From Brain Signals to Clinical AI: Visual Imagery Decoding, Endovascular Brain–Computer Interfaces, and EEG Biomarkers

Monday, November 2, 2026
3:00 PM – 4:00 PM
HTB 10.114

Yanagisawa Takufumi

Graduate School of Medicine, Osaka University

-

Abstract: Advances in artificial intelligence are transforming how we read and interpret signals from the human brain, opening new paths toward clinical brain–computer interfaces (BCIs) and neural biomarkers. In this talk, I will give an overview of our work at Osaka University across three areas. First, I will present our visual imagery BCI, in which intracranial recordings from patients are decoded into the latent spaces of pretrained AI models. This approach allows us to infer what a person is seeing or imagining, and to let them control these semantic representations through mental imagery. Second, I will introduce our endovascular EEG platform, a minimally invasive approach that records neural signals from within blood vessels without open brain surgery, with potential applications in communication support for patients with ALS and other neurological disorders.

Finally, I will present our deep learning approach to dementia diagnosis based on resting-state EEG. A convolutional neural network trained on data from a single hospital distinguished patients with Alzheimer’s disease, dementia with Lewy bodies, or idiopathic normal-pressure hydrocephalus from healthy individuals. We are now establishing a multicenter EEG cohort with PET imaging to develop accessible, AI-driven biomarkers for early diagnosis.

Speaker Bio: Takufumi Yanagisawa, M.D., Ph.D., is Professor of Neuroinformatics at the Graduate School of Medicine, The University of Osaka, and Founder and CEO of ivec Inc. A neurosurgeon and neuroscientist, he specializes in analyzing human neural signals, including electrocorticography (ECoG), electroencephalography (EEG), and magnetoencephalography (MEG). His research combines neural decoding and artificial intelligence to develop brain–computer interfaces for communication, neurological diagnosis, and neurofeedback therapy. He is currently leading the clinical translation of a minimally invasive endovascular EEG brain–machine interface. He received his M.D. and Ph.D. from Osaka University, after earning B.A. and M.Sc. degrees in physics from Waseda University. He is a board-certified neurosurgeon and epileptologist, has authored more than 85 peer-reviewed articles, and is an inventor on 19 patents. His honors include the Young Scientists’ Prize of the Commendation for Science and Technology from Japan’s Minister of Education, Culture, Sports, Science and Technology, and the international BCI Award