Details

Autor(en) / Beteiligte

• Ma, Ling
• Hu, Zhijun
• Yue, Ding
• Wu, Guangyu
• Shi, Xiaohua
• Sirejiding, Shalayiding
• Liu, Kaijiang

Titel

Multimodal federated learning framework evaluation for lymph node metastasis in gynecologic malignanciese

Ist Teil von

• 2023 IEEE 4th International Conference on Pattern Recognition and Machine Learning (PRML), 2023, p.269-273

Ort / Verlag

IEEE

Erscheinungsjahr

2023

Quelle

IEEE Electronic Library (IEL)

Beschreibungen/Notizen

• Gynecologic malignancies, including cervical cancer, endometrial cancer, and ovarian cancer, represent a significant threat to women's health, and their diagnosis and treatment are crucial aspects of global public health. While imaging techniques play an important role in diagnosing gynecologic tumors, early detection and accurate diagnosis present challenges. Hence, we propose a framework for assessing lymph node metastasis based on multi-modal federated learning. By utilizing MRI images, we develop a hybrid neural network model that effectively predicts lymph node metastasis by combining a multi-layer perceptron (MLP) and a convolutional neural network (CNN). Through federated learning, we ensure data privacy while facilitating data sharing and collaborative computation among multiple hospitals. Validation using a dataset from a single hospital illustrates the considerable benefits of the model in enhancing the sensitivity and specificity of lymph node metastasis diagnosis in gynecologic malignancies. This research offers a novel and efficient approach for assessing lymph node metastasis in gynecologic malignancies, showcasing the vast potential of federated learning in the analysis of medical images.