Publications & Media

A Hybrid Deep Learning Approach for Breast Tumor Classification, Segmentation and Phytochemical Recommendation

Status: Under review at an international journal (submitted March 2026)

A two-stage AI framework that classifies breast cancer metastases in histology images (95%+ accuracy) and recommends traditional medicinal plant compounds. Designed for low-resource healthcare settings where pathologists are scarce.

Preprint on request

PyDataQuality — Enterprise Data Quality Engine for Python

Automates the tedious 80% of data science: validation, profiling, and cleaning of new datasets with a single line of code. Used by data teams to produce quality reports in seconds instead of minutes. Includes AI-powered remediation prompts.

GitHub Repository

OncoVision: METABRIC Risk Stratifier for Breast Cancer

Implements a Random Survival Forest model (C-index 0.7045) on the METABRIC cohort of 1,980 patients. Deployed as an interactive Streamlit app, it stratifies patients into low, medium, and high risk groups with event rates from 35.9% to 87.5%. A clinical decision support tool for personalised oncology.

GitHub Repository