Your browser doesn't support javascript. This means that the content or functionality of our website will be limited or unavailable. If you need more information about Vinnova, please contact us.

An AI based digital pathology decision support tool to identify and classify lung cancer.

Reference number
Coordinator ContextVision AB - Head quarters
Funding from Vinnova SEK 995 280
Project duration June 2018 - June 2019
Status Completed
Venture Swelife and Medtech4Health- Project proposals to improve health
Call Förbättrad prevention, diagnos och behandling – Swelife och Medtech4health

Purpose and goal

The overall goal of the project was to create an AI-based decision support tool (DST) to assist lung pathologists in histopathologist evaluation of lung biopsies. Our AI-based software is designed to: 1. State whether cancer is present in the tissue sample 2. Indicate areas where cancer is present in the biopsy 3. Classify these areas into different types of lung cancer At the end of the Vinnova project, we have not yet been able to train the network. All lung biopsies were completed-annotated and quality assured in June. Now the actual training and evaluation of the result remains.

Expected results and effects

Previous results when using TMAs have shown very promising results. Now it remains to be seen how well this can be repeated in the much more difficult lung biopsy images.

Planned approach and implementation

The project has been carried out in a number of work packages: WP 1. Business development, which resulted in a product specification and business plan WP 2. Generation of training data, where we identified over 200 patient biopsies where different types of lung cancer are represented. WP 3. Manual annotations, a very extensive work has been done where each biopsy has been minutely annotated and each small cancer area outlined. WP 4. Algorithm design, the training of the AI network itself has not yet begun.

The project description has been provided by the project members themselves and the text has not been looked at by our editors.

Last updated 19 September 2019

Reference number 2018-00184

Page statistics