Invitation to to the May Edition of NAMP webinars: Machine Learning-Based Computational Magnetic Resonance Strategies for Rapid Diagnosis of Coronavirus Disease 2019 (COVID-2019) Using Magnetic Resonance Fingerprinting Data


In commemoration of the International Medical Physics Week (IMPW), NAMP would be hosting a special webinar. Please find more details below.

Organizer: Professor O.B Awojoyogbe

Speaker: Dr. Michael Dada   

Moderator: Dr. Iyobosa Uwadiae   

Dr. O.M. Dada, Department of Physics, Federal University of Technology Minna, Niger State, Nigeria is an international leading expert of Computational Magnetic Resonance Imaging for medical physics applications, working in the field since 2011.
Since 2018, he leads the Artificial Intelligence for medical applications in the MRI research group, Department of Physics, Federal University of Technology Minna, Nigeria. O.M. Dada has extensive expertise in teaching computational MRI at various levels and organized many computational short courses addressed to the use of this MRI code in medical physics.

O.M Dada has actively participated in several international workshops and conference sessions dedicated to computational MRI codes applied to medical physics. He is an Editor of African Journal of Medical Physics.


COVID-19 a contagious disease caused by Severe Acute Respiratory Syndrome Corona virus 2 (SARS-CoV-2), as declared by WHO is a global pandemic. This virus has continued to be a serious threat to the world. During imaging, medical doctors face high risk of infection as they come in direct or indirect contact with the infected patients. Due to the ability of the virus to survive on surface, most of the times the MRI scanner gets contaminated. Preparing the patient for the imaging increases high risk of the spread of the virus and it is important to note that constant decontamination of imaging equipment cannot be fully relied on. The goal is to develop a machine-learning based computational magnetic resonance (MR) models for rapid diagnosis of Corona virus 2019 (COVID-19) by simulating clinically-consistent magnetic resonance fingerprinting data for computation of MR Dataset. We use different models to train our magnetic resonance finger printing data and develop a high accuracy deep learning algorithm with the best performing model for the diagnosis and monitoring of COVID-19. The trained model can be deployed as an APK application making it available as an android application for personal diagnosis. Special application software/libraries such as Jupyter notebook, python 3, numpy, pandas, seaborns, Streamlit and Heroku are implored for building and deployment of the model. From the analytical solution of the time independent Bloch NMR flow equation obtained in terms of Bessel functions, magnetic resonance transverse magnetization was obtained coupled with other clinical data to build the machine learning model for early diagnosis and monitoring of coronavirus 2019 (COVID-19). A GUI was built with the model of best performance using “streamlit” library and deployed to a web page hosted by Heroku cloud applications. Generally, virus is known to have a high tendency of mutating, the model can easily be updated as soon as changes are noticed, for a better and reliable diagnosis of virus related diseases.

Leave a Reply