Homepage of Deepak Roy Chittajallu

2012 - 2014

2007 - 2010

2004 - 2007

• Developed a method that couples marker-controlled watersheds with a hierarchical learning-based region merging algorithm for the segmentation of hundreds of nuclei from 3D Intravital fluorescence microscopy data.
• Developed a supervised hierarchical classification scheme for the automatic identification of the cell cycle state of each cell in 3D Intravital fluorescence microscopy data.
• Developed a learning-based method for quantifying the extent of DNA damage of each individual cell in 3D intravital fluorescence microscopy data.
• Developed a user-friendly GUI front end that enables a biologist to run the underlying computational framework on any dataset and view the results.
• Developed a set of software tools to procure ground truth for the validation of underlying algorithms and for generating training data for the algorithmic components based on machine learning.

• Co-developed a supervised hierarchical classification scheme for the detection of coronary calcifications in non-contrast CT data
• Co-developed a novel heart-centered coordinate system in application to the automated detection of coronary artery zones in non-contrast CT data
• Developed a graph-based method for the automatic delineation of the inner thoracic region in non-contrast CT data
• Co-developed an automatic method for the segmentation of the diaphragm in non-contrast CT data
• Developed Fuzzy-Cuts - A knowledge-driven graph based method for medical image segmentation
• Developed a Shape-driven MRF Model for the segmentation of organs in medical images
• Developed the initial set of image-derived features for the development of a new coronary artery calcium score with particular emphasis on leveraging the previously unexplored fact that the location of the calcium deposits within the coronary artery is an important factor in determining its contribution to cardiovascular risk.
• Supervised a new Ph.D. student and helped him get started on the topic

• Developed methods for automatic and semi-automatic estimation of breast volume from a 3D surface scan of the patient’s torso.
• Co-developed a novel geometric model that captures the overall shape of the breast including its key shape variations. The model is accompanied with a physics-based deformable model framework that fits the breast shape model to real data.
• Developed the UH-CARES software for computer-aided breast reconstructive surgery that includes three modules, namely:
• Automatic Non-Invasive Breast Volume Estimation
• Graphical environment that allows the surgeon to interactively create a female breast model using a novel physics based deformable female breast model
• Patient database maangement.
• Conducted Simulation studies for post-mastectomy breast reconstructive surgery using ABAQUS.

• The project involved the development of an autonomous security system that continuously monitors the activity captured by various surveillance cameras and reports concerned personnel of any suspicious activity. I was in charge of developing two modules: (i) Motion detection and activity perception; and (ii) Video capture and playback.

Cardiovascular Disease (CVD) is one of the leading causes of death, both in United States and all around the globe. One of the primary causes of CVD is coronary artery atherosclerosis, also known as coronary artery disease (CAD). Recent studies have shown that the presence of calcified coronary plaques, as detected from non-contrast computed tomography (CT) data has a significant predictive value for CAD. Consequently, several CAD risk scores have been developed based on the data collected by CT. However, in spite of a vast amount of CAD-related information available from CT, only a small fraction of it is being used in the existing risk scores. This is due to the lack of robust image analysis methods for the automated extraction of CAD-related information from non-contrast CT data. The long term goal of our research, to which this dissertation contributes, is to develop a set of computational methods for the automated extraction of CAD biomarkers from non-contrast CT data.

The specific objectives of this dissertation are: (i) To develop a general knowledge-based method for the segmentation of organs in medical images with particular emphasis on the incorporation of knowledge-driven constraints into the segmentation problem, and to apply this method to the problem of heart segmentation in non-contrast CT data; (ii) To develop a method for the delineation of the inner thoracic region in non-contrast CT data; and (iii) To develop a method for the estimation of coronary artery zones in non-contrast CT data.

The main contributions of this dissertation are: (i) development of a general knowledge-driven Markov Random Field model for image segmentation that uses prior information about appearance, location and shape to collectively constrain the solution space of the segmentation problem; (ii) development of a graph-based method for the delineation of the inner thoracic region in non-contrast CT data; and (iii) development of a learning-based method for the estimation of coronary artery zones in non-contrast CT data. The accuracy and robustness of the proposed methods is demonstrated by an extensive set of experimental results.

With the exception of skin cancer, breast cancer is the most common cancer diagnosis among women worldwide. Post-mastectomy breast reconstruction is the third most frequently performed reconstructive procedures. The continuous creation and adoption of new surgical techniques and devices have made it possible for the surgeons to reconstruct the breast in many different ways, thus making the decision-making job of the surgeon much more difficult. Current trial and error methods may lead to an unsatisfactory surgical outcome. Hence, there is an urgent need for a system that can pre-operatively simulate the surgical options, visualize potential outcomes of the surgery, and, consequently, help the surgeon and the patient in the process of surgical planning and decision making.

In this thesis, we present a set of computational tools for breast segmentation and breast shape modeling. More specifically: 1) we propose a new algorithm for automatic segmentation of the 3D data of a breast from a 3D scan of the patient's torso; and 2) we present a second generation deformable model for compact representation of the breast shape as estimated from 3D triangular mesh data of the patient's breast. These techniques can serve as the foundation of a number of applications in pre-operative planning for breast reconstruction and for breast plastic surgery in general.

• Recipient of an award for "Best overall PhD Work between 2007 and 2010" from the Department of Computer Science, University of Houston, Texas, USA
  
  
• Was selected to showcase my Ph.D. research at the CVPR Doctoral Consortium for two consecutive years 2009 and 2010.
  
  
• My poster entitled "Fuzzy-cuts: A knowledge-driven graph-based method for medical image segmentation," was among 8 of the 73 posters selected for oral presentation at the International Computer Vision Summer School (ICVSS) 2009.
  
  
• Secured 1268th rank in the Entrance Examination for Undergraduate Admission (EAMCET - 2000) in the year 2000 out of a total of 150,000 students in the State.
  
  
• Secured 496th rank in Indian Institute of Information Technology (IIIT - HYDERABAD) entrance examination in the year 2000 out of a total of 50,000 students from all over India.
  
  
• Certificate of Merit in Mathematics Olympiad - 1997-98.
  
  

• Bassam Almujahid
  Junion Ph.D. Student
  Research Topic: Design and development of a new coronary artery calcium score based on image derived features of coronary calcium deposits extracted from non-contrast CT data
  

01/2011 - 05/2011

• Neva Waynesboro
  NSF's Research Experience for Undergraduates (REU) Program
  Research Topic: Automatic segmentation of the heart in non-contrast cardiac CT data using a Markov Random Field Model with Shape Priors
  

06/2010 - 08/2010

• Paul Balanca
  Research Intern, Computational Biomedicine Lab
  Research Topic: Automatic delineation of the inner thoracic region in non-contrast CT data using a graph-based method.
  

02/2009 - 08/2009

• Barbara Morales-Quinones
  NSF's Research Experience for Undergraduates (REU) Program
  Research Topic: Automatic Detection of the Inferior Boundary of the Heart in Non-Contrast Cardiac CT Data Using Dynamic Programming
  

06/2008 - 08/2008