Rajaram Lab

Understanding Tissue Organization VIA MAchine Learning



We are driven by the belief that the spatial organization of tissue provides a powerful  window into cell-cell interactions, a crucial component of disease progression and response. Yet, technical challenges have left this valuable source of information untapped: human comprehension can be overwhelmed by the complexity and size of individual tissue samples while purely automated approaches are hindered by the relative scarcity of independent tissue samples. By building approaches that leverage recent advances in machine learning within the framework established by biology, we hope to establish tissue organization itself as a predictive and investigative biomarker.



Our research lies at the intersection of cancer biology machine learning and pathology. We develop approaches to:

1. Incorporate Biological Priors

From a machine learning perspective, a fundamental and unique challenge is the massive (gigapixel) size of individual tissue samples relative to the smaller number (hundreds to low thousands) of independent patients. To overcome this, we develop machine machine learning approaches that incorporate biological knowledge accumulated over decades of experience with the unparalleled pattern extraction ability of deep learning neural networks.

2. Generate Robust Tissue Profiles

Tumor heterogeneity is a primary challenge in diagnosing and treating cancer. Practical constraints allow only a small fraction of a tumor to be profiled. How can we ensure we do not miss the part of the tumor that is driving tumor behavior? We combine spatial statistics with non-invasive modalities (e.g. radiology), to develop smart sampling strategies to ensure that we reliably capture the heterogeneity of a tumor.

3. Interpret Tissue Architecture

The biological origins of tissue phenotypes apart from nuclear morphology are poorly understood and are largely unused in pathology. We relate our tissue profiles to other assays (such as genetics or expression) to gain a better understanding of the underlying biology and as a first step towards generating models of progression and response.


Lab Members


Satwik Rajaram

Assistant Professor

Satwik's formal training is in theoretical physics. His switch to biology occured during his Ph.D (at the University of Illinois at Urbana Champaign) as he developed statistical-physics-inspired approaches to help humans comprehend high-dimensional biological data.

Satwik's post-doctoral research in the joint labs of Drs. Altschuler and Wu (at UTSW and then at UCSF) explored the implications of cellular heterogeneity, primarily through quantitative microscopy. He developed machine learning approaches that automatically identified relevant image phenotypes - sometimes invisible to the human eye - and accurately profiled heterogeneous cellular populations. His work also explored the associated experimental design questions, such as how few cells are needed to represent a population's heterogeneity reliably.


Anthony Vega

Postdoctoral Researcher

Tony is a native Texan who received his bachelor’s degree in Physics at St. Mary’s University and his PhD in Molecular Biophysics from UT Southwestern. In his graduate work, Tony worked on developing various computer vision tools to characterize protein interactions and organization at the cell membrane at the single-molecule level.

In a joint post-doctoral position with the Diamond lab, Tony’s current research focuses on using machine learning to elucidate the pathological features of neurodegenerative diseases. Specifically, he is interested in deciphering the relationship of pathological features at the tissue level and those at the biochemical level, and importantly how this relationship may manifest in mixed-disease pathologies.


Hafez Eslami Manoochehri

Graduate Student

Hafez received his M.S. degree in Computer Science from The University of Texas at Dallas, in 2015, where he is currently pursuing a Ph.D. degree. Since 2018, he has been with the Bioinformatics Department, UT Southwestern, as a research collaborator.

His research interests include the image analysis and machine learning in biological applications. He is interested in finding the correlation between genetics and tissue morphology. In particular, his research is focused on analyzing spatial clonal evolution models of tumor heterogeneity in histopathology images using deep learning.


Paul Acosta

Graduate Student

Paul received his bachelor’s degree in Biomedical Engineering at the University of Arizona. His undergraduate work focused on improving the noninvasive methods of identifying tumor acidosis by integrating chemical exchange saturation transfer (CEST) MRI and T2 exchange MRI with machine learning dimensionality reduction. Paul also engaged in work in immunology, investigating transduction events of a T cell co-receptor through a cell-free membrane reconstitution.

Currently, his graduate research focuses on investigating the relationship between tissue morphology and driver mutation functional status in clear cell renal cell carcinoma. Paul believes that there is an abundance of untapped information in tissue morphology, and through deep learning applications we can use this information to improve histopathological practices. Specifically, he is interested in being able to predict the functional status of driver mutations through tissue morphology and finding the biological features that drive this classification.


Vandana Panwar

Senior Research Associate

Vandana is the point person in the lab for classical pathology and digital slide annotation and characterization. She brings with her a wealth of anatomic pathology experience both in clinical and research settings, with a particular focus on genitourinary tumors.

Vandana received her MD degree in Pathology from Sarojini Naidu Medical College,India. She gained her passion for research during her appointment at the OBGYN Department at UCLA . Vandana has been at UT Southwestern since 2016, driving digital pathology efforts in genitourinary cancers research projects resulting in multiple publications in journals including the Journal of Clinical Oncology and the Journal of Urology.


Vipul Jarmale

Machine Learning

Vipul obtained his Master's in Computer Science with focus on Data Science in May 2018. He's currently involved in designing and optimizing deep learning pipelines. His responsibilities also include building tools to facilitate data generation and visualization.


Jay Jasti

Jay has his Ph.D in Chemical Engineering from the University of Michigan. He worked in various research organizations (Scientific Research Labs at Ford Motor Co, Unocal Science and Technology, Mobil Research and Development Corporation). More recently Jay has worked for Verizon Communications in various assignments. He recent experience has been in the general area of Fraud, Identity and Security (Identity proofing and verification, anomaly detection).

Jay is currently working on developing a kidney tumor classifier for H&E-stained Whole Slide Images using deep learning technologies.



We believe that it is a question of when, not if, machine learning revolutionizes pathology. To get there, there are several, biological, experimental and computational challenges to be overcome. We are constantly on the lookout for smart and motivated scientists (graduate students or postdocs) to join us in this effort. 

  • Unique multi-disciplinary training:  We are a machine learning lab at one of the top academic medical centers in the world, working on problems at the cutting edge of computation and biology. Do you want to learn how to leverage your machine learning background to tackle hard problems in biology? Or maybe you want to be trained in deep learning approaches that would scale up your pathology skills? Come talk to us.
  • Flexible startup-up like environment: We are a young lab in a new department committed to a fast pace. We are not afraid to break the old rules of academia to in search of what works best for our science. We've adopted open office plans with no reserved offices to maximize interactions. We have the ability to hire scientific programmers so that we can focus on algorithms+biology.
  • World-class computational facilities: the BioHPC and Bioinformatics Core Facility in our department provide the strong computational and analytical infrastructure necessary to build your projects.
  • High-standard of living: We offer competitive compensation that rewards industry experience and fellowships in a city with low cost of living.

For a more formal description of what we are looking for please look at our Jobs page.