EUVIMEDEUVIMED entdecken
Open Access

Artificial Intelligence–Based Diagnosis of Kaposi Sarcoma Using Digital Photographs in Dark-Skinned Patients in Uganda

Sarah J Coates iD, Feng Yang, Cody Hill, Zhiyun Xue, Sivaramakrishnan Rajaraman iD, Aggrey Semeere iD, Racheal Ayanga iD, Miriam Laker-Oketta, Helen Byakwaga iD, Robert Lukande, Matthew Semakadde, Micheal Kanyesigye und 8 weitere

JCO Global Oncology · 2026-03 · 12 (3)

DOI: 10.1200/go-25-00236 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JCO Global Oncology, 2026-03. Verfasst von Sarah J Coates und 19 weiteren Autor:innen. Beteiligte Einrichtung: Department of Dermatology, University of California, San Francisco, San Francisco, CA. Hauptthemen laut MEDLINE: Sarcoma, Kaposi, Artificial Intelligence, Photography, Skin Neoplasms. Die Arbeit nennt 29 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NCI NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE This study sought to evaluate the accuracy of artificial intelligence (AI)–based interpretation of digital surface images of skin lesions to diagnose Kaposi sarcoma (KS) among dark-skinned patients in Uganda. METHODS Patients were evaluated at skin biopsy services in Uganda because of clinical suspicion of KS. In a cross-sectional design, 482 consecutive participants were enrolled. Lesions were photographed using a digital camera, and punch biopsies were obtained. Histopathologic interpretation was considered the gold standard. Using training (approximately 70% of images) and validation (approximately 10% of images) sets, we developed a prediction model using a rule-based combination of You Only Look Once version 5 and 8 object detection classifiers. We determined sensitivity, specificity, and positive and negative predictive values of the AI-based prediction model in a test set (approximately 20% of images) and compared these with the accuracy of a dermatologist's visual interpretation of images. RESULTS Four hundred seventy-two participants (1,385 images) were evaluable. Of these, 36% was female; the median age was 34 years; and 94% had HIV, 332 had KS, and 140 had no KS by histopathology. In the test set, the AI-derived prediction model achieved 89% sensitivity (85%-94%) and 51% specificity (40%-61%) for diagnosing KS; the positive predictive value was 81% (75%-86%), and the negative predictive value was 67% (55%-78%). The area under the receiver operating characteristic curve was 0.72. A dermatologist evaluating the same images, with emphasis on sensitivity, achieved a sensitivity of 93% (89%-96%) and a specificity of 19% (11%-28%). CONCLUSION Among dark-skinned patients in Uganda with lesions suspicious for KS, evaluation of digital surface images by an AI-based prediction model produced moderate accuracy for diagnosing KS. While currently inadequate for clinical use, this inaugural assessment is sufficiently promising to justify future evaluation of larger data sets and evolving technologies.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Sarcoma, KaposiArtificial IntelligencePhotographySkin NeoplasmsHumansFemaleUgandaMaleAdultCross-Sectional StudiesMiddle AgedAdolescentYoung AdultSensitivity and Specificity

Autor:innen und Institutionen

  • Department of Dermatology, University of California, San Francisco, San Francisco, CA.
  • Division of Imaging, Diagnostics, and Software Reliability, Office Science and Engineering Laboratories, Center for Devices and Radiological Health, US Food and Drug Administration, Silver Spring, MD.
  • Division of Intramural Research, National Library of Medicine, Bethesda, MD.
  • Infectious Diseases Institute, Makerere University College of Health Sciences, Kampala, Uganda.
  • Pathology Department, Makerere University College of Health Sciences, Kampala, Uganda.
  • HIV Clinic, Masaka Regional Referral Hospital, Masaka, Uganda.

ORCID-Profile: Sarah J Coates, Sivaramakrishnan Rajaraman, Aggrey Semeere, Racheal Ayanga, Helen Byakwaga, Ethel Cesarman, David Erickson, Sameer Antani, Jeffrey Martin

Förderung

  • NCI NIH HHS

Zitieren

APA

Coates, S. J., Yang, F., Hill, C., Xue, Z., Rajaraman, S., Semeere, A., Ayanga, R., Laker-Oketta, M., Byakwaga, H., Lukande, R., Semakadde, M., Kanyesigye, M., Wenger, M., LeBoit, P., McCalmont, T., Cesarman, E., Erickson, D., Maurer, T., Antani, S., & Martin, J. (2026). Artificial Intelligence–Based Diagnosis of Kaposi Sarcoma Using Digital Photographs in Dark-Skinned Patients in Uganda. JCO Global Oncology, 12(3). https://doi.org/10.1200/go-25-00236

Vancouver

Coates SJ, Yang F, Hill C, Xue Z, Rajaraman S, Semeere A, et al. Artificial Intelligence–Based Diagnosis of Kaposi Sarcoma Using Digital Photographs in Dark-Skinned Patients in Uganda. JCO Global Oncology. 2026;12(3). doi:10.1200/go-25-00236

BibTeX
@article{Coates2026,
  author = {Coates, Sarah J and Yang, Feng and Hill, Cody and Xue, Zhiyun and Rajaraman, Sivaramakrishnan and Semeere, Aggrey and Ayanga, Racheal and Laker-Oketta, Miriam and Byakwaga, Helen and Lukande, Robert and Semakadde, Matthew and Kanyesigye, Micheal and Wenger, Megan and LeBoit, Philip and McCalmont, Timothy and Cesarman, Ethel and Erickson, David and Maurer, Toby and Antani, Sameer and Martin, Jeffrey},
  title = {Artificial Intelligence–Based Diagnosis of Kaposi Sarcoma Using Digital Photographs in Dark-Skinned Patients in Uganda},
  journal = {JCO Global Oncology},
  year = {2026},
  volume = {12},
  number = {3},
  doi = {10.1200/go-25-00236}
}
Kostenlos starten →