Researchers Ankita Murmu, Balázs Győrffy and others. conducted a review titled “Artificial Intelligence Methods Available for Cancer Research.” This review was published in Frontiers of Medicine, Volume 18, Issue 5.
Cancer is a heterogeneous and multifaceted disease with significant global impact, and early diagnosis and selection of effective treatments remain challenging. The availability of large-scale multilevel datasets has made artificial intelligence (AI) technologies increasingly important in cancer research. This review summarizes various AI techniques, including traditional machine learning (decision trees, random forests, gradient boosting, support vector machines, Bayesian networks, K-nearest neighbors, etc.), neural network algorithms (deep learning, natural language processing), and large-scale language models. These methods are widely applied in fields such as cancer prediction, diagnosis, prognosis, treatment response prediction, and drug development. This review also discusses relevant guidelines for AI in healthcare, such as SPIRIT-AI, CONSORT-AI, and MI-CLAIM, and points out current challenges such as data heterogeneity, bias, privacy issues, lack of standardized reporting, and difficulties in clinical translation. We also look forward to future directions such as explainable AI, personalized treatments, rare cancer research, and non-invasive detection tools.
For more information, please read the full paper below. https://doi.org/10.1007/s11684-024-1085-3
