Shulou(Shulou.com)12/24 Report--

CTOnews.com, December 1, under the leadership of Bixiaan, a professor at the School of Information Science and Engineering of Hunan normal University, the brain science and artificial intelligence team innovated and developed an AI algorithm that can diagnose Alzheimer's disease (AD).

Technical framework of deep learning algorithm for disease classification and risk prediction. The team deeply analyzed brain images and genetic data, took them as macro and micro views of AD, and proposed a deep learning algorithm for disease classification and risk prediction, which can accurately generate brain functional network views.

A large number of experiments applied to AD show that the accuracy of multi-stage diagnosis and risk prediction is 74.2% and 84.5% respectively, which is 10 percentage points higher than the existing advanced diagnosis methods.

Photo Source: Qingdao Emergency Center the research results were published on IEEE TPAMI on November 6. Bisha'an is the first author and newsletter author. Liu Tianming, a professor at the University of Georgia, USA, is the co-author. Li Xiang, an assistant professor at Harvard Medical School and Massachusetts General Hospital, participated in the research. Hunan normal University is the first paper unit and communication unit.

In the biomedical field, AD is defined as a brain degenerative disease caused by a variety of complex factors. At present, researchers have developed a variety of detection techniques to help clinical workers understand the disease, such as magnetic resonance imaging, PET and other medical imaging methods.

Bishop explained:

In the Alzheimer's Disease Neuroimaging Program (ADNI) database, we extracted data from 197patients with early mild cognitive impairment, 203patients with advanced mild cognitive impairment and 233patients with Alzheimer's disease, and achieved gratifying results.

At present, it is possible to draw blood for genome-wide tests to determine the early stages of Alzheimer's disease and risk, but it is not known how genes characterize changes in brain function, and brain changes are invisible. Based on the team algorithm, individual gene data can be directly mapped to the corresponding brain network.

CTOnews.com attached the reference address of the paper: X.-a. Bi et al., "Structure Mapping Generative Adversarial Network for Multi-view Information Mapping Pattern Mining," in IEEE Transactions on Pattern Analysis and Machine Intelligence, doi: 10.1109/TPAMI.2023.3330795.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.