Project Details
Description
INTRODUCTION
The use of voice, combined with Artificial Intelligence methods, paves the way to many applications for clinical practice as a potential diagnostic tool or to remote monitor clinical parameters. However, little work has been done in the field of diabetes and in particular type 1 diabetes (T1D). In addition, diabetes distress, which is defined by the burden of stress, fears or emotions linked to the daily management of diabetes, is frequent (1 in 4 people with T1D). We know that it is associated with a poorer quality of life, poorer self-monitoring of blood sugar, an increased risk of depression, poor treatment adherence, greater glycemic variability, increased risk of hypo- and hyperglycemia and consequently with an increased risk of major diabetes-related complications. It is therefore crucial to be able to better identify early high levels of diabetes distress and to be able to monitor this key parameter between clinical visits and intervene when needed to prevent or slow the occurrence of associated major events.
OBJECTIVE
The objective of the CV_VOCADIAB project is to identify candidate vocal biomarkers for the remote monitoring of diabetes distress in people with type 1 diabetes.
MATERIAL AND METHODS
The project is based on the analysis of at least 500 people with type 1 diabetes participating in the CoLive Voice survey, all over the world, for whom standardized (multilingual) audio recordings as well as detailed clinical data are available. Machine Learning algorithms will be trained to predict the presence or absence of diabetes distress, from voice data, alone or in combination with clinical data.
IMPACT
We will identify voice biomarkers for real-life monitoring of diabetes distress in people with type 1 diabetes. Such algorithms will then be integrated into devices for monitoring people with diabetes (smartphone apps, vocal assistants, connected mirrors, telemedicine. ..) and will allow a rapid, unbiased, and regular assessment of this crucial parameter and thus improve the daily life of people living with type 1 diabetes.
The use of voice, combined with Artificial Intelligence methods, paves the way to many applications for clinical practice as a potential diagnostic tool or to remote monitor clinical parameters. However, little work has been done in the field of diabetes and in particular type 1 diabetes (T1D). In addition, diabetes distress, which is defined by the burden of stress, fears or emotions linked to the daily management of diabetes, is frequent (1 in 4 people with T1D). We know that it is associated with a poorer quality of life, poorer self-monitoring of blood sugar, an increased risk of depression, poor treatment adherence, greater glycemic variability, increased risk of hypo- and hyperglycemia and consequently with an increased risk of major diabetes-related complications. It is therefore crucial to be able to better identify early high levels of diabetes distress and to be able to monitor this key parameter between clinical visits and intervene when needed to prevent or slow the occurrence of associated major events.
OBJECTIVE
The objective of the CV_VOCADIAB project is to identify candidate vocal biomarkers for the remote monitoring of diabetes distress in people with type 1 diabetes.
MATERIAL AND METHODS
The project is based on the analysis of at least 500 people with type 1 diabetes participating in the CoLive Voice survey, all over the world, for whom standardized (multilingual) audio recordings as well as detailed clinical data are available. Machine Learning algorithms will be trained to predict the presence or absence of diabetes distress, from voice data, alone or in combination with clinical data.
IMPACT
We will identify voice biomarkers for real-life monitoring of diabetes distress in people with type 1 diabetes. Such algorithms will then be integrated into devices for monitoring people with diabetes (smartphone apps, vocal assistants, connected mirrors, telemedicine. ..) and will allow a rapid, unbiased, and regular assessment of this crucial parameter and thus improve the daily life of people living with type 1 diabetes.
| Acronym | CV_VOCADIAB |
|---|---|
| Status | Finished |
| Effective start/end date | 1/09/21 → 30/08/22 |
Funding
- Société Francophone du Diabète (SFD): €40,000.00
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.