Federated Lung Cancer Prediction Using Histopathological Medical Images
Browne Judith Ayekai, Chen Wenyu, Mamo Tadiyos Hailemichael, Linda Delali Fiasam, Agbesi Victor Kwaku, Fortune Agbley, Williams Ayivi, Francis Sam, Juliana Mantebea Danso, Delanyo Kwame Bensah Kulevome, Cobbinah Bernard Mawuli · 2022 19th International Computer Conference on Wavelet Active Media Technology and Information Processing (ICCWAMTIP) · 2022
Machine learning is increasingly significant in health science because it can infer valuable information from high-dimensional data. However, combining research and patient data from various organizations and hospitals is frequently not practical due to privacy concerns. In this research, we conduct a study of federated learning for lung cancer prediction to demonstrate the effectiveness of collaborative and decentralized learning in a context where data is privacy preserved. We also conducted visual interpretation using GradCAM to validate the robust performance of the decentralized method in predicting lung cancer.