Joyoni Deydey

Assistant Professor

Ph.D. in Electrical and Computer Engineering, 1999 - Carnegie Mellon University

Louisiana State University
Department of Physics & Astronomy
459-A Nicholson Hall, Tower Dr.
Baton Rouge, LA 70803-4001
(225) 578-4289-Office
(225) 578-9048-Lab
deyj@lsu.edu

Teaching

Fall:         MedP 7111: Advanced Medical Imaging Physics

Spring:    MedP 4111: Introduction to Medical Imaging

Spring:    (Co-Instructor): Radiological Physics for Residents (LSU Health Sc)

Research

Research Interest in Medical Imaging, Image Processing and Deep Learning: Imaging system design and optimization reducing patient dose, imaging time; Image reconstruction with physical modeling; Deep-learning for oncological prediction and staging; Mathematical (pde) models of tumor growth and treatment; Segmentation and registration. My research focuses on new systems and algorithms designed to help large patient populations with new imaging advances: for example, higher resolution and sensitivity systems for lowering the dose to patients, better quality diagnostic images, more efficient acquisition etc.; pathological quantification algorithms; predictions for oncology; correction of motion artifacts for better diagnosis.

Current Projects Include

System Design and Optimization

  • Optimizing a Novel Gamma Camera design for Cardiac SPECT: We are investigating a high-sensitive and/or high-resolution gamma-camera design with a system of Ellipsoid detectors with pinhole collimation for Cardiac SPECT, which is an important modality for assessing myocardial perfusion with millions of patients undergoing nuclear cardiology scans per year. In course of our research, we have built a comprehensive multi-pinhole system simulator and iterative reconstruction.

           Example Relevant Publications/Patents/Disclosures

          J. Dey and S.J. Glick, "SPECT Camera Design", Patent No., US 8,519,351 B2, Aug 27, 2013

          J. Dey, “Improvement of Performance of Cardiac SPECT Camera using Curved Detectors With Pinholes”, IEEE Trans. Nuclear Science, vol.59, no.2,pp.334-347, April 2012

  • Phase-Contrast X-ray not only provides attenuation images provided by conventional CT but also scatter and phase images, affording higher soft-tissue contrast in images compared to conventional CT. We are investigating a novel single grating far-field system to bring Phase Contrast X-ray a step closer to the clinic. In collaboration with Professor Les Butler (Dept of Chemistry), Dr. Kyungmin Ham (CAMD) and Dr. Jonathan Dowling (Dept of Physics).

    Example Relevant Publications/Patents/Disclosures

    J. Xu, J. Dey, N. Bhusal, et al., “Two-dimensional Linear Quadratic Grating for X-ray Interferometry”, presented SPIE Medical Imaging, 2018  (oral)

    J. Dey, J. Xu, N. Bhusal, , K. Ham, , V. Singh, “A Novel Phase Contrast System”,  presented, IEEE NSS-MIC, Oct 2017 (oral)

           J. Dey, et. al., "Phase Contrast X-ray Interferometry," Provisional Patent July 24, 2017

Deep-learning and Radiomic Methods for Oncological Applications

  • Convolution neural networks for lung-cancer prediction for large patient datasets (student Ivan Hidrovo senior thesis, April 2018).
  • Staging preclinical SPECT/CT data using support vector machine using key radiomic features (student Megan Chesal honors thesis, April 2018).

Building Bioanalytical Tools using Mathematical Models

  • Tumor progress and disease treatment monitoring by extracting biophysical-model-parameters from images: Oncology Imaging is performed using modalities including CT, MRI, FDG-PET, SPECT etc. Applying realistic mathematical models to serial-images of tumors extracts biologically relevant information from images, such as cell-motility, growth-rate etc. We are investigating a new mathematical Finite Element tumor-model where effects of the necrotic core are considered in addition to cell-motility, growth, apoptosis and migration. We also acquired 6 preclinical serial SPECT/CT datasets and fitted an existing ode compartmental volume model. In collaboration with J. M. Mathis, PhD, (Dept of Comparative Biomedical Sciences, School of Vet Medicine, LSU) and S. W. Walker, PhD (Dept of Mathematics, LSU).

          Example Relevant Publications/Patents/Disclosures

           I. Hidrovo, J. Dey, M. E. Chesal, D. Shumilov, N. Bhusal and J. M. Mathis, "Experimental Method and Statistical Analysis to Fit Tumor Growth Model Using SPECT/CT Imaging: A                  Preclinical Study", Quant Imaging in Med and Surg,  vol. 7, no. 3, pp. 299-309, June 2017,doi: 10.21037/qims.2017.06.05
 
          J. Dey, S. W. Walker, J. M. Mathis, D. Shumilov, K. M. Kirby and Y. Luo, "Modeling and analysis of a physical tumor model including the effects of necrotic core," in Proc 2015 IEEE               Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), San Diego, CA, 2015, pp. 1-4.
 
  • Fat-mass and Lean-Mass Growth Models: Mathematical modeling of fat-mass and lean-mass growth upon overfeeding different diets (collaboration with S. Heymsfield, MD, and L. M. Redman, PhD, from Pennington Biomedical Research Center).

    Example Relevant Publications/Patents/Disclosures

    D. Shumilov, S. B. Heymsfield, Leanne M. Redman, Steven R. Smith, George A. Bray, K. Kalluri, J. Dey, “New Compartment Model Analysis of Lean-Mass and Fat-Mass Growth with Overfeeding”, Nutrition, vol. 32, no. 5, pp. 590-600, May 2016

Patents

          J. Dey and S.J. Glick, "SPECT Camera Design", Patent No., US 8,519,351 B2, Aug 27, 2013

Students

Ph.D.

  • Jingzhu Xu

MSc

  • Hanif Soysal
  • Narayan Bhusal

Undergraduate

  • Ivan Hidrovo (REU, Summer 2016, Summer 2017, Acad year 2017-2018)
  • Megan E. Chesal (Spring 2016, Acad year 2017-2018)
  • David Sanchez (REU, Summer 2017)

 

Back to Top