United KingdomHealthcare and life sciences2023 – ongoingFrancis Crick Institute
Image classification of Parkinson's disease mechanisms in laboratory research
Francis Crick researchers classified Parkinson's disease models with up to 95% accuracy
Problem
Researchers needed to distinguish cellular mechanisms associated with Parkinson's disease. Assessing microscope images manually was slow and prone to error.
Approach
Faculty worked with the institute to develop a computer vision model that analysed microscope images of neurons. It distinguished healthy cells from disease models and classified different disease mechanisms. It also identified image regions that influenced its predictions.
Results
Takeaway
The reported accuracy concerned laboratory cell images. Clinical diagnosis of patients remained a separate validation task.
Categories
- Industry
- Healthcare and life sciences
- Business function
- IT and R&D
- Innovation type
- Process innovation
- AI technology
- Computer vision
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
Faculty described a completed research collaboration using neuron images to distinguish disease mechanisms. It reported classification accuracy of up to 95% and said the research was published.
Full details are in the original sources.
