Student project Healthcare AI Deep Learning Computer Vision

Lymphocytosis Classification - Medical Imaging Project

Automated Detection of Reactive vs Tumoral Lymphocytosis from Blood Smears

Posted by Clement Wang on March 1, 2024

Project Overview

This project aimed to build an automated system to distinguish reactive from tumoral lymphocytosis using blood smear images along with patient metadata. The dataset, collected at Lyon Sud University Hospital, included 204 patients—142 for training and 42 for testing. The goal was to support clinicians in identifying cases that require flow cytometry, reducing costs and improving diagnostic accuracy.

This work was part of the Deep Learning for Medical Imaging course by Olivier Colliot and Maria Vakalopoulou.

Approach

We received multiple cell images per patient along with tabular patient information. I experimented with three modeling strategies:

  1. Tabular-only model using patient metadata.
  2. CNN-based image model ignoring tabular data.
  3. Multi-modal model combining images and tabular features.

Due to the small dataset, results were very noisy, with metrics fluctuating on both the validation set and the Kaggle leaderboard. I decided to drop the course because I felt like the design of the competition was wrong. My final submission placed top 6 out of 30 participants, even though I was around 15th on the public leaderboard before the private evaluation shuffled rankings.

The experience highlighted the limitations of small datasets in deep learning for medical imaging and the importance of robust evaluation.

Code Repository

All code for this project is available on GitHub: Lymphocytosis Classification

CATALOG
    FEATURED TAGS
    Student organization Deep Learning Student project Computer Vision Development Gap year Automatants Healthcare AI Optimization Game Internship Machine Learning Time Series Competition Reinforcement Learning Transformers AWS Consulting Contrastive Learning Graph Neural Networks Image Generation Infrastructure Multi-Agent Learning Natural Language Processing Team Management