Meta-analysis finds AI tools show high accuracy for Schistosoma haematobium detection in sub-Saharan AfricaAI tools show high accuracy for detecting a parasitic infection in Africa

PLoS neglected tropical diseases Published May 12, 2026 Study authors: Desale Sisay, Alemu Getaneh, Hailu Tadesse PubMed ↗ DOI ↗ Editorial oversight: Dr. Lars van Dijk, PhD · Surgical, Procedural & Diagnostic

AI-generated summary of the cited source, checked by automated accuracy review. How we work

Key Takeaway

Consider AI-assisted tools for Schistosoma haematobium screening in endemic areas, but note high heterogeneity limits generalizability.

This is a meta-analysis of 10 studies (15 datasets, 5,564 urine samples) conducted in sub-Saharan Africa. It synthesized the diagnostic accuracy of AI-assisted tools for detecting Schistosoma haematobium infection, compared to microscopy and/or molecular reference standards. The authors reported a pooled sensitivity of 88% (95% CI 83%-91%) and a pooled specificity of 89% (95% CI 83%-93%). The pooled diagnostic odds ratio was 54.00 (95% CI 30.41-95.88), and the SROC curve AUC was 0.94 (95% CI 0.92-0.96), indicating strong discrimination and excellent overall accuracy. The authors acknowledge that heterogeneity across studies was high (I² = 100%), suggesting results varied by the specific AI platform and study context. They note that AI-assisted tools showed promise for detecting infections and could help screen populations in endemic areas, but further validation in field settings and comparison to highly sensitive reference tests is needed. Practice relevance is restrained, as the findings are specific to the included studies and contexts.

A new review of research from sub-Saharan Africa looked at how well computer-based AI tools can detect a parasitic infection called Schistosoma haematobium in urine samples. The review combined results from 10 different studies, which included over 5,500 urine samples. The AI tools were compared to traditional methods like microscopy and molecular tests.

The main finding was that the AI tools were very accurate. They correctly identified the infection about 88% of the time and correctly ruled it out about 89% of the time. The overall ability of the AI tools to tell who had the infection and who did not was excellent, with a score of 0.94 out of 1.0.

These results suggest that AI-assisted tools could be a helpful new way to screen large populations in areas where this infection is common. They might make screening faster and more accessible. However, the studies included in the review were very different from each other, which means the results might not be the same in every situation.

More research is needed to test these AI tools in real-world field settings and compare them to the most sensitive reference tests available. The tools show promise but are not yet proven to be better than traditional methods in all cases.

What this means for you:

AI tools accurately detect a parasitic infection in African studies, offering a promising new screening option.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedMay 2026

PMID42085501

View Original Abstract ↓

BACKGROUND: Urogenital schistosomiasis caused by Schistosoma haematobium remains endemic in sub-Saharan Africa. Diagnosis traditionally relies on urine microscopy to detect parasite eggs; however, its sensitivity declines in low-intensity infections. Artificial intelligence (AI)-assisted image analysis offers a promising approach to automate egg detection and enhance diagnostic accuracy, but its performance compared with standard microscopy is not well established. METHODS: We conducted a systematic review following the PRISMA guidelines and checklist. Studies evaluating AI-assisted detection of S. haematobium compared with microscopy and/or molecular reference standards, published up to August 2025, were identified through searches in PubMed/MEDLINE, HINARI, Epistemonikos, Science Direct, Google Scholar and grey literature sources. Eligible studies were selected based on pre-defined inclusion and exclusion criteria. The quality of included studies was assessed using the QUADAS-2 tool. Heterogeneity among studies was evaluated using the Cochrane Q test and I² statistic. Data was analyzed using STATA version 14.1 and Review Manager version 5.4.1. RESULTS: Ten studies (15 datasets, 5,564 urine samples) conducted in sub-Saharan Africa met the inclusion criteria. AI-assisted tools demonstrated high diagnostic accuracy. The pooled sensitivity was 88% (95%CI 83%-91%) and pooled specificity was 89% (95% CI 83%-93%). The pooled diagnostic odds ratio was 54.00 (95% CI 30.41-95.88), indicating strong discrimination between infected and uninfected cases. The SROC curve yielded an AUC of 0.94 (95% CI 0.92-0.96), reflecting excellent overall accuracy. Heterogeneity across studies was high (I² = 100%), suggesting results varied by the specific AI platform and study context. CONCLUSION: AI-assisted microscopic diagnosis of S. haematobium achieved very good in this meta-analysis. These automated tools, whether smartphone-based or bench-top systems, showed promise for detecting infections and could help screen populations in endemic areas. With further validation in field settings and comparison to highly sensitive reference tests, AI diagnostic technology may become a valuable tool to improve case detection and support schistosomiasis control and elimination efforts.

Meta-analysis finds AI tools show high accuracy for Schistosoma haematobium detection in sub-Saharan AfricaAI tools show high accuracy for detecting a parasitic infection in Africa

Study Details

Machine Learning Model Performance for Predicting Contrast Induced Nephropathy in Coronary Interventions

Machine learning models help predict kidney damage during heart procedures

Clinical research that matters. Delivered to your inbox.

Meta-analysis finds AI tools show high accuracy for Schistosoma haematobium detection in sub-Saharan AfricaAI tools show high accuracy for detecting a parasitic infection in Africa

More on Schistosomiasis

Study Details

Machine Learning Model Performance for Predicting Contrast Induced Nephropathy in Coronary Interventions

Machine learning models help predict kidney damage during heart procedures

Clinical research that matters. Delivered to your inbox.

Related in Radiology & Imaging

From Other Specialties