QNRF-Funded Research Develops Automated Embedded Prototype Colorectal Tumor Detection System

Research outcomes will enhance the speed and accuracy of cancer diagnosis through automated methods

Cancer is the second leading cause of death all over the world – accounting for one in six deaths in 2018. In Qatar, colorectal cancer is the third most commonly diagnosed cancer and a leading cause of cancer-related deaths, according to the Qatar National Cancer Registry. While medical research has provided treatment approaches that can increase life expectancy to some extent, it is evident that timely quality screening for early treatment is critical.
With rapid technological advancements in the field of image processing and machine learning, histopathologic analysis, which is the common mode of cancer diagnosis, can now be automated through the application of image analysis and pattern recognition techniques on the biopsy images, while a computer-aided diagnostic system can detect and classify the various tumor cells and ensure a reliable and rapid screening procedure.
QNRF is supporting studies to develop technologies to provide an easier and faster approach for cancer detection and identification. A research project titled, "Automated Classification and Diagnosis of Tissue Patterns in Colorectal Tumors Using Non-Visible Multispectral Imagery" received funding under the sixth cycle of QNRF’s National Priorities Research Program under project ID (NPRP6-249-1-053l).
Led by Prof. Sumaya Al-Maadeed, Qatar University (QU), the team included Dr. Rafif Al Saady, Al-Ahli Hospital (Qatar), Prof. Ahmed Bouridane, Northumbria University (UK), and Prof. Mohamad Sawan, Polytechnique Montreal (Canada).
The team investigated several approaches for cancer detection and identification on biopsy images, toward understanding the biochemistry of colorectal cancer data and its subsequent diagnosis and prognosis.
This project investigated the deployment of multispectral imaging to exploit the richer texture information content at different spectral bands and develop a prototype system. The approach was proven to be useful for understanding the biochemistry of colorectal cancer data.
The team implemented a completely automated and embedded system for the classification of colorectal tumors, which was carried out by the group at QU and Al-Ahli Hospital in collaboration with the teams in the UK and Canada. The result was a multispectral image database for colorectal tumor biopsies.
Two patents have recently been filed for the project. The first is a real-time colorectal tumor detection device which could process multispectral images and deliver the diagnostic results. The second is a new miniature endoscopic capsule capable of detecting flatulent lesions of colorectal cancer, based on fluorescence imaging. Given the high incidence of colorectal cancer in Qatar, the computer-aided diagnosis system will help improve the health care system in public and private hospitals by enabling the early detection of colon cancer and allowing for successful treatment.
Among its other positive aspects, the research allowed QU to develop and maintain a working research laboratory dealing with the computational aspects of medical imaging, which could encourage students to take up research-based degrees. The collaboration with Al-Ahli Hospital will ensure that the technology can be assessed independently by pathologists and be deployed to solve remote diagnostic/treatment cases in Qatar and worldwide.
The team is hoping to extend the research to the study of other types of cancers.