Dr Mustapha Oloko-Oba graduated from UKZN’s School of Mathematics, Statistics and Computer Science with a PhD in Computer Science. He was supervised by Professor Serestina Viriri.
Of the firm belief that “travelling is part of education”, he relocated from Nigeria to register at UKZN. The University’s ranking as a leading university in Africa and its innovative research, curricula, dynamic teaching and learning, and state-of-the-art laboratories influenced his choice of institution.
Oloko-Oba developed a computer-aided detection system to diagnose Tuberculosis (TB) from a chest radiograph (CXR). He employed state-of-the-art deep learning models through ensemble learning to detect and distinguish infected CXRs from healthy ones.
The model was trained on the Shenzhen dataset and validated on the Montgomery dataset to improve accuracy and generalisation on new (unseen) datasets. The EfficientNet model’s performance was found to be comparable with state-of-the-art techniques and it outperformed existing TB classification systems.
Oloko-Oba noted that TB is among the leading causes of death in developing countries. According to the World Health Organization, South-East Asia and Africa account for about 69% of global cases. Tuberculosis causes economic distress and exacerbates poverty and vulnerability. The automatic detection system will assist with early diagnosis, correct misdiagnosis and address the shortage of skilled radiologists. Early detection increases the chances of a cure and millions of deaths could be averted.
The system will also assist healthcare centres, especially in TB-burdened regions, to detect the disease early and prescribe appropriate treatment.
Oloko-Oba is currently involved in online sentiment analysis. His future plans include ongoing research and publications, and collaborating with industry, research institutes and individuals to create new ideas and design innovations for the benefit of humanity.
His supervisor Professor Viriri noted that Oloko-Oba was a dedicated researcher with innovative ideas and an expert in machine learning, especially applied to the medical image analysis niche.
Oloko-Oba described his PhD studies as a wonderful experience, enhanced by the warm reception he received from staff and students in the School of Mathematics, Statistics and Computer Science. He said that he was privileged to learn about South Africa’s diverse cultures and even picked up some isiZulu.
He devotes his spare time to advocating for equity, diversity and inclusion.
Words: Leena Rajpal
Photograph: Sandile Ndlovu
