Preprint
Computer Vision

Artificial Intelligence in Oncology: Current Capabilities, Future Opportunities, and Ethical Considerations

Jacob T. Shreve(Department of Oncology, Mayo Clinic, Rochester, MN), Sadia A. Khanani(Department of Radiology, Mayo Clinic, Rochester, MN), Tufia C. Haddad(Department of Oncology, Mayo Clinic, Rochester, MN)
July 1, 2022American Society of Clinical Oncology Educational Book77 citations
DOI

77

Citations

2

Influential Citations

American Society of Clinical Oncology Educational Book

Venue

2022

Year

Abstract

The promise of highly personalized oncology care using artificial intelligence (AI) technologies has been forecasted since the emergence of the field. Cumulative advances across the science are bringing this promise to realization, including refinement of machine learning– and deep learning algorithms; expansion in the depth and variety of databases, including multiomics; and the decreased cost of massively parallelized computational power. Examples of successful clinical applications of AI can be found throughout the cancer continuum and in multidisciplinary practice, with computer vision–assisted image analysis in particular having several U.S. Food and Drug Administration–approved uses. Techniques with emerging clinical utility include whole blood multicancer detection from deep sequencing, virtual biopsies, natural language processing to infer health trajectories from medical notes, and advanced clinical decision support systems that combine genomics and clinomics. Substantial issues have delayed broad adoption, with data transparency and interpretability suffering from AI’s “black box” mechanism, and intrinsic bias against underrepresented persons limiting the reproducibility of AI models and perpetuating health care disparities. Midfuture projections of AI maturation involve increasing a model’s complexity by using multimodal data elements to better approximate an organic system. Far-future positing includes living databases that accumulate all aspects of a person’s health into discrete data elements; this will fuel highly convoluted modeling that can tailor treatment selection, dose determination, surveillance modality and schedule, and more. The field of AI has had a historical dichotomy between its proponents and detractors. The successful development of recent applications, and continued investment in prospective validation that defines their impact on multilevel outcomes, has established a momentum of accelerated progress.

Analysis

Why This Paper Matters

This paper provides a timely and comprehensive overview of artificial intelligence in oncology, a field where AI has already achieved regulatory approval and clinical deployment. It is significant because it bridges the gap between technical AI advances and practical clinical implementation, addressing both the successes (e.g., FDA-approved computer vision) and the persistent challenges (e.g., bias, interpretability). For AI practitioners, it highlights the need for robust validation and fairness in model development, especially as oncology moves toward highly personalized, multimodal AI systems.

The paper also contextualizes the historical dichotomy between AI proponents and detractors, arguing that recent successes and continued investment have created momentum for accelerated progress. This perspective is valuable for understanding the current landscape and anticipating future directions, such as living databases and highly convoluted modeling.

Technical Contributions

  • Computer vision: Several FDA-approved applications for image analysis (e.g., mammography, pathology) are already in clinical use.
  • Multi-cancer detection: Deep sequencing of whole blood enables early detection of multiple cancer types from a single liquid biopsy.
  • Virtual biopsies: AI models infer tumor characteristics from imaging, reducing need for invasive tissue sampling.
  • Natural language processing: Extracts health trajectories from unstructured clinical notes.
  • Clinical decision support: Combines genomics and clinomics (clinical + omics data) to guide treatment selection and dosing.
  • Multimodal modeling: Future direction integrating diverse data types (imaging, genomics, notes) to better approximate biological complexity.

Results

The paper does not report new experimental results but summarizes existing evidence: computer vision tools have demonstrated clinical utility in radiology and pathology, with several receiving FDA clearance. Emerging techniques like liquid biopsy and NLP show promise but lack large-scale prospective validation. No specific accuracy, sensitivity, or specificity metrics are provided.

Significance

This review serves as a roadmap for AI researchers and clinicians, emphasizing that while technical capabilities are advancing rapidly, real-world adoption hinges on solving issues of interpretability, bias, and data transparency. The paper’s forward-looking discussion of living databases and multimodal AI sets a research agenda for the next decade. For Neura Market’s audience, it underscores the importance of building fair, explainable, and clinically validated AI systems to realize the promise of personalized oncology.

oncologycomputer-visiondeep-learningclinical-decision-supportnatural-language-processingmultiomicsethicshealthcare-ai
MultimodalComputer Vision

Related Papers