Artificial intelligence methods available for cancer research.

Area of Science:

• Oncology
• Bioinformatics
• Artificial Intelligence in Healthcare

Background:

• Cancer remains a complex disease with challenges in early diagnosis and treatment selection despite technological advances.
• Large-scale, multi-level datasets offer potential but require advanced bioinformatic tools for clinical application.
• Artificial intelligence (AI) is emerging as a key technology in transforming cancer data into actionable insights.

Purpose of the Study:

• To review the diverse applications of AI methods in cancer research.
• To explore the benefits and limitations of AI in clinical decision support.
• To summarize guidelines for AI in healthcare and discuss future impacts on cancer research.

Main Methods:

• Review of machine learning techniques including Bayesian networks, support vector machines, decision trees, random forests, gradient boosting, K-nearest neighbors, and deep learning.
• Exploration of large language models for novel problem-solving in oncology.
• Analysis of existing reporting guidelines for AI in healthcare to address reproducibility challenges.

Main Results:

• AI methods, including deep learning, demonstrate significant value in predictive, prognostic, and diagnostic cancer studies.
• Large language models are being explored for new applications in cancer research.
• Lack of standardized reporting guidelines hinders the reproducibility of AI studies in clinical settings.

Conclusions:

• AI holds substantial potential to advance cancer diagnosis, treatment, and research.
• Adherence to reporting guidelines is essential for enhancing the reliability and clinical translation of AI models.
• Future cancer research will likely be significantly shaped by the continued development and application of AI.