Major Bio-Medical Imaging Modalities
Our projects involve robust Machine Learning (ML)/Deep Learning (DL)/Artificial Intelligence (AI) techniques and we apply them to various bio-medical imaging modalities in an organ/disease-agnostic manner. We do not discriminate against any type of data/organs/diseases as we strongly believe in "Building Healthcare - Data as Capital" motto! :-)
Dermotology
Epifluorescence - Denoising, Segmentation, Clustering
Immunofluorescence - RF-HEp-2, HEp-2SegZoo
Fundoscopy - Retinal-seg
Confocal - Denoising, Deconvolution, Segmentation
IHC
FISH
Cryo-EM - Denoising
ErythroNet
MALDI-MS
ErythroNet
Brain MRI - MAC-Multiphase segmentation, MSP-Mid saggital plane, Skull stripping, Symmetry, SIMMER, Bleeds, Denoising, iSPi
Liver MRI
MREntNet
MRA
fMRI - LeaS
Mammography - Segmentation, Enhancement, Registration
DBT
Thermal imaging
Denoising
Chest X-ray (CXR) - PICCLineNet, TracNet, Analysis, LungSeg, LungBoundary, TBScreen
Cardiomegaly
Organ Segmentation
CBXIR
DEXA
Multimodal AI
Multimodal data that spans the domains of Bioinformatics and Clinical, Health, Medical Informatics can be leveraged in unified frameworks and Prasath Lab is interested in leveraging the recent multimodal AI (contrastive learning, foundational models) to solve biomedical informatics problems.
Selected publications:
P. Rajdeo, S. Deoghare, B. Iyer, S. Debsarkar, N. Chatterjee, M. Imam, B. Aronow, V. B. S. Prasath. Application of multimodal deep learning in medicine. In preparation, 2024.
H. Voleti, S. Deoghare, V. B. S. Prasath. Anonymous Submission. Submitted, 2024.Â