Research
Our Research Vision
At SinhaLab, our central goal is to understand the process of aging using big data via Machine Learning. We build computational methods to better understand the fundamental mechanisms of aging and develop strategies to prevent age-related decline.
Our current focus is to use high-resolution tissue pathology images available at scale to establish the principles of how different tissues architecture are deteriorating (from images), identify its molecular regulators (multi-omics), lifestyle and genetic risks (EHR) and detect it from blood.
Computational Approaches to Aging
We apply and develop novel computer vision techniques to extract meaningful biological insights from tissue images. Our lab has pioneered several innovative methods that integrate multi-modal data to understand aging processes.
Our multi-modal deep learning architecture integrates imaging data with genomic information to provide richer insights into aging processes than either data type alone could offer. This two-step deep learning framework is guided by transcriptomics for enhanced tissue analysis, revealing aging signatures that would be otherwise undetectable.
Current Research Directions
Tissue-Specific Aging Mechanisms
How do aging processes differ across tissue types? Can we identify universal vs. tissue-specific aging signatures from histopathology images? We're developing approaches to characterize aging across diverse human tissues at unprecedented resolution.
Digital Pathology for Normal Tissue Analysis
Can computational approaches reveal age-related changes in normal tissues that evade conventional analysis? We're developing novel digital pathology methods to detect subtle architectural changes in non-diseased tissues that correlate with aging.
Aging-to-Cancer Transformation
What age-associated tissue changes create vulnerability to malignant transformation? We're investigating how aging-related alterations in tissue architecture and molecular pathways contribute to cancer development and progression.
Circulating Aging Biomarkers
Can aging signatures be detected from routine blood tests? We're developing computational methods to identify aging biomarkers in blood that correlate with tissue-level changes, potentially enabling non-invasive aging assessment.
Our Interdisciplinary Approach
We are a diverse team comprising expertise spanning machine learning, bioengineering, computer vision, and omics analysis. Our interdisciplinary approach allows us to address complex questions about the aging process that weren't previously accessible through traditional methods.
By integrating cutting-edge computational approaches with biological insights, we aim to develop a comprehensive understanding of aging that can inform interventions to promote healthier aging.