Nigerian AI Scientist to Roll Out Software That Cuts MRI Time By 90%

Nigerian AI researcher Mary-Brenda Akoda, founder of GenScan AI, plans to launch her AI medical imaging solution called GenMRI in Nigeria and the UK by the second half of 2027. GenMRI is designed to reduce MRI scan times by up to 90 percent without requiring hospitals to purchase new equipment. In an interview at a Lagos event, Akoda, a Google DeepMind Scholar, ex-Microsoft engineer, and Imperial College London researcher, explained that the software can transform a 20-minute scan into a mere two minutes and an hour-long scan into six minutes.

By accelerating MRI scans, hospitals can process more patients per day, shorten waiting lists, and decrease the costs associated with repeat scans and sedation, especially for vulnerable populations like the elderly, children, and infants. After training the software on over 8,800 MRI images and conducting an early assessment with a consultant radiologist in the UK, the radiologist determined that the GenMRI-enhanced images were just as accurate as conventional MRI images.

Technical evaluations revealed up to 99.7 percent similarity to traditional MRI images, with scan speeds ranging from four to ten times faster. The system was also validated on over 15,000 cardiac MRI images and 7,700 cardiac MRI images, with a blind test on 1,125 knee MRI images that had never been seen before. Akoda, driven by personal experiences with a delayed stroke diagnosis and overburdened healthcare systems, emphasized that swift diagnosis is crucial, particularly in critical conditions like stroke and cancer, where every second counts.