The Brain Imaging Collaboration Suite

I developed the Brain Imaging Collaboration Suite (BrICS), a web platform for analyzing MR spectroscopic data and using it for radiation therapy planning. BrICS is available for use at, and a public demo is available through Dr. Shim’s lab at BrICS has been used for multiple clinical trials at Emory University and with collaborations at the University of Miami and the Johns Hopkins University.

Gurbani et. al., The Brain Imaging Collaboration Suite (BrICS): A Cloud Platform for Integrating Whole-Brain Spectroscopic MRI into the Radiation Therapy Planning WorkflowTomography, 2019; 5(1).

Deep Learning for MR Spectroscopy Quantification

A key step in the processing of MR spectroscopic imaging (MRSI) includes the quantitation of each metabolite, typically done through fitting a model of the spectrum to the data. For high‐resolution volumetric MRSI of the brain, which may have ~10,000 spectra, significant processing time is required for spectral analysis and generation of metabolite maps. In this project, a novel unsupervised deep learning architecture that combines a convolutional neural network with a priori models of the spectrum was developed. This architecture, which we dubbed a  convolutional encoder–model decoder (CEMD), combines the strengths of adaptive and unbiased convolutional networks with models of magnetic resonance and is readily interpretable. The CEMD architecture performs accurate spectral fitting for volumetric MRSI in patients with glioblastoma, provides whole‐brain fitting in 1 min on a standard computer, and handles a variety of spectral artifacts.

Gurbani et. al., Incorporation of a spectral model in a convolutional neural network for accelerated spectral fitting, Magn Med Res, 2019; 81(5): 3346-3357.

Artifact Filtering in Spectroscopic MRI Using Machine Learning

Proton magnetic resonance spectroscopy (a.k.a. spectroscopic MRI) is a noninvasive modality capable of generating volumetric maps of in vivo tissue metabolism without the need for ionizing radiation or injected contrast agent. Magnetic resonance spectroscopic imaging has been shown to be a viable imaging modality for studying several neuropathologies. However, a key hurdle in the routine clinical adoption of spectroscopic MRI is the presence of spectral artifacts that can arise from a number of sources, possibly leading to false information. Conventional methods of addressing artifacts rely on statistical analysis of curve fits on spectroscopy, which are insufficient in removing all artifacts. We are working on developing neural networks which can operate directly on “raw” frequency-domain spectroscopy data and determine automatically which spectra have sufficient quality to be used in clinical assessment.

Gurbani et. al., A convolutional neural network to filter artifacts in spectroscopic MRI, Magn Reson Med, 2018;00:1-11.


Please see my Google Scholar profile.