EUVIMEDEUVIMED entdecken
Crossref

From Prediction to Delivery: Reframing Oncology Artificial Intelligence for Global Cancer Care

Emma Ruoyi Zhang iD, James Fan Wu iD, Milit S Patel iD, Erin Jay G Feliciano iD, Nishwant Swami iD, Nancy Y Lee iD, Fabio Ynoe Moraes iD, Edward Christopher Dee iD

JCO Global Oncology · 2026-11 · 12 (11)

DOI: 10.1200/go-26-00468 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JCO Global Oncology, 2026-11. Verfasst von Emma Ruoyi Zhang und 7 weiteren Autor:innen. Beteiligte Einrichtung: University of California Los Angeles, Los Angeles, CA. Die Arbeit nennt 18 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence (AI) in oncology should be judged not only by how well models perform, but also by whether these advances improve cancer care delivery.Many oncology AI studies emphasize technical metrics such as discrimination, area under the curve, or concordance with expert review.These metrics are necessary but insufficient, particularly in global oncology.A delivery-oriented approach instead begins with a practical question: what health-system bottleneck is the AI tool trying to solve?For health systems facing limited diagnostic capacity, uneven specialist access, costly molecular testing, and fragmented follow-up, the more consequential question is whether AI can help more patients receive timely, safe, and effective care.[1][2][3] Scalability should not be confused with simplicity.Clinical AI depends on data pipelines, hardware, interoperability, governance, cybersecurity, training, maintenance, and financing.A tool that identifies more cancers without strengthening downstream care may expose unmet need without resolving it.2,3 Global oncology AI therefore requires delivery-oriented end points, such as screening completion, diagnostic timeliness, treatment access, and quality of life, where feasible.Selected AI presentations at the ASCO 2026 Annual Meeting are notable because they move beyond prediction alone.[4][5][6][7] They suggest that AI could support global cancer care by extending scarce specialist capacity, supporting nonspecialist clinicians, extracting more value from routine clinical data, and reducing reliance on unevenly available infrastructure.This piece uses these abstracts to examine AI as delivery infrastructure for global oncology.Lung cancer screening illustrates the gap between identifying eligible patients and delivering completed care.Eligible patients may exist within a health system but still not be reliably contacted, scheduled, scanned, or followed.Even when eligibility is established, screening rates remain low. 4 Zhang et al 5 provide a useful example of how AI may extend centralized resources from a wellresourced health system to rural facilities.Their navigator-led, AI-assisted platform used electronic health record tools to identify patients eligible for lung cancer screening.Navigators then coordinated outreach, scheduling, and follow-up.5 The study is important because AI was embedded within a broader care-delivery intervention rather than deployed as a standalone predictive tool.This distinction matters for global oncology.AI can flag missed opportunities, but navigators, referral pathways, imaging capacity, and follow-up systems are what convert identification into completed care.In resource-constrained settings, screening programs often face fragmented records, limited primary-care capacity, geographic barriers, and incomplete follow-up after abnormal results.2,3 A centralized AI-assisted platform, especially when paired with navigators and hub-and-spoke networks, points toward a scalable model.This example should still be interpreted cautiously.Generalizability depends on electronic health record quality, data completeness, navigation workforce, imaging access, and follow-up pathways.Without these enabling systems, AI may identify unmet need without resolving the barriers that prevent diagnosis or treatment.AI is not a panacea for global oncology delivery, but it may be valuable when integrated into broader health-system strengthening.

Autor:innen und Institutionen

  • University of California Los Angeles, Los Angeles, CA.
  • Division of Hematology and Oncology, Department of Medicine, Medical College of Wisconsin, Milwaukee, WI.
  • Department of Molecular Biosciences, The University of Texas at Austin, Austin, TX.
  • Department of Hematology/Oncology, Northwell Health Cancer Institute, New Hyde Park, NY.
  • School of Medicine and Public Health, Ateneo de Manila University, Pasig City, Philippines.
  • Department of Cancer Medicine, The University of Texas MD Anderson Cancer Center, Houston, TX.

ORCID-Profile: Emma Ruoyi Zhang, James Fan Wu, Milit S Patel, Erin Jay G Feliciano, Nishwant Swami, Nancy Y Lee, Fabio Ynoe Moraes, Edward Christopher Dee

Zitieren

APA

Zhang, E. R., Wu, J. F., Patel, M. S., Feliciano, E. J. G., Swami, N., Lee, N. Y., Moraes, F. Y., & Dee, E. C. (2026). From Prediction to Delivery: Reframing Oncology Artificial Intelligence for Global Cancer Care. JCO Global Oncology, 12(11). https://doi.org/10.1200/go-26-00468

Vancouver

Zhang ER, Wu JF, Patel MS, Feliciano EJG, Swami N, Lee NY, et al. From Prediction to Delivery: Reframing Oncology Artificial Intelligence for Global Cancer Care. JCO Global Oncology. 2026;12(11). doi:10.1200/go-26-00468

BibTeX
@article{Zhang2026,
  author = {Zhang, Emma Ruoyi and Wu, James Fan and Patel, Milit S and Feliciano, Erin Jay G and Swami, Nishwant and Lee, Nancy Y and Moraes, Fabio Ynoe and Dee, Edward Christopher},
  title = {From Prediction to Delivery: Reframing Oncology Artificial Intelligence for Global Cancer Care},
  journal = {JCO Global Oncology},
  year = {2026},
  volume = {12},
  number = {11},
  doi = {10.1200/go-26-00468}
}
Kostenlos starten →