UM PhD student receives Lotfi Zadeh Best Paper Award for medical biometrics systems on accurate diabetes diagnosis
Shu Ting, a PhD student from the Department of Computer and Information, Faculty of Science and Technology (FST), University of Macau (UM), received the Lotfi Zadeh Best Paper Award for her medical biometrics systems on accurate diabetes diagnosis, at the 17th International Conference on Machine Learning and Cybernetics (ICMLC) and the 15th International Conference on Wavelet Analysis and Pattern Recognition (ICWAPR). The award is presented to the best overall paper from both conferences. Shu’s supervisor is FST Assistant Professor Zhang Yibo and the paper was co-authored by FST Chair Professor Tang Yuan Yan .
The paper, titled ‘Computer-Assisted Non-Invasive Diabetes Mellitus Detection System via Facial Block Analysis’, proposes a computer-assisted non-invasive system that can help medical professionals detect diabetes mellitus accurately and in real time. The underlying technology behind this system is medical biometrics, which involves image processing and pattern recognition. According to Dr Zhang, the team first captured a facial image using the device and then extracted four blocks located on the face. Texture features would be extracted from each block forming a feature vector. Finally, classification is performed to determine whether or not the individual has diabetes. The proposed system operated by a technician can be set up at hospitals or health centres, allowing more people to check their health status. The proposed system takes 9.9 miliseconds to show the results and can achieve an average accuracy of 97.5 percent. Shu appreciates the guidance from her supervisor Dr. Zhang, from whom she learned how to conduct research and work in a team. Shu also praised the inspiring research environment at UM, saying this has encouraged her to pursue a PhD degree in computer sciences to take her career to the next level.
Both held in Chengdu, China, this year’s ICMLC and ICWAPR attracted over 160 experts and researchers from around the world. The ICWAPR aims to provide a platform for researchers in wavelet analysis and pattern recognition to explore the synergy by combining these closely related areas of study, while the ICMLC is a leading international conference in the field of machine learning and cybernetics.