The [IEEE Long Island (LI) Signal Processing Society (SPS)](https://ieee.li/society-chapters/signal-processing-society-sp/) in collaboration with [North Jersey Social Implications of Technology Society](https://r1.ieee.org/northjersey/chapter/ssit/) presents the following Technical Lecture:

Accurate image segmentation holds significance for vital clinical applications such as diagnosis and surgery planning. While deep neural networks have excelled in achieving superior segmentation outcomes via fully supervised learning, their reliance on substantial annotated training data is a challenge. Procuring extensive labeled datasets for medical images is labor-intensive and costly due to the need for clinical expertise in annotations. Thus, an opportunity for improvement is evident. Hence, the critical need to devise strategies for attaining medical images with scant annotations while harnessing untapped potential within unlabeled data during training. We harness the power of self-supervised representation learning and semi-supervised learning in this regard and perform extensive experiments on images from multiple modalities: Computer Tomograhpy (CT) scan, Magnetic Resonance Imaging (MRI) scan, Histopathology studies, etc. Our recent research showcases that even with minimal annotations estimate of x<10%, we achieve comparable or superior performance compared to fully supervised approaches.

Speaker(s): Mr. Hritam Basak, Student,

Agenda:
Technical support set-up: 5:30pm EST
Introductions 6pm-6:05pm EST
Technical Lecture: 6:05pm-6:50pm EST
Q&A: 6:50pm-7pm EST

Virtual: https://events.vtools.ieee.org/m/369575