Estimating chemotherapy side effect in histopathologically diagnosed patients over 65 with metastatic breast cancer using deep learni̇ng

Objective: In this study, it is aimed to make an analysis study to predict the side effects of chemotherapy by using machine learning and deep learning methods in the treatment stage of histopathologically diagnosed breast cancer patients over 65 years of age. With this study, it was aimed to give physicians ideas about whether to give chemotherapy to the patient. Material and methods: The TensorFlow library, which is an open source library developed by Google, will be used and the library which contains many deep learning algorithms has been developed by using ptyhon language. The library is developed by more than 1000 developers. Success in deep learning methods depends on determining the data feature set to be selected. For this study, a data set consisting of the most frequently observed features and side effects of chemotherapy was selected. On the developed system, relevant information about the existing patients was entered first, and then the results of chemotherapy treatment in these patients were indicated. As a result of this process, the related data set was subjected to learning process by using machine-learning methods, and as a result of this procedure, when a new patient arrives, it was tried to predict whether chemotherapy would be applied for this patient. Results: When there is sufficient data for the deep learning model, deep learning approaches can be successfully applied to clinical decision making and treatment approaches. When appropriate, accurate and sufficient data set is applied, more successful results will be achieved and the success rate will reach 90%. Conclusion: By using deep learning techniques in the diagnosis of breast cancer, the accuracy of the diagnostic method and the appropriate treatment method can be increased. It is thought that the results presented in this study will lead to further research in the diagnosis and treatment of other diseases by the deep learning technique.