Source: TG- Ophthalmo Topic Driver - PowerPoint Presentation

Slide1

Source:

TG-Ophthalmo Topic DriverTitle:Att.3 - Presentation (TG-Ophthalmo)Purpose:DiscussionContact:Arun ShroffE-mail: arun@xtend.aiAbstract:This PPT summarizes the content of the TDD for the TG on ophthalmology (L-017-A01), for presentation and discussion during meeting L.

FGAI4H-L-017-A03

E-meeting, 19-21 May 2021

Meeting L Topic Group UpdateOphthalmology (TG-Ophthalmo )E-meeting, 19 May – 21 May 2021Arun Shroff, Topic Driver, TG-Ophthalmology

Topic Group – OphthalmologyStandardized benchmarking of artificial intelligence for Ophthalmology.Conditions/Sub-topics in this group:

Diabetic Retinopathy (DR) Age Related Macular Degeneration (AMD)Glaucoma (GC) Pathological Myopia (PM)Red Eye (RE) – (Added Meeting G)

The Health ChallengeDiabetic Retinopathy (DR) At risk population: 422M worldwide with diabetes (2014)35%, 148M have DR 11%, 48M have Vision Threatening DR (64M by 2040)

Leading cause of blindness among adults worldwideAge Related Macular Degeneration (AMD) Damages macula and impairs central vision 196M by 2020Third leading cause of vision loss overall, leading cause for those over 50

The Health ChallengeGlaucoma (GC) Damages optic nerve & leads to vision loss 80M by 2020Pathological Myopia (PM)

Global Prevalence is 0.9% to 3.1%35% of people with myopia have High Myopia, which can develop into PMRed Eye [Added Meeting G] 2-3% visits to primary health centers & emergency facilities due to eye problems & majority are due to Red Eye. May denote more serious conditions like keratitis, iritis, glaucoma, which could lead to vision loss

Impact of AI Bridge acute shortage of healthcare professionals in LMICs, rural areas.

Provide earlier detection and prevent vision loss for millions. Decrease healthcare costs via earlier interventions Increase overall efficiency and scalability of current screening methods.

Benchmarking: DR ClassificationsBinary: [0 (Nogradable

Image)] (Optional) 1 (Non-referable Retinopathy = Normal or Mild) 2 (Referable Retinopathy = Moderate, Severe, PDR) Multi-class Classification: [0 (Nongradable Image) ] (Optional) 1(Normal) 2 (Mild) 3 (Moderate NPDR) 4 (Severe NPDR) 5 (PDR)

Benchmarking: AMD, GC, PM Classifications

AMD: [0 (Image Nongradable)]1 (No/early stage AMD 2 (Intermediate/advanced stage AMD) GC: [0 (Image Nongradable.]1 (No GC)2 (GC)Optic Disk SegmentationPM: [0 (Image Nongradable)]1 (No PM/HM)2 (HM: high myopia)3 (PM)

Benchmarking Metrics

Classification Tasks: Sensitivity:True Positive/(True Positive + False Negative)Specificity: True Negative/(True Negative + False Positive)AUC (Area Under ROC);Sensitivity Vs (1-Specificity) plotted at different points of the model

Accuracy, F1 Score

Cohen’s Kappa / Quadratic Kappa Score

Segmentation Tasks:

IOU , Dice Coefficient

Topic Group HistoryMeeting B - New York (Nov 2018)AI for Ophthalmology Use Case submitted in response to the Call for ProposalsMeeting C – Lausanne, (Jan 2019)Topic Group “Ophthalmology” established (2 Members)Meeting D- Shanghai, (Apr 2019)First version of Topic Description Document (TDD) – Version 1.0 completed

Meeting E to K: TDD Revised & Updated during each meetingNew members added

TG-

Ophthalmo MembersCurrently 22 members – From 15 organizations & 8 countriesDedicated mailing list : fgai4htgophthalmo@lists.itu.int​Mailing List has 35 members​​

TG-

Ophthalmo Output DocumentsTDD: Topic Group Description Document (FGAI4H-L-017-A02) Topic Group Call for Participation (FGAI4H-L-017-A01)Topic Group Collaboration Site: https://extranet.itu.int/sites/itu-t/focusgroups/ai4h/tg/SitePages/TG-Ophthalmo.aspx​​

Working on converting the TDD to the new TDD template (J-105)Outreach via Email & Social MediaCollaborating with WG-DAISAM (Data & AI Solution Assessment Methods)​ to setup benchmarking challenge for TG-Ophthalmo:Provide a public dataset that can be shared with participants to train their algorithmProvide a challenge description with metrics that will be used during the evaluation

Provide a private dataset for evaluationProgress Since Meeting K

Organizations contacted for datasets for benchmarking: EYEPACS, USA - Dr. Jorge Cuadros – discussion ongoingNHS Moorfields Eye Hospital, UK - Dr. Pearse A Keane, Consultant Ophthalmologist INSIGHT UK: Health Data Research Hub for Eye Health – Dr. Alastair Denniston, Director

Offers large, anonymized sets of patient data for researchRequire certain criteria to be met, application process and formal reviewEmail exchange and discussions are ongoing Progress Since Meeting K

Next StepsUpdates to TDD: Incorporate new TDD TemplateComplete sections on ethics, benchmarking, reporting.Split into subtopics Dataset Procurement: Complete the INSIGHT application for dataBenchmarking:Work with DISAM to setup benchmarking challenge.

 Outreach / Community BuildingIncrease engagement from membersGet more experts on board and involved.

Thank you!