An artificial intelligence tool has been developed to help make real-time diagnoses during surgery, improving the quality of images to increase the accuracy of rapid diagnostics, a new study has shown.
According to the study published in Nature Biomedical Engineering, the method leverages artificial intelligence to translate between frozen sections and the gold-standard approach.
“We are using the power of artificial intelligence to address an age-old problem at the intersection of surgery and pathology,” said the corresponding author of the study Faisal Mahmood, PhD, of the Division of Computational Pathology at US-based Brigham and Women’s Hospital.
“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,” he added.
To make final diagnoses, pathologists use formalin-fixed and paraffin-embedded (FFPE) tissue samples — this method preserves tissue in a way that produces high-quality images but is labour-intensive and can take several days, according to the study.
Mahmood and co-authors developed a deep-learning model that can be used to translate between frozen sections and more commonly used FFPE tissue.
As a result of their research, the team demonstrated that the method can subtype different types of cancer, including gliomas and non-small-cell lung tumours.
Moreover, the authors note that prospective clinical studies should be conducted in the future in order to confirm and validate the AI method in real-world hospital settings in terms of diagnostic accuracy and surgical decision-making.
“Our work shows that AI has the potential to make a time-sensitive, critical diagnosis easier and more accessible to pathologists,” said Mahmood.
“And it could potentially be applied to any type of cancer surgery. It opens up many possibilities for improving diagnosis and patient care,” Mahmood added.
(Only the headline and picture of this report may have been reworked by the Business Standard staff; the rest of the content is auto-generated from a syndicated feed.)