ARUSHA: THE Arusha based Nelson Mandela African Institution of Science and Technology (NM-AIST), has developed a new Mobile Application, which is able to accurately detect and analyse Pneumonia from Chest X-ray images.
The technology is geared to assist doctors in hospitals in identifying and diagnosing pneumonia more easily.
The new artificial intelligence aided innovation is also able to determine the confidence level and specific areas affected by the pneumonia complications and can also suggest the necessary medical remedies and treatments.
The Vice-Chancellor of NM-AIST, Prof Maulilio Kipanyula said with the new discovery, analysing chest x-ray images of pneumonia patients will now be digitally possible as opposed to the current practice where most of the chest X-ray images were manually interpreted and analysed.
“Our role as an African Centre of excellence in science and technology, is pursuit and promotion of excellence in research and innovation. We are committed to remain relevant to the society and industry by delivering quality research and innovations as well as educating African scientists by equipping them with technical, entrepreneurial and leadership capacities to solve African problems,” said Prof Kipanyula.
He further noted that there was no impact without use, saying a research result or an innovation can generate an outcome only after its actual use and maximising the value and impact of the new knowledge generated into tangible products, services, or policy change.
He said his institution is determined to be a leader in bringing innovation into use in solving society and industry challenges, saying “In parallel, we work with public institutions and provide them with technical solutions addressing their pressing concerns”.
The Mobile Application, to be called “LungGuard” has been developed by Mr Japheth Mumo Kimeu, a second year Master’s Student pursuing Masters of Science in Embedded and Mobile Systems at NM-AIST.
“We are currently in the process of patenting before ensuring that we scale the mobile application services to most of the hospitals in the country and beyond,” said Mr Kimeu.
He said to invent the applications, he first analysed the current challenges facing medical doctors and radiographers deciphering and analysing the x-ray images manually, saying the practice is pruned to a number of challenges, including unreliable and uncertainty of the results.
“When you do analysis on the x-ray images manually, you are likely not going to quantify as to what extent someone has been affected by the pneumonia complications not even clear of a specific place affected in the lung” he said.
With support from supervisors, Professor Michael Kisangiri, Dr Judith Leo and Ms Hope Mbelwa, Mr Kimeu was able to develop the artificial intelligence aided application.
Sponsored by CENIT@EA, Mr Kimeu, whose research title was “Deep Learning Approach for Enhancement of Pneumonia Medical Imaging Analysis” said he was prompted to pursue the research after noting that pneumonia complications were accountable for about 14 percent of all maternal mortality rate in the world.
The research was also voted the best by the Artificial Intelligence Advancement International Conference in the African Context (AAIAC) which ended recently at the Arusha International Conference Centre in Arusha.
NM-AIST in Arusha is one of a network of Pan-African Institutions of Science and Technology located across Sub-Saharan Africa (SSA), a proud brainchild of the late Nelson Mandela, the former President of South Africa.
This realization was made possible through collaboration, showcasing the synergy between NM-AIST and University of Dodoma, facilitated by the Artificial Intelligence for Development Lab (AI4D) and Institute of Electrical and Electronics Engineers (IEEE), with funding from AI4D.