New Strategies and Prognosis for Diagnosis and Treatment of Pituitary Adenomas: Research Based on Medical Big Data and AI

Ming Feng, MD, Peking Union Medical College Hospital, Chinese Academy of Medical Science

Pituitary adenoma is one of the most common tumors in the nervous system. Because of the atypical early symptoms and the lack of understanding of clinicians, pituitary adenoma is easily misdiagnosed as hypertension, diabetes and other common diseases, because of which treatment is delayed. And how to accurately carry out early diagnosis, surgical efficacy prediction, treatment strategy formulation is still lack of guidelines to regulate clinical diagnosis and treatment. However, it is difficult to make accurate judgment only based on the personal clinical experience of doctors. Based on the massive clinical multimodal data, this study uses weak supervision learning to extract and structure the medical record information. And machine learning is used to extract the MRI image features. Then a multidimensional diagnosis and prediction model of pituitary adenoma is established. Finally, the prospective cases will be used to verify and optimize the model. This study can assist clinicians to make accurate diagnosis and individualized treatment strategies for pituitary adenoma, to predict postoperative tumor recurrence and guide follow-up. This study breaks through the limitations of the existing diagnosis and treatment technology and mode, and provides new methods and ideas for the accurate diagnosis, individualized treatment and evaluation of pituitary adenoma by using the methods of big data and artificial intelligence.