Research area

I focus on multi-scale characterization of early-stage cervical cancer and precancerous conditions, combining nuclear morphological biomarkers with molecular gene signatures. My work spans two complementary tracks: cytomorphological analysis of precancerous changes in Pap smear images and molecular biomarker analysis of early-stage invasive cervical cancer using gene expression data. The goal is to enable earlier detection and more precise, personalized diagnosis by linking structural and molecular disease signatures.

Methods

• Nuclear Morphological Analysis: Quantifying chromatin texture, radial organization, and structural features of koilocytotic and normal nuclei in Pap smear images to distinguish precancerous from healthy cells

• Differential Gene Expression: Analyzing RNA-sequencing data to identify significant expression changes between cancerous and non-cancerous tissue, with a focus on genes linked to nuclear and chromatin organization

• Machine Learning: Building and validating classification models to test the diagnostic value of morphological and molecular features, including feature importance analysis for interpretability

• Data-Driven Insights: Applying bioinformatics, texture analysis, and statistical methods to ensure robust, reproducible, and interpretable findings

Applications: 

The identified morphological and molecular biomarkers aim to:

• Enable earlier and more accurate detection of precancerous and early-stage cancerous changes

• Provide interpretable, image-based diagnostic tools alongside molecular markers

• Facilitate targeted therapy development and treatment personalization