Kampala, Uganda: Victoria University Researcher Dr. Lillian Tamale has broken new ground with a landmark study demonstrating that simpler artificial intelligence models can transform tuberculosis (TB) detection in low-resource settings across Africa.
Published in the July 2025 edition of JMIRx Med, the study titled “Improving Tuberculosis Detection in Chest X-Ray Images Through Transfer Learning and Deep Learning” revealed that the lightweight VGG16 convolutional neural network (CNN) significantly outperforms more complex architectures like ResNet152 and Inception-ResNet-V2 in classifying TB from chest X-ray images.
Dr. Tamale, who heads the Department of Computing and Information Science at Victoria University, led the collaborative research alongside Mr. Alex Mirugwe of Makerere University and Dr. Juwa Nyirenda from the University of Cape Town.
The trio assessed six CNN models: VGG16, VGG19, ResNet50, ResNet101, ResNet152, and Inception-ResNet-V2, using a dataset of 4,200 chest X-ray images, including 700 TB-positive and 3,500 normal cases.
From the study, VGG16 emerged as the best performer, achieving an impressive 99.4% classification accuracy alongside high precision, recall, F1-score, and computational efficiency.
“This is a game-changer for AI in public health,” Dr. Tamale said. “With VGG16, even remote or under-resourced clinics can adopt AI-powered diagnostics without needing expensive computing infrastructure. It’s a major step toward equitable healthcare access.”
Among the study’s most surprising findings was that data augmentation techniques, like flipping and rotating images, did not improve the model’s performance, challenging established practices in medical AI training.
According to the World Health Organization, TB remains one of the world’s deadliest infectious diseases, with over 10 million new infections and 1.5 million deaths each year. Many clinics in sub-Saharan Africa still lack affordable, reliable screening tools, a gap this research aims to fill.
Victoria University praised Dr. Tamale’s achievement in a statement on Friday morning, saying: “Congratulations Dr. Lillian Tamale on your successful academic publication on how to detect TB using AI. This reaffirms the University’s commitment to the use of Artificial Intelligence and modern technology.”
Dr Tamale’s study is available in JMIRx Med and has also been published in preprint form on medRxiv to encourage wider access and replication.
Victoria University’s leadership in this innovation highlights the emerging role of African universities in developing transformative digital health solutions. It also reinforces the institution’s growing reputation as a hub of cutting-edge AI, not only in Uganda but on the continent.
If you would like your article/opinion to be published on Uganda’s most authoritative news platform, send your submission on: [email protected]. You can also follow DailyExpress on WhatsApp and on Twitter (X) for realtime updates.
