AI in Oncology: Transforming Cancer Care with Advanced Technologies

The integration of artificial intelligence (AI) in oncology has evolved from a concept to a tangible reality, presenting numerous opportunities for revolutionizing cancer care. Ongoing studies indicate that AI's applications in cancer diagnosis, treatment, and research are expanding.

Dr. Tufia C. Haddad from Mayo Clinic in Rochester, Minnesota, emphasizes the current prevalence of AI in the field of cancer diagnostics. Machine learning and deep learning models are enhancing the accuracy and efficiency of cancer diagnoses, particularly in the realm of oncology imaging.

Several AI technologies have gained approval from the US Food and Drug Administration (FDA) for oncology applications, with a notable focus on radiology. In 2021, Luchini et al reported that AI devices approved for oncology were predominantly utilized in radiology (54.9%) and pathology (19.7%). Breast cancer, in particular, saw significant AI implementation (31.0%).

AI's impact is evident in breast radiology, where it assists in tasks like breast density determination, mammogram quality assessment, and risk stratification. Dr. Laurie Margolies from Mount Sinai in New York notes that AI is playing a crucial role in identifying breast cancers and assessing atherosclerotic disease risk.

Beyond radiology, AI has demonstrated effectiveness in colorectal cancer screening and diagnosis. The FDA's authorization of GI Genius in 2021 marked a milestone, as it became the first device to use AI for lesion detection during colonoscopy.

Moreover, AI is making strides in pathology labs, where it aids in reading digital pathology slides to refine cancer diagnoses. Dr. Olivier Elemento from Weill Cornell Medicine in New York highlights AI's role in drug discovery, as some companies leverage AI to identify novel and safe targets in oncology and design new drug candidates.

Optimizing AI for Clinical Use

Dr. Haddad underscores AI's role in optimizing radiation treatment planning, enhancing tumor and organ contouring efficiency, and remotely assessing patients' symptoms and vital signs for supportive care.

Ongoing research explores additional AI applications in oncology, including identifying patients at high risk for pancreatic cancer using abdominal imaging and electronic health records. AI holds promise in transforming biomarker assessment and molecular characterization of cancers, potentially eliminating the need for invasive biopsies.

Large language models (LLMs) are emerging as an exciting area in AI, with the potential to answer medical questions accurately. Medical centers possess vast datasets that could train high-quality LLMs for personalized treatment recommendations.

Despite these advancements, challenges persist in optimizing AI for clinical use. Issues such as data sharing, implementation in electronic health record software, user interfaces, and resource costs need addressing. Ethical considerations, guidelines, and guardrails are crucial to ensure the safe and unbiased use of AI in oncology.

In conclusion, AI in oncology represents a transformative moment with the emergence of generative AI and LLMs, poised to have an unprecedented impact on cancer research and care delivery.