities of ocular fundus imaging, combining with clinical data and the experience of clinicians to diagnose different fundus diseases.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Area under the receiver operating characteristic curve of the deep learning system	The investigators will calculate the area under the receiver operating characteristic curve of the deep learning system and compare this index with human ophthalmologists.	Baseline

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Intersection-Over-Union of the models' explanation accuracy	The investigators will calculate the Intersection-Over-Union (IOU) (or Jaccard similarity) between the lesion-image attention mapping regions and ground truth regions of the deep learning system.	Baseline
Sensitivity and Specificity of the deep learning system	The investigators will calculate the sensitivity and specificity of the deep learning system.	Baseline

Collaborators and Investigators

• Sponsor: Sun Yat-sen University
• Investigators: Principal Investigator: Yingfeng Zheng, M.D. Ph.D, Zhongshan Ophthalmic Center, Sun Yat-sen Univerisity,Guangzhou, Guangdong, China, 510060

Study record dates

Study Major Dates

• Study Start (Actual): January 1, 2011
• Primary Completion (Estimated): December 1, 2023
• Study Completion (Estimated): July 1, 2024

Study Registration Dates

• First Submitted: November 15, 2022
• First Submitted That Met QC Criteria: November 15, 2022
• First Posted (Actual): November 18, 2022

Study Record Updates

• Last Update Posted (Actual): July 11, 2023
• Last Update Submitted That Met QC Criteria: July 10, 2023
• Last Verified: July 1, 2023

More Information

Terms related to this study

• Keywords: Deep learning; Retinal disease; Choroidal disease
• Additional Relevant MeSH Terms: Eye Diseases
• Other Study ID Numbers: 2021KYPJ164

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No