Groundbreaking AI tool gives real-time diagnosis during surgery

Researchers come up with a revolutionary system that uses AI to improve image quality and improves diagnosis accuracy

Tech Desk January 04, 2023

Dil Faroza’s eye surgery was never performed. PHOTO: AFP/FILE

A new study from Mahmood Lab at the Brigham and Women’s Hospital along with collaborators from Bogazici University created a method that uses AI to translate between frozen sections and the gold-standard approach, which improves the quality of images for a rapid and accurate diagnosis.

During surgery, surgeons often have to collect, analyse and diagnose a disease while the patient is on the operating table. Samples have to be sent to the pathologist for a speedy and accurate examination, using two methods; the gold-standard approach which can take long, and a method involving freezing tissue which can complicate a diagnosis.

Corresponding author Faisal Mahmood, PhD, of the Division of Computational Pathology at BWH said, “We are using the power of artificial intelligence to address an age-old problem at the intersection of surgery and pathology. Making a rapid diagnosis from frozen tissue samples is challenging and requires specialised training, but this kind of diagnosis is a critical step in caring for patients during surgery.”

Pathologists use a formalin-fixed and paraffin-embedded tissue sample which preserves it and produces high-quality images. However, this process is not only laborious but can take 12 to 48 hours.

The AI would be able to translate between frozen sections and more commonly used FFPE tissue. The method was tested with pathologists making a diagnosis using images that had gone through the AI method and traditional cryosectioning images.

Published in Nature Biomedical Engineering, the findings of the study showed that AI improved image quality and diagnostic accuracy.

Mahmood is eager to try out the method in a real hospital setting, saying, “Our work shows that AI has the potential to make a time-sensitive, critical diagnosis easier and more accessible to pathologists. And it could potentially be applied to any type of cancer surgery. It opens up many possibilities for improving diagnosis and patient care.”