Dmitrii TodorovMany symptoms of Parkinson’s disease, despite a long history of research, are still not completely understood. In particular, there is no clear understanding on how Parkinsonian rest tremor is generated and even how it can be detected and distinguished from voluntary movements in neural recordings. In my BCI-flavour study I use multimodal (MEG and STN LFP) brain data, recorded from tremor-dominant PD patients and employ several data analysis methods (PCA, LDA, t-SNE, UMAP) on it to evaluate how distinguishable are distinct behavioral states (rest, tremor, voluntary movements). We also describe which data features (both spectral and time-domain ones) contribute most to such classification (see second file page for additional details).
Link to an outline (
audio to be listened to, while looking at the poster pdf -- same as attached file)
Google Meet: https://meet.google.com/ixh-gagf-cec