Dissertation Project

Breast cancer classification research with image, tabular, and exploratory fusion testing.

A research project onimage, tabular, andexploratory fusion testing.

This project brings the main breast cancer classification workflows into one place, with dedicated testing views for image-based analysis, tabular inputs, and exploratory multimodal experiments.

Open testing pageRead the notebook stack
Image branch89%Patient-level ROC AUC
Tabular branch96%Wisconsin benchmark accuracy
Image accuracy77%
Best synthetic fusion accuracy100%
Notebook chapters12
Curated figures18
Scroll narrative

One interface. Three research modes.

01

Image-only mode

Upload pathology tiles and get corrected patient-level branch probabilities with a serious diagnostic-style presentation.

02

Tabular-only mode

Use all 30 morphological features in a structured interface designed for transparent tabular testing.

03

Synthetic fusion demo

Explore combined image and tabular inputs in an experimental multimodal workflow for comparison and analysis.

Notebook system

Curated docs, not raw dumps

The notebook page turns the executed research sequence into a browsable reading surface with side navigation, figure cards, and direct notebook downloads.