[Back to Homepage] [Brief CV] [Education] [Employments] [Mentoring] [Research] [Presentations] [Publications] [Softwares] [Teaching]
Date of thesis defense: 03 February 2026
Title of the thesis: Statistically Consistent and Novel Image Registration Methods for Various Transformations in Presence of Zooming
Primary Contribution of the thesis: This thesis aims to make the first stride towards widening the scope of rigid-body image registration to include zooming or scaling under the framework of jump regression analysis. The proposed solutions to various image registration problems perform well on a much wider range of situations compared to many popular state-of-the-art methods. Thus, the proposed solutions should be useful in various downstream applications, such as image comparisons, image monitoring, and so forth. Moreover, the central ideas underlying the proposed framework are adaptable and may serve as a foundation for developing a broader class of techniques in data science, including methods for nonlinear image registration.This thesis has produced the following two journal papers so far (as of December 2025), with one more journal paper expected in due course:
Current job (as of February 2026): Principal Biostatistician at Novartis Healthcare Pvt. Ltd., Hyderabad, India.
Date of thesis defense: 14 January 2025
Title of the thesis: Statistical Monitoring of Image Data under Jump Regression Framework
Primary Contribution of the thesis: This thesis leverages jump regression analysis to formulate statistical methodologies for image comparison and monitoring in real-world applications across diverse scientific disciplines. By forging a meaningful connection between image processing and statistical process control, two fields that have largely evolved in isolation, this work opens a new methodological frontier. The proposed approaches address fundamental challenges in image monitoring by unifying state-of-the-art techniques from both domains, thereby laying the groundwork for more robust, scalable, and interpretable image-based monitoring systems and advancing the field towards broader practical impact.This thesis has produced the following four journal papers:
Current job (as of February 2026): Postdoctoral Research Associate at the department of Biostatistics and Bioinformatics in Rollins School of Public Health of Emory University, USA, Postdoctoral supervisor: Prof. Amita Manatunga.
Date of submission of the thesis: 18 July 2025
Tentative title of the thesis: A Regression Tree Framework for Denoising and Monitoring of Image Data
Primary Contribution of the thesis: This thesis develops a regression tree–based framework founded on Oblique-axis Regression Trees (ORT) for estimating discontinuous regression functions in finite-dimensional spaces, and demonstrates its potential through effective image denoising. By rethinking the assumptions imposed on the underlying regression function, the resulting theoretical analysis attains a level of structural simplicity that departs from much of the existing regression tree literature. Extending the ORT framework beyond static estimation, the thesis introduces an original paradigm for monitoring drift patterns in image sequences. Although gradual temporal variations, commonly referred to as drifts, are ubiquitous in sequential imaging data, principled statistical approaches for drift monitoring remain largely unexplored. By addressing this challenge, this thesis aims to open new avenues for the monitoring of evolving image data.This thesis has produced the following two journal papers so far (as of December 2025), with one more journal paper expected in due course:
Current job (as of December 2025): Assistant Professor at Dhirubhai Ambani University, Gandhinagar, India.
[Back to Homepage] [Brief CV] [Education] [Employments] [Mentoring] [Research] [Presentations] [Publications] [Softwares] [Teaching]