Assoc.Prof. Dr.Sc. Vu Hoang Linh (left), Rector of VNU University of Science, handing over the project product to the representative of Med-Aid Company (USA).
Assoc.Prof. Dr. Nguyen Thi Hong Minh - Project team representative, presenting the project results to the US partners.
The aim of the project is to apply artificial intelligence, especially machine learning methods, to automatically zoning human body parts through tomography images. The members of the research team used modern algorithms to process medical images, localizing 12 parts of the human body in 3 areas: the head (left eye, right eye, brain), the chest (left/ right lung, spinal cord, heart) and the abdomen (left/ right top femur, bladder, rectum, prostate). The product has an accuracy of 92% for the head and chest areas, 83% for the abdomen, while the accuracy rate required by Med-Aid, USA is 80%.
Mr. John Cong Nguyen, Chairman of Med-Aid Company, said: “This is the first project between a Vietnamese university and Med-Aid (USA). The results of the project are beneficial for the field of cancer radiation therapy, helping to reduce treatment time, and more importantly, to increase and ensure treatment quality for cancer patients”.
The tomography images of a medical diagnosis are always a series of slices. Normally, it takes a doctor 2-3 hours on average to process a patient’s image, but thanks to the application of artificial intelligence, doctors only spend a few minutes getting the intended results with high accuracy zoned images; and therefore, doctors can easily identify tumors to treat and avoid damage to other body parts”.
VNU University of Science has a large network of cooperation with domestic and foreign businesses and experts in training and scientific research. In addition to teaching and research activities, the lecturers also participate in science and technology consultancy to many large domestic and global companies.