Thailand’s Siriraj Hospital expands AI use in mammogram, CT brain services

AI was initially used in radiology, pathology, and ophthalmology.

Thailand’s Siriraj Hospital will expand the integration of artificial intelligence (AI) into its mammogram and CT brain analysis services by the end of the year, building on the technology's existing implementation in the facility’s key areas, including radiology, pathology, and ophthalmology.

The initiative aims to improve accuracy in detecting abnormalities, diagnosing diseases, and optimising medical personnel workflow.

Moreover, the hospital has utilised AI to enhance centroblast cell counting in a study for precise lymphoma grading.

Through this adoption, researchers trained the You Only Look Once (YOLO) model, an object detection algorithm, using 1,669 centroblast cell images from the whole slide images (WSI) of five patients.

According to the hospital, the YOLOv5x6 model achieved a precision of 0.808, a recall of 0.776, a mean average precision (mAP) of 0.800 at 0.5 Intersection over Union (IoU), and an overall mAP of 0.647.