Home▸AI & Machine Learning▸How CNNs Learn to Classify White Blood Cells Under the Microscope

🩸 How CNNs Learn to Classify White Blood Cells Under the Microscope

Watch a stained blood cell pass through a convolutional neural network's layers as feature maps activate and the model outputs a class prediction across eosinophils, lymphocytes, monocytes and neutrophils.

AI & Machine Learning3DAdvanced60 FPS

A stained white blood cell on a microscope slide passes through a convolutional neural network's layers as feature maps light up, ending in a live probability read-out across four leukocyte classes.

🔬 What It Demonstrates

Each translucent slab is a convolutional layer whose glowing tiles represent feature maps activating in response to shapes like nucleus lobes and granule texture, feeding forward into a final classification.

🎮 How to Use

Pick a cell type, dial in image noise to simulate a poor blood film, and change the network's depth. Press Run Inference to fire a forward pass and watch the output bars update.

💡 Did You Know?

NHS haematology labs already use CNN-based digital morphology analysers to pre-sort blood film images, letting biomedical scientists focus their attention on the cells the model is least confident about.

⚙ Under the hood

artificial intelligenceneural networkscomputer visionclassificationconvolutional neural networkblood cellsmicroscopyThree.js

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install