Although largely treatable and preventable, cervical cancer kills an estimated 340,000 women each year. Of the women who succumb to the disease annually, 90% live in low- and middle-income countries (LMICs) where many women are not screened or screened with a technique known as Visual Inspection with Acetic Acid (VIA) testing, which has low rates of accuracy. Healthcare providers in LMICs need new tools to spot and treat pre-cancerous lesions more easily, effectively, and affordably—especially because alternatives such as annual Pap smears require expensive lab infrastructure that is largely out of reach.
To help bridge gaps, GH Labs developed an Automated Visual Evaluation (AVE) smartphone-based app that applies machine learning to detect pre-cancerous lesions. Our integrated software incorporates data entry, image capture and analysis, and data export to auto-capture and interpret a quality in-focus image in less than two minutes, computing the likely presence of pre-cancer/cancer with high accuracy.
Research has also shown AVE can detect pre-cancerous lesions amid challenging cervical confounders, confirming high accuracy among women living with HIV (WLHIV) and HIV-negative women. In addition, longitudinal cohort studies demonstrate the potential superior accuracy of AVE to current clinical tools.
With our partners, GH Labs continues to assess AVE as an effective decision aid alongside its geographic portability and provider and patient acceptance in five African countries and India. We will publish results following the study’s completion in March 2023.