Cell microscopy visual
Research direction Cell-Level Analysis
Current focus Concept Evidence
PhD Research Project

HistoVision Cell-Level Intelligence for Digital Pathology

HistoVision is a research-driven digital pathology system focused on cell-level detection, interpretable visual evidence, and transparent analysis of histopathology images. The long-term vision is to support image-based analysis from tissue regions to individual cells, combining automated detection, concept evidence, and explainable model outputs.

Enter Research Panel View research pathway
Cell Detection Research foundation
Concept Evidence Current development stage
Explainability Long-term direction
Research workflow

Research development pathway

HistoVision is being developed through a staged research workflow, moving from cell detection toward concept evidence, interpretability, and cell-level classification in digital pathology.

Step 1 Cell Detection
Completed / Accepted paper

Cell Detection

This stage focuses on detecting cell instances in histopathology image regions. The first stage has led to an accepted research paper and provides the technical foundation for the next stages.

Detection Cell localisation Accepted work
Step 2 Concept Dataset
Current active stage

Concept Dataset Construction

The current research panel supports the construction of a concept-level annotation dataset using selected cell crops. This stage helps collect interpretable visual evidence for later model development.

Annotation Concept evidence Cell crops
Step 3 Interpretability
Planned stage

Interpretability Modelling

This stage will connect model predictions with explicit visual evidence, so the system can explain decisions through meaningful cell-level concepts rather than only black-box outputs.

Explainability Visual evidence Model reasoning
Step 4 Cell Classification
Planned stage

Cell Classification

This stage will use cell-level evidence and learned concepts to support more transparent classification of cellular patterns in histopathology images.

Classification Cell patterns Transparent output
Future Image Analysis
Future direction

Image Upload & Analysis

The long-term direction is to extend HistoVision toward user-submitted histopathology images, automated detection, interpretable analysis, and structured reporting.

Upload Automated analysis Reporting
Step 1 · Completed

Cell Detection

Detection and localisation of cells in histopathology image regions. This stage has led to an accepted research paper.

Step 2 · Current

Concept Dataset Construction

Current research panel for concept-level annotation using selected cell crops.

Step 3 · Planned

Interpretability Modelling

Connecting predictions with explicit visual evidence and concepts.

Step 4 · Planned

Cell Classification

Transparent classification using cell-level evidence and learned concepts.

Future

Image Upload & Analysis

Future extension toward uploaded histopathology images and interpretable reporting.