Description:
 
Background
Various medical diseases are difficult to diagnose correctly or consistently. 
Variations in patient populations can cause inaccuracies in the patient test results.
Variations in clinical technique can cause one, highly-trained practitioner to diagnose a patient with a disease and another, highly-trained practitioner to diagnose the same patient as normal.
Retinopathy of prematurity (ROP) is a proliferative retinal vascular disease that affects approximately two-thirds of premature infants weighing less than 1250 grams at birth. 
Most cases of ROP are mild and can be resolved. However, 5-10% of cases progress to severe ROP can lead to retinal detachment and permanent blindness if untreated. 
The most critical feature of severe, treatment-requiring ROP is the presence of "plus disease."
 
Technology Overview
This technology proposes an "intelligent teleophthalmology" platform for retinopathy of prematurity that is designed to be a part of routine clinical practice in the NICU as well as rural screening programs. The core of this software is an algorithm (DeepROP) which can automatically diagnose plus disease from fundus photographs. The algorithm was trained on a large multi-institutional dataset collected as part of the ongoing i-ROP cohort study, consisting of more than 5,500 images from 718 patients.
This algorithm was evaluated on non-overlapping splits of the training data to assess performance. 
 
Benefits
- The system allows clinicians to objectively assess the severity of disease from fundus photographs, eliminating subjectivity and bias
- Provides a more granular "risk score" that is sensitive to change across consecutive exam
- Provides an enhanced view of the retinal vasculature which can be viewed alongside the raw image, and serves as a quality control measure
 
Applications
- To diagnose a 'plus disease' in retinopathy
- Can be incorporated into any camera systems
- Medical industries
 
Opportunity
- License
- Partnering
- Research Collaboration
 
Patent Information:
Category(s):
Diagnostics
For Information, Contact:
Mark Saulich
Associate Director of Commercialization
Northeastern University
m.saulich@northeastern.edu
Inventors:
Deniz Erdogmus
Efstratios (Stratis) Ioannidis
Jayashree Kalpathy-Cramer
James Brown
Michael Chiang
Paul Chan
Keywords:
COVID