Artficial Intelligence

The participants are expected to have basic knowledge in medical imaging and programming. It is not necessary to be able to do programming by themselves. But some feeling of programming would be beneficial for them to follow the practical guidance and to run the scripts provided by the course. Artificial intelligence is an evolving disciplinary to assist and extend the human ability to solve lots of problems. It has also been developed for physicians to resolve the critical information from complex medical imaging data, which improves the accuracy and robustness of diagnosis and treatment. The course is going to provide basic knowledge of artificial intelligence, machine learning and deep learning for the postgraduate students of medicine and medical sciences. Specifically, it will focus on translating these knowledge to the processing and computer aided diagnosis on medical imaging. The course will discuss applying artificial intelligence on several medical imaging modalities such as CT, MRI, PET and ultrasound. It will cover a number of clinical applications such as cancer lesion detection, neurological differential diagnosis and early diagnosis. In addition to the basic concepts of artificial intelligence and their clinical applications, this course will provide several practical lessons. It will give a guidance of simple programing using language python. Then it will provide scripts of several real clinical problems. The participants can run these scripts to feel the conditions, benefits and problems of artificial intelligence. An in-depth understanding of the theoretical part can achieved after these practice. In this course, the participants will understand and practice the following concepts of artificial intelligence in medical imaging: I. Recall basic concepts of medical imaging, understand the difference between human and machine to look at imaging data, understand the possibilities and restrictions of applying artificial intelligence. II. Intuitively understand basic concepts of machine learning technologies, such as clustering, classification, support vector machine, decision tree and so on. III. Understand the concept of feature extraction. In particular, radiomics as features of medical imaging will be introduced and discussed. IV. Intuitively understand basic concepts of neural network. In particular, the concepts of deep learning, such as constitutional neural network, autoencoder, will be introduced. V. Understand the basic concepts of computer aided diagnosis using artificial intelligence, get an overview the clinical applications of computer aided diagnosis such skin lesion diagnosis, early dementia diagnosis an so on. VI. Get the basic knowledge of the script language python. Understand the structure of a runnable file and the meanings of basics codes such as data structure, flow control. VII. Able to run the scripts provided by the course. Understand the relation between the codes and the results. Passed / not passed In order to complete the course, the students are expected to participate in all lectures and on the last day, there is a multiple choice short exam as a feedback to teachers on what has been learned, and a discussion. The students are going to be asked for an anonymous evaluation of teachers, too.