Deep Learning

To automatically extract the valve-related motion parameters, the proposed AI-based system was developed and analysed on a large dataset. We investigated the robustness and feasibility of the system extensively on MV and TV related motion parameters. The system achieved human-level accuracy and can improve the workflow efficiency, automation and standardization of valve-related acquisitions or analyses.

Small Android demo application which classifies singing vs speaking vs silence of the last 2 seconds based on microphone sensor data.

The analysis of mitral valve motion is known to be relevant in the diagnosis of cardiac dysfunction. Dynamic motion parameters can be …

Phase-contrast (PC) MRI is used to evaluate blood hemodynamics; however, it can be time consuming to process PC-MR data. In this work, …

Research internship project where I conducted literature-review and compare state-of-the-art analysis used in EEG signal processing. Implemented DL pipeline to decode grasping action from EEG

Automatic derived parameter for valve motion assessment and interactive visualization dashboard for cluster analysis and anomaly detec.

M.Sc. Thesis carried out in collaboration with Siemens Healthineers. Research state-of-the-art literature on cardiac MRI slice tracking and medical landmark detection. Develop a robust novel valve tracking 3D-CNN algorithm based on heatmap regression

The project aims to combine multi-sequence CMR data to classify the myocardial pathology. It willl be segmented into into normal, infarcted and edema regions, which is essential in the diagnosis and treatment management for patients suffering from myocardial infarction (MI)