QNRF's GSRA scholar develops an innovative solution to help treat hepatic diseases
Liver-related diseases reap the lives of around two million people around the globe every year. Half of these deaths are related to complications caused by liver cirrhosis, and the other half are due to hepatitis and hepatocellular carcinoma (HCC). Moreover, the liver is also a hub for metastasis – the spread of cancer cells – originating from adjacent organs such as the pancreas and stomach.
Therefore, it is crucial to develop innovative technologies and methods which can help doctors further understand the effect of these diseases on the liver and how they can be prevented and treated more effectively. 3D modeling is one of these technologies which can accurately replicate the anatomy and pathology of organs. It can help doctors treat various diseases, teach medical students more effectively, and educate the patients better about their bodies.
A QNRF Graduate Student Research Award recipient, Ayman Al-Kababji from Qatar University, has contributed to this increasingly important field through his project titled, “ Diagnosing Liver’s Lesions for Medical Analysis ” (GSRA6-2-0521-19034).
Ayman started the project by surveying the state-of-the-art studies and finding the most suitable and comprehensive dataset for liver tissue segmentation. He also obtained access to a Raad2 supercomputer and applied machine learning (ML) and artificial intelligence (AI) algorithms to produce the liver tissues segmentation present the liver in a 3D format.
Figure 1: System Model
Lastly, a clinical real-time desktop application, named AI Radiologist, was created to provide all the steps of image pre-processing, tissues segmentation, and 3D interpolation for clinicians to make it easily deployable in the field. If used in a clinical environment, this solution can help surgeons pre-plan well for hepatectomy procedure by visualizing the organ, even printing it in critical circumstances, consequently increasing the operation success rate, and in the end, the patient’s chances of survival and quality of life.
Figure 2: Top view of our segmentation and 3D interpolation of a record from an online dataset.
According to Ayman, his research project reflects the research priority areas outlined in the Qatar National Research Strategy. The AI Radiologist automatically delineates the liver, tumors, and vessels in real-time (in a matter of minutes), providing an accurate 3D model of the liver tissues to surgeons for various clinical reasons, such as hepatectomy planning and adaptive radiation therapy planning.
He said: “Thank God, we have accomplished most of what we planned to do. Our project has successfully applied state-of-the-art AI algorithms to tackle a serious problem to help save people. We have developed a convenient AI Radiologist application, catering to the clinicians’ needs who might use it in the future. Moreover, through this project, we are gaining professional experience dealing with global businesses to 3D print the results of our research to showcase and implement its outcomes.”
Acknowledging the support of his supervisors, he said: “I am grateful to Prof. Faycal Bensaali, my main supervisor, for taking this challenging project with me and providing guidance every step of the way. I also would like to thank my co-supervisor, Dr. Sarada Prasad Dakua from Hamad Medical Corporation (HMC), who aided us with his expertise in the field. He is currently helping us acquire a budget to 3D print a ground-truth liver segmentation and our algorithms’ segmentation. We might also try to deploy AI Radiologist in a clinical scenario in HMC to observe its efficacy.”