Skip to main content
Fig. 4 | Infectious Agents and Cancer

Fig. 4

From: Validation in Zambia of a cervical screening strategy including HPV genotyping and artificial intelligence (AI)-based automated visual evaluation

Fig. 4

Step by step precancer/cancer stratification: low prevalence Zambia study population, after knowing HIV status, after knowing HPV status. This figure explains step by step risk discrimination in a population after knowing each screening test result. In total, there are 931 patients screened in this study in Zambia (67 excluded due to no J8 image). After testing for HIV, precancer/cc risk of HIV+ patients increases to 15% while the risk decreases to 2.2% for HIV-negative patients. After HIV, if the patients get tested for HPV genotype, we can observe even further risk discrimination. A 15% precancer/cc risk of HIV-positive patients increases to 48% if they are positive for HPV type 16. Similarly, the risk decreases from 15 to 1.4% if the patients are HPV HR-negative. For HIV-negative patients, the precancer/cc risk increases from 2.2 to 22% if they are positive for HPV type 16. Similarly, if HIV-negative patients are negative for any HPV HR types then their precancer/cc risk decreases to 0.23%. In the above figure, the number of patients (N) observed in each category and their precancer/cc risk are displayed separately for each category. *67 individuals have no J8 images, they have images captured by other camera types. **37 of HIV+ individuals and 28 of HIV− individuals, and 2 of the HIV missing individuals do not have any J8 images (which add up to 67 from the previous step), so they are not included in this analysis. 22 individuals have missing HIV result (317 HIV+, 592 HIV − , and 22 missing HIV will add up to the previous step, screening population)

Back to article page