Computer Assisted Diagnosis and Intervention based on artificial intelligence and multi-disciplinary computation anatomy model

Abstract

This presentation gives computer assisted diagnosis and intervention systems based on artificial intelligence and multi-disciplinary computational anatomy models. Multi-disciplinary computation anatomy is new academic area that tries to understand human anatomy from four different disciplinaries including space, time, function and pathology and assist diagnostic and therapeutic procedures based on multidisciplinary anatomical structure recognition. Multi-disciplinary computational anatomy is asking us to understand anatomical structures in the space spanned by these four disciplinaries and to provide clinicalassistance information. Artificial intelligence or machine learning techniques can be utilized for achievingaccurate analysis and information presentation.This talk convers the following topics: (a) anatomical structure analysis based on machine learning, (b) multi-scale registration covering 10 times scale gap, (c) laparoscopic surgery assistance based on anatomical structure analysis and (d) real-time endoscopic diagnosis assistance using artificial intelligence technique. In anatomical structure analysis, we present machine learning framework for whole-body anatomical structure recognition based on machine learning. This will cover macroscopic anatomical structure to microscopic structures. In multi-scale registration, we provide a method for registering macro and microscopic anatomical structures. We also show super-resolution technique that enables us to convert clinical CT images to very high-resolution images using micro CT image information. Surgical navigation system using analysis results of multi-disciplinary computational anatomical models will be demonstrated in this talk. Finally, we demonstrate real-time colonoscopic procedure assistance system.