FirstVet - Enriched medical records for the best possible veterinary advice
Reference number | |
Coordinator | FirstVet AB |
Funding from Vinnova | SEK 442 500 |
Project duration | December 2020 - June 2021 |
Status | Completed |
Venture | AI - Competence, ability and application |
Call | Start your AI-journey for businesses - autumn 2020 |
Important results from the project
FirstVet´s veterinarians today spend a lot of time documenting video calls with FirstVet´s customers in continuous text. Through the project "Enrich the record keeping for the benefit of the animals", FirstVet has created the opportunity to reduce the administration for the veterinarians with the help of a machine learning algorithm that provides suggestions for existing treatments based on the veterinarian´s medical history.
Expected long term effects
** Denna text är maskinöversatt ** The result: Through text analysis, the machine learning algorithm produces suggestions for recommended treatments that are based on the anamnesis text. The veterinarian then has the opportunity to select, modify and / or ignore the proposals. Effects: The veterinarian´s choice is saved and included in the data base, in this way the machine learning algorithm will continuously learn and improve its precision over time.
Approach and implementation
The project took place in collaboration with Advectas and we divided the project into four phases. The first phase we went through data models, how the veterinarians work and the data needed for the project. Advectas began to test different methods iteratively to compare and cluster journal cases. Evaluation of the pilot was done by FirstVet´s veterinarians by reviewing the results of these methods. The solution was adjusted according to the veterinarian´s feedback and put into production in AWS. The transfer to FirstVet has taken place and will be integrated with the veterinary tool.