blog

F1000 presents an interview with Gates Foundation grantee Sasha Frade, one of the co-authors of “Malaria RDT (mRDT) interpretation accuracy by frontline health workers compared to AI in Kano State, Nigeria”. The Research Article was submitted to VeriXiv, a Verified Preprint Server built and run by F1000. Read on to learn about their research and why they chose to upload their work to VeriXiv.

How would you summarize your preprint?

Our preprint investigates the accuracy of malaria rapid diagnostic tests (mRDTs) interpreted by frontline health workers (FHWs) in Kano State, Nigeria, compared to artificial intelligence (AI) computer vision technology. We analyzed the performance of both groups in identifying malaria-positive, negative, and faint positive test results. The study found that AI algorithms performed as well as trained health workers, particularly excelling in interpreting faint positive results. This suggests that AI could significantly assist FHWs in improving diagnostic accuracy in malaria-endemic regions.

What inspired you to investigate the topic of your recent preprint?

The inspiration came from the persistent challenge of misinterpreting malaria RDT results, which can hinder effective treatment and control efforts. In high-burden settings like Nigeria, accurate testing is critical for malaria management. We wanted to explore whether AI-based tools could enhance diagnostic accuracy, particularly in low-resource settings where training and supervision may be limited. This research aims to address these gaps and offer potential solutions using technology.

What key findings are reported in your preprint and what real-world impacts do you hope they will have?

Key findings include the high accuracy of AI-based interpretations, especially with faint positive mRDTs, where AI outperformed FHWs. Overall accuracy yielded a 96.4 weighted F1 score for the AI compared to 95.3 for FHWs. AI algorithms were additionally able to accurately classify 90.2% of the 163 mRDTs that showed a faint positive line, compared to only 76.1% for the FHWs. We hope that these findings will lead to the integration of AI tools into public health systems, improving the quality of diagnostics and, ultimately, malaria treatment outcomes in endemic regions.

Why is publishing your work as a preprint important to you? And what motivated you to publish with VeriXiv?

Publishing as a preprint allows for rapid dissemination of our findings, which is particularly important in fields like public health, where timely access to research can inform policy and practice. VeriXiv provided an accessible platform with strong foundations in global health research, making it an ideal venue for sharing our work. The open access model also aligns with our goal of contributing to the wider scientific and medical communities without barriers to entry.

What are your next steps for your future research on this topic?

Our next steps involve scaling the use of AI-based RDT interpretation tools in different settings and for other infectious diseases. We plan to conduct further studies that expand beyond malaria to include multi-disease diagnostic platforms. Additionally, we aim to investigate the impact of AI on clinical decision-making and workflow efficiency, particularly in under-resourced health systems where these tools could have the most significant impact.

We want to thank Sasha for taking the time to contribute their thoughts to the blog. Read the full Research Article today on VeriXiv to explore the study and its findings in more depth.