AI improves cataract surgery in the Global South

Although appropriate surgical treatment of cataracts is guaranteed in high-income countries, the results of surgical treatment in the Global South are often unsatisfactory. Video recording allows you to analyze possible surgical errors, improve training and clearly optimize surgical results. However, until now it has taken a very long time. Promising approaches currently exist to automate surgical video analysis using artificial intelligence (AI).

However, deep learning algorithms for video analysis of surgical procedures commonly used in countries of the Global South have not yet been developed.

Researchers from Bonn University Hospital, University of Bonn, Sankara Eye Foundation India and Microsoft Research India want to change this and develop a corresponding algorithm. The aim is to improve cataract surgery outcomes in the Global South in the long term. As an important first step, they compiled an overview of previous AI approaches to analyzing cataract surgery. The results were announced below. translational vision science and technology.

Cataracts, which affect 70 million people worldwide, are the most common cause of blindness and severe visual impairment. In one year alone, he performed over 30 million cataract surgeries around the world.

“However, the burden of cataracts is unevenly distributed around the world, as blindness from cataracts is around 10 times more common in countries of the Global South than in high-income countries,” says Dr. Maximilian Wintergarst, head of the Department of Ophthalmology at University Hospitals. . He also conducts research at the University of Bonn, Bonn (UKB). “Against this backdrop, treating cataracts in the Global South remains an important and significant challenge,” says Wintergarst.

Resource conservation concepts needed to improve surgical outcomes

“However, surgical outcomes for cataract surgery in the Global South are sometimes poor, largely due to inadequate training. Depending on country circumstances, up to 30% of patients undergoing surgery ~50% of surgical outcomes are unsatisfactory,” Wintergarst explains.

“Therefore, there is an urgent need to improve surgical outcomes there.” Video recording of surgeries allows for subsequent review of surgical procedures, video-based training, monitoring of surgical quality, and improves surgical outcomes. Proven to be directly linked to improved results. “But it is expensive and generally not available in countries of the Global South.”

An algorithm is currently being developed to automatically analyze surgical videos of cataract surgery techniques commonly used in the Global South as part of a joint research project between UKB, University of Bonn, Sankara Eye Foundation India and Microsoft Research India. . The algorithm should work not only with traditional image material, but also with videos recorded with a smartphone specially attached to a surgical microscope. The researchers hope this will improve cataract surgery outcomes in medical facilities across the Global South over the long term.

There is no AI yet to analyze common surgical techniques in the Global South

As a first important milestone, they conducted a systematic literature review of existing algorithms for automated video analysis of cataract surgery. There are promising initial approaches, but so far only videos from countries with high-income health systems have been used.

Professor Thomas Schulz, study co-author from the research group, said: “The studies reviewed vary widely in quality, have few publicly available datasets, and have little source code publicly available, making them difficult to reproduce. This has been shown to be the case.” b-it and University of Bonn Computer He is Leader of the Science Institute and Principal Researcher of the Lamar Institute for Machine Learning and Artificial Intelligence.

“While our literature review shows that surgical instrument tracking and surgical stage detection work well, the assessment of surgical skill and complications remains a challenge at this time.”

The next step is to develop an algorithm to automatically analyze surgical videos of cataract surgery techniques commonly used in the Global South. As researchers at the University of Bonn and UKB have already shown, recording images from smartphones could improve eye care in India, for example.

“Low-cost smartphone-based video recording of cataract surgery could therefore be a targeted solution to improve training and surgical outcomes in countries of the Global South,” said Director of UKB Eye Clinic. Professor Frank Holtz added.

“However, comparisons between smartphone-based video recording systems and established traditional video recording systems are needed to assess the feasibility, including the quality of the data obtained.”