AI-Supported Approach May Improve Cervical Cancer Screening in Low-Resource Settings

Limited accuracy of visual inspection with acetic acid, together with uneven access to high-quality human papillomavirus (HPV) testing and follow-up services, can hinder identification of women at risk of severe precancerous lesions. To improve detection and triage under these constraints, a new study evaluated an AI-supported screening strategy for resource-limited settings.

The Barcelona Institute for Global Health (ISGlobal) led the study of the HPV-Automated Visual Evaluation (PAVE) strategy, which combines HPV testing using self-collected samples, extended genotyping, and automated assessment of cervical images. The testing identifies 13 carcinogenic HPV types, with women who test positive proceeding to cervical imaging using a portable colposcope. A deep-learning algorithm then classifies the images as severe, indeterminate, or normal.

The cross-sectional study enrolled more than 50,000 women aged 25 to 49 years between 2022 and 2025 across Brazil, Cambodia, the Dominican Republic, El Salvador, Eswatini, Honduras, Malawi, Nigeria, and Tanzania. Women with positive HPV results underwent follow-up imaging and biopsy to confirm the diagnosis.

Among nearly 49,000 women with valid results, more than one in four tested positive for HPV, and most attended follow-up assessment. The study identified more than 500 cases of severe precancerous lesions, known as CIN3+, with a higher prevalence among women living with HIV. Risk varied with both HPV type and the severity of abnormalities identified by the AI system. HPV16 carried the highest risk, particularly when accompanied by severe automated image findings.

The research was conducted within the PAVE Consortium and published in The Lancet Obstetrics, Gynaecology & Women’s Health. The findings showed that self-sampling, HPV extended genotyping, and automated visual evaluation could support risk-based management of HPV-positive women. The approach also supports rapid testing and real-time image analysis in a same-day screen–triage–treat model.

“Among nearly 49,000 women with valid results, more than one in four tested positive for HPV, and most attended follow-up assessment. We identified over 500 cases of severe precancerous lesions—known as CIN3+, the most advanced stage before cervical cancer—with higher prevalence among women living with HIV,” said Silvia de Sanjosé, associate researcher at ISGlobal and first author of the study.

“Even as HPV vaccination expands globally, effective screening will remain essential for decades, particularly for populations that have not benefited from vaccination. The PAVE approach offers a feasible solution to extend high-quality cervical cancer screening to underserved populations, contributing to global efforts to reduce cervical cancer incidence and mortality,” said Mark Schiffman, epidemiologist at the National Institutes of Health (NIH) and co-leader of the study.

Related Links
ISGlobal