Image-Quality Triage Tool for a Tele-Radiology Network
Overview
What this challenge is about.
Train a multi-label CNN to flag rotation, clipping, and motion in chest X-rays, then build a Gradio demo and write a product memo. Get a verifiable certificate.
The scenario
The vendor (around 60 staff, around 4M reads per year across 14 partner networks) loses 6-9% of reads to non-diagnostic images today, and a workable retake-prompt at acquisition would translate to direct radiologist throughput gains.
The Brief
What you'll do, and what you'll demonstrate.
Build and demo an image-quality triage model for chest X-rays that flags non-diagnostic images at acquisition with a defended false-positive budget.
Earning criteria — what you'll demonstrate
- Apply multi-label CNN classification to medical-imaging quality control
- Translate per-flag probabilities into a single workflow decision
- Defend a false-positive budget against operational impact
- Demo a working ML tool in a workflow-relevant interface
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Aligned coursework coming soon.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
- Medical Imaging
Apply medical imaging to solve real industry problems and demonstrate production-level capability.
- Classification
Apply classification to solve real industry problems and demonstrate production-level capability.
- Convolutional Neural Networks
Apply convolutional neural networks to solve real industry problems and demonstrate production-level capability.
- Model Evaluation
Apply model evaluation to solve real industry problems and demonstrate production-level capability.
- Demo Development
Apply demo development to solve real industry problems and demonstrate production-level capability.
- Pytorch
Apply pytorch to solve real industry problems and demonstrate production-level capability.
Careers
Career paths this challenge builds toward
Completing this challenge demonstrates skills that transfer directly to these roles:
Computer Vision Engineer
Shipping a CNN-based quality-triage tool with a working demo is exactly the day-one CV-engineer deliverable at tele-radiology and imaging-vendor startups.
This challenge sharpens
- medical-imaging
- convolutional-neural-networks
- classification
AI Engineer
Wrapping a multi-label model into a workflow-placed demo with a defended false-positive budget is the AI-engineer's bread-and-butter at applied healthtech teams.
This challenge sharpens
- demo-development
- classification
- pytorch
Applied AI Scientist
Tying model decisions to operational metrics like technologist retake load is the applied-AI-scientist's craft at any healthtech product team.
This challenge sharpens
- model-evaluation
- medical-imaging
- classification