Computer-aided diagnosis (CAD) in medical image analysis (MIA) is one of the expected fields that artificial intelligence (AI), especially deep learning (DL), improves the performance. However, DL alone is not enough to analyze medical images. DL process is just one processing step in the overall CAD system. Rather, the main role of researcher is to develop methods to synthesize data that are processed by the DL models, and methods that derive satisfactory results from the inferred results of the DL model. In order to discuss how to develop AI-based-CAD systems, I will introduce some CAD applications using DL models. It will include fatigue fracture detection in 3-D computer tomography (CT) images, tooth recognition in dental panorama radiograph, finger joint detection in hand radiograph. Through the applications, I am going to summarize the strategy to develop AI-based-CAD systems.
Syoji Kobashi received BE in 1995, ME in 1997, and Doctor of Engineering in 2000, all from Himeji institute of Technology. He was an assistant professor at Himeji Institute of Technology (2000-2004), an associate professor (2005-2016), currently a professor (2016-) and the manager of advanced medical engineering research center (2016-), University of Hyogo. And, he was a guest associate professor at Osaka University, WPI immunology frontier research center (2010-2016), and was a visiting scholar at University of Pennsylvania (2011- 2012). His research interests include computer-aided diagnosis in medical images and artificial intelligence. He received 16 international awards, including Lifetime Achievement Award (WAC, 2016), Franklin V. Taylor Memorial Award (IEEE-SMCS, 2009). He has been serving on Program Chair of WAC2021, Publication Chair SMC2018, of and many others. Moreover, he is an editor-at-large of Intelligent Automation & Soft Computing journal, an editor-in-chief of International Journal of Biomedical Soft Computing and Human Sciences, etc. He is the senior member of IEEE.