Recently, Artificial Intelligence (AI) accomplished many complex tasks in medicine. As a representative of the new generation of AI technology, deep learning has been widely used in the diagnosis and treatment of malignant tumors. However, the problem with many existing models is the lack of transparency and interpretability. These powerful models have been generally considered “black boxes”, not providing any information about what exactly makes them arrive at their predictions. In addition, the limited information expression ability of single modality and static data restricts the application of AI in the diagnosis and treatment of malignant tumors.
Addressing the above key problems will help to improve the interpretability of deep learning models, maximize the use of multimodal and dynamic data and fully leverage the power of AI. Therefore, we are proud to offer this platform to disseminate state-of-the-art deep learning techniques for endocrine tumor diagnosis and treatment.
This Research Topic aims to develop prediction models for malignant tumors based on cross-modal deep learning technology, investigate interpretable deep learning methods for early identification of malignant tumors, use multimodal deep learning to automatically identify pathological types and detailed features of tumors based on dynamic or high-resolution data, and develop visualization methods that can be used to enhance the interpretability of deep learning models.
The work presented here will highlight the diversity and advancement of research conducted across the field. We welcome Original Research, Reviews, Perspectives, Methodological and Systematic Reviews on the following sub-topics, including but not limited to:
• The prediction of lymph node metastasis (LNM) in malignant tumors based on cross-modal deep learning technology;
• Early identification of malignant tumors based on "eXplainable Artificial Intelligence (XAI)" technology;
• Automatic identification of pathological types of malignant tumors in endocrine systems using multimodal machine learning or deep learning models;
• Evaluation of therapeutic effects of endocrine system malignancies by deep learning and multimodal ultrasound imaging;
• Automatic identification of detailed features of malignancies such as capsular invasion, extraglandular invasion, or microcalcifications of malignant tumors in endocrine systems based on dynamic or high-resolution data;
• Risk assessment and prognosis judgment of minimally invasive treatment of endocrine tumors combined with artificial intelligence.
Recently, Artificial Intelligence (AI) accomplished many complex tasks in medicine. As a representative of the new generation of AI technology, deep learning has been widely used in the diagnosis and treatment of malignant tumors. However, the problem with many existing models is the lack of transparency and interpretability. These powerful models have been generally considered “black boxes”, not providing any information about what exactly makes them arrive at their predictions. In addition, the limited information expression ability of single modality and static data restricts the application of AI in the diagnosis and treatment of malignant tumors.
Addressing the above key problems will help to improve the interpretability of deep learning models, maximize the use of multimodal and dynamic data and fully leverage the power of AI. Therefore, we are proud to offer this platform to disseminate state-of-the-art deep learning techniques for endocrine tumor diagnosis and treatment.
This Research Topic aims to develop prediction models for malignant tumors based on cross-modal deep learning technology, investigate interpretable deep learning methods for early identification of malignant tumors, use multimodal deep learning to automatically identify pathological types and detailed features of tumors based on dynamic or high-resolution data, and develop visualization methods that can be used to enhance the interpretability of deep learning models.
The work presented here will highlight the diversity and advancement of research conducted across the field. We welcome Original Research, Reviews, Perspectives, Methodological and Systematic Reviews on the following sub-topics, including but not limited to:
• The prediction of lymph node metastasis (LNM) in malignant tumors based on cross-modal deep learning technology;
• Early identification of malignant tumors based on "eXplainable Artificial Intelligence (XAI)" technology;
• Automatic identification of pathological types of malignant tumors in endocrine systems using multimodal machine learning or deep learning models;
• Evaluation of therapeutic effects of endocrine system malignancies by deep learning and multimodal ultrasound imaging;
• Automatic identification of detailed features of malignancies such as capsular invasion, extraglandular invasion, or microcalcifications of malignant tumors in endocrine systems based on dynamic or high-resolution data;
• Risk assessment and prognosis judgment of minimally invasive treatment of endocrine tumors combined with artificial intelligence.
