How Tobi Oyekanmi is Using AI in Powering New Discoveries in Health, Environmental Safety

By Fadekemi Ajakaiye

In an era where technology increasingly shapes healthcare and environmental monitoring, Tobi Titus Oyekanmi, a researcher at New Mexico Highlands University, has been pioneering the use of artificial intelligence (AI) to tackle pressing global challenges. His research explored how AI can assist doctors in diagnosing pneumonia from X-rays and how machine learning models can be applied to monitor background radiation for environmental safety.

Both groundbreaking studies were accepted and presented at the 44th Annual Conference of the Nigerian Institute of Physics (NIP), held at Lagos State University of Science and Technology (LASUSTECH), Ikorodu, from May 8–12, 2023. The presentations marked a defining moment in Oyekanmi’s early research career, bridging computer science with applied physics and public health innovation.

In his collaborative study titled “Detection and Interpretation of X-Ray Scans for Pneumonia Using Convolutional Neural Networks,” Oyekanmi and his colleagues developed a deep learning model capable of identifying pneumonia from chest X-rays. The research focused on leveraging Convolutional Neural Networks (CNNs) to automate diagnosis and support clinicians, particularly in regions where access to radiological expertise remains limited.

The team trained the model on 5,216 pediatric X-ray images, achieving 76% classification accuracy in distinguishing between normal and pneumonia cases. Implemented using Python and TensorFlow, the system employed multiple CNN layers to extract and learn critical visual patterns from medical images.
“Our vision was to demonstrate that deep learning can act as an assistive diagnostic tool,” Oyekanmi explained. “Even at this early stage, the results showed that AI could offer valuable second opinions in healthcare diagnostics, especially in low-resource settings.”

Building on his passion for the intersection of computing and physical science, Oyekanmi also co-authored a study focused on the application of machine learning to background radiation monitoring. The research explored how predictive algorithms can help interpret environmental radiation data, improving the efficiency of safety assessments and early warning systems.

Using statistical and learning-based models, the study analyzed variations in naturally occurring radiation levels to detect anomalies and predict safe exposure thresholds. These insights could inform radiation safety regulations, industrial monitoring, and environmental protection initiatives.

“Our goal was to show that machine learning isn’t just for text or images, it can play a role in environmental physics and public health,” Oyekanmi said. “By applying these models, we can create faster, smarter, and more adaptive systems for radiation monitoring.”

Oyekanmi’s dual contributions to AI in healthcare and environmental monitoring reflect a growing movement among Nigerian scientists who are integrating artificial intelligence into traditional scientific disciplines. The studies presented at the NIP conference demonstrated how cross-disciplinary research can address both medical and environmental challenges bridging the gap between computing, physics, and social impact.

By combining computational modelling with domain-specific expertise, researchers like Oyekanmi are positioning Nigeria as a rising hub for applied AI innovation in Africa. His work highlights how collaborative efforts between computer scientists and physicists can produce solutions that have both scientific and societal relevance.

While both studies achieved promising results, Oyekanmi acknowledges the challenges of data availability and computational limitations in conducting high-quality research in developing environments. However, he remains optimistic about the future.

“Every model begins as a proof of concept,” he noted. “The key is to keep improving accuracy, validate models in real-world conditions, and ensure that AI systems remain transparent and accessible to the communities they’re built to serve.” Looking ahead, Oyekanmi envisions a more integrated approach to AI where medical and environmental models share data and methodologies to build a unified framework for improving public health and ecological safety.