student profile: Mr Raquib-ul Alam


Map

Thesis work

Thesis title: Automatic Artifact Rejection with Limited Samples in TMS-EEG with Deep Learning

Supervisors: Alistair MCEWAN , Steve VUCIC , Luping ZHOU , Boris GUENNEWIG

Thesis abstract:

�p�With the recent advancements of computational power in computers, scope of deep learning (DL) algorithms is expanding rapidly in medical image and signal processing studies. Artificial neural networks (ANN), an architecture of DL, have numerous applications in classification and segmentation of medical modalities. For example, diabetic retinopathy screening, skin lesion classification, and brain tumor segmentation are performed using ANN. DL is just beginning to pave its way into the world of neuroscience and it can potentially be used as a powerful processing and analyzer tool in the field of transcranial magnetic stimulation (TMS) electroencephalogram (EEG) recording. TMS is a non-invasive technique that is used in research, diagnostic and therapeutic applications in neuroscience and clinical neurology/psychiatry. The more recent integration of TMS with EEG recordings has unfolded tremendous potential but is fought with technical challenges such as the need for averaging across a high number of samples, and a range of physiological and technical artefacts. Since DL has performed exceptionally in classifying temporal data and rejecting noise in recent years, it may feasibly be employed to obtain noise-free TMS-EEG readings with potentially fewer stimulations.�/p�

Note: This profile is for a student at the University of Sydney. Views presented here are not necessarily those of the University.