Development of Artificial Intelligence Tools for the Detection of Stress Markers and Consideration of Stress States in the Monitoring of Subjects With Type 1 Diabetes (SMART-T1D)

Centre d'Etudes et de Recherche pour l'Intensification du Traitement du Diabète

Status: Enrolling

Conditions: Diabetes Type 1

Treatments: Other: Voice recording 4 times a day

Study type: Interventional

Funder types: Other

Identifiers

NCT06985862

2024-A02431-46

Details and patient eligibility

About

Stress refers to all the reactions of an organism subjected to exogenous or endogenous stress. In the context of diabetes, stress plays a critical role. There are two forms of stress: acute and chronic, both of which can have a significant impact on patients' glycaemic control. Acute stress, if repeated, can cause rapid increases in blood glucose levels, while chronic stress can lead to insulin resistance. It is therefore essential to develop tools for recognising and quantifying stress states specific to patients with diabetes. These tools would provide a better understanding of the role of stress in diabetes management, paving the way for more targeted therapeutic interventions and improving patients' quality of life.

We are currently training algorithms using advanced machine learning and artificial intelligence techniques to recognise and quantify stress states using existing databases, including voice and physiological data. These technological advances will make it possible to identify moments of stress more accurately and provide appropriate responses, thereby contributing to better diabetes management.

The SMART-T1D study is an ancillary study of the EVASTRESS study.

Enrollment: 35 estimated patients

Sex: All

Ages: 18+ years old

Volunteers: No Healthy Volunteers

Inclusion criteria

• Patient who has signed the SMART-T1D free and informed consent form
• Patient able to speak and read French

Exclusion criteria

• Mute patient.
• Patient with severe speech problems that may prevent voice recordings from being made.

Trial design

Primary purpose: Prevention

Allocation: N/A

Interventional model: Single Group Assignment

Masking: None (Open label)

---

35 participants in 1 patient group

Voice recording 4 days a day

Experimental group

Description:

* Morning (first recording): Text reading (article 25.1 of the Declaration of Human Rights). * Noon (second recording): Counting from 1 to 20 at normal speed * Evening (third recording): Prolonged phonation of the vowel 'a' without catching your breath. * Bedtime (fourth recording): Free expression describing stressful moments of the day and their impact on diabetes management, for at least 30 seconds.

Treatment:

Other: Voice recording 4 times a day