Improving the Diagnoses of Mental Illnesses using Artificial Intelligence Based on Text Data
Reference number | |
Coordinator | Lunds universitet - Institutionen för psykologi |
Funding from Vinnova | SEK 3 211 866 |
Project duration | June 2018 - June 2022 |
Status | Completed |
Venture | Digital health |
Call | 2017-04570-en |
Purpose and goal
Depression and anxiety are one of the greatest public diseases of our time. A faster and more accurate diagnosis can speed up recovery. The project aims to create, test, verify and improve a decision support system to assess the degree of depression and anxiety based on texts generated by patients analyzed with artificial intelligence (AI). Patients describe their mental health in texts, which are analyzed with AI to link the texts to the assessment of patients´ degree of depression and anxiety.
Expected results and effects
The project has resulted in a working prototype for decision support for assessments of patients with depression and anxiety problems in clinical contexts. This work has consisted of the preparation of open text questions; an AI engine for analyzing this data; a user interface where the results are presented for clinicians; feedback from patients, clinicians and patient organizations. The results show that depression and anxiety can be assessed via a decision support based on text answers in a clinical environment that improves the accuracy of identifying depression and anxiety.
Planned approach and implementation
The project has been carried out with project partners from Lunds Universitet, Region Skåne and patient organizations. The project has had six intermediate goals. To build a prototype for a decision support consisting of open text issues linked to depression and anxiety. 2. Evaluate the prototype with participating patient organizations. To comply with CE marking and GDPR. 4. To test the decision support system in clinical contexts. 5. Analyze collected data using AI methods. Disseminate the results of the project to clinics, politicians, patients, clinics and researchers.