shiva khoshnoud

* Digital biomarker pipeline — Built end-to-end analysis pipelines that turned messy, real-world EEG/ECG/EMG into validated digital biomarkers: preprocessing, quality control, and feature extraction, with predictive/ML modeling where the question called for it (e.g. the Parkinson's DBS study). Python (NumPy, pandas, scikit-learn, PyTorch) and MATLAB. * Parkinson's DBS outcome study (first-author, npj Digital Medicine) — Led the longitudinal analysis and built the predictive models linking DBS parameters and

Multimodal biosignal research on brain–body interaction and the psychophysiology of time perception — from raw EEG/ECG/respiratory signals to peer-reviewed findings. * Time-perception psychophysiology — Analysed multimodal physiological time-series (EEG, ECG, respiratory) to quantify brain–heart interaction and the psychophysiology of time perception during experimental tasks, turning subjective experience into quantitative markers derived from physiology. * Statistical modeling & reproducible workflows

* Biosignal classification (ML) — Developed signal-processing and machine-learning classifiers for biomedical signals: EEG-based ADHD characterisation and arrhythmia detection from ECG (linear predictive coefficients, probabilistic neural networks). * Peer-reviewed output — Published multiple peer-reviewed studies on biosignal classification and feature extraction