Skip Ribbon Commands
Skip to main content

Lung Imaging Program

Lung Imaging Program: AI and Deep Learning in Radiology – Team SCRATCH

Paired CXR images showing segmentation of lungs, heart and bowel gas. Orange spot ("heat map") illustrates the left pleural effusion as the identified pathology in the CXR by an Artificial Intelligence algorithm.

The use of Artificial Intelligence (AI) and Deep Learning (DL) in radiology has grown exponentially in recent years. The SGH Campus Radiology AI Team CHaracters (SCRATCH) is a campus-level collaboration to enhance the involvement of radiologists in AIDL research and translation to clinical practice. The inaugural team comprises radiologists from SGH (Dr Lionel Cheng, Dr Tran Nguyen Tuan Anh) and NCCS (Dr Thng Choon Hua, Dr Gideon Ooi). SCRATCH is currently working with the Institute of High-Performance Computing in A*STAR to refine AI algorithms for use in radiology. Our current project focuses on developing an AI tool to detect a broad range of pathology on chest radiographs, with an aim to deploy this into the clinical workflow. We are also collaborating with the larger SingHealth AI initiatives such as AIMx. If you have an interest in imaging informatics, image analytics, natural language processing, deep learning, and data science, or perhaps have an insatiable desire to annotate radiological images, feel free to contact the team consisting of Dr Lionel Cheng and Dr Tran Nguyen Tuan Anh for more information.