Medical Image Diagnostics based on Deep Transfer Learning

• Main Authors: Guan-Hong Shih, 施冠宏
• Other Authors: Chuan-Jun Su
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/fvstx4

碩士 === 元智大學 === 工業工程與管理學系 === 107 === In recent years, artificial intelligence and deep learning have developed rapidly and have achieved good results. At present, many fields are slowly introducing artificial intelligence to assist their work and make their work more efficient. The development of artificial intelligence in medical care has recently been introduced by more and more hospitals to assist doctors. Reducing the burden on doctors, allowing doctors to use reduced time and spirit to treat patients, and allowing more people to access medical services in the same amount of time. This study uses chest X-rays as an example to determine whether there is pneumonia and to determine which type of pneumonia. However, in order to correctly judge X-ray films, professional medical knowledge and rich experience are required to correctly and quickly determine what type of pneumonia is correct. So we try to construct a model to judge. Transfer learning can be used for a small amount of data. In medical imaging, there are problems such as patient privacy and research ethics. Therefore, it is relatively difficult to obtain data, so the amount of data is also small. Finally, a user interface is provided to facilitate the use of medical personnel, and a reference is given from the side to achieve the purpose of reducing workload and increasing efficiency of consultation.