What is the solution about?
➡️Our product aims to give predictions based on Machine Learning, and in particular, Deep Learning, to forecast the mortality risk of a particular COVID-19 patient given CT scans, tabular data, and other possible sources of information.
The solution is divided in two different layers. The first layer uses images as input, and the second layer model combines this prediction with the other sources of information.
What did you achieve?
➡️After data was retrieved from the assigned hospital partner, it was pre-processed and used to validate the model performance obtaining satisfactory results, with an AUROC of 0.84, above the indicated KPI threshold of 0.8.
Regarding computational performance, tests show computational time was below 1 second per patient and memory usage below 16GB for the complete validation process.
Who are the end users of the technology?
➡️ We consider any clinical or administrative manager at any clinical institution treating COVID-19 patients a target customer.
Let us know the impact on society and patients
➡️We hope that our product can help clinicians give appropriate personalized care to COVID-19 patients and better deal with the impact of COVID-19, facilitating the reduction of mortality rates related to this disease while also allowing a more appropriate allocation of resources.
Give us a hint about the learnings from the COVID-X programme
➡️- Practical information about regulation in healthcare for Machine Learning products.
– How to use Continuous Integration / Continuous Deployment (CI/CD) tools like Jenkins.
– Very interesting pitching techniques, with a different approach to the ones we have used in the past
How was your experience within COVID-X?
➡️“The Covid-X programme has offered us constant support and motivation. When we have faced difficulties or roadblocks, we have always found help from someone in the consortium. Getting advice and help from inspiring and knowledgeable mentors and coordinators has been very motivating.” A. Blanco – Head of Business Development Sigesa.
What was the part you enjoyed the most within the COVID-X programme?
➡️- Some webinars and mentors were really helpful, especially regarding regulation and CE marking.
– Technical tutorials and examples of CI/CD workflows.
– The positive atmosphere and motivation among teams and mentors.
– Constant availability of very experienced professionals for help and support
#COVIDXSuccessStories #GameON #DataVsCovid #TheCovidX