June 17, 2024
Utility Communications

Four Advanced AI Search Engines for Histopathology Images Fall Short of Clinical Readiness: Study

New research indicates that four cutting-edge AI search engines designed to automate the process of locating and retrieving digital histopathology slides may not be prepared for clinical application. The findings were shared by Dr. Helen Shang, a third-year internal medicine resident and incoming hematology-oncology fellow at the David Geffen School of Medicine at UCLA, and Dr. Mohammad Sadegh Nasr of the University of Texas at Arlington, who co-led the study.

The researchers assessed the performance of the AI algorithms powering the histopathology image databases and discovered that their results were below expectations. Some algorithms demonstrated less than 50% accuracy, which is unacceptable for clinical application practice, according to Dr. Shang.

Despite the growing number of AI algorithms being developed for medical tasks, there is a lack of focus on thorough, external validations. The field also needs to establish a standardized testing protocol for AI algorithms prior to their clinical adoption, Dr. Shang added.

*Note:
1.Source: Coherent Market Insights, Public sources, Desk research
2.We have leveraged AI tools to mine information and compile it