DNIC Using Deep Learning and Artificial Intelligence

Université de Sherbrooke

Status

Suspended

Conditions

Chronic Pain

Treatments

Other: Conditioned pain modulation test

Study type

Observational

Funder types

Other

Industry

Identifiers

NCT04896827

2021-4227

Details and patient eligibility

About

Chronic pain (CP) is disabling for people triggering important costs for society. A deficit of diffuse noxious inhibitory controls (DNIC) is one of the CP mechanisms. DNICs are evaluated in research setting using a CPM protocol (conditioned pain modulation). There is a lack of reference values on the effectiveness of DNICs. Wider research on DNIC will help to understand CP and to develop a clinical screening test evaluating DNICs. This study aims more specifically to determine whether it is possible to develop a facial recognition system to automate pain measurement and the effectiveness of pain control mechanisms.

Full description

This study aims:

To develop and validate a predictive tool (using deep learning and artificial intelligence) to estimate the efficacy of pain control mechanisms.
To estimate references values for facial expressions of pain control mechanisms in healthy and in chronic pain participants.

The target population will be healthy volunteers and volunteers with chronic pain, male and female, stratified by age.

The reference values (healthy volunteers) will be established via a non-parametric method for a standard conditioned pain modulation (CPM) protocol in which two "stimuli tests" of the same intensity and nature (heat) will be applied before and after the application of another "conditioning stimulus" (cold water bath). The perceived pain difference between the 1st and 2nd stimuli tests will reflect the intensity of the DNICs. Participants' facial expressions will be captured simultaneously by three cameras during the CPM testing.

These results will be compared to those from volunteers suffering with chronic pain. The clinical decision rule will result from clinical and paraclinical elements correlating with the amplitude of the efficacy of CPM (serum noradrenaline, intensity of pain, heart rate and blood pressure measurements, psychometric questionnaires assessing anxiety, depressive feelings and pain catastrophizing). Logistic regression analysis will determine the best predictors of a CPM deficit.

Enrollment

244 estimated patients

Sex

All

Ages

18 to 79 years old

Volunteers

Accepts Healthy Volunteers

Inclusion and exclusion criteria

Healthy participants

Inclusion Criteria:

18-79 years old
No chronic pain
Able to provide consent

Exclusion Criteria:

Cardiovascular disease (arrhythmia, cerebrovascular accident, infarction...)
Raynaud syndrome
Severe psychiatric disease (dementia, schizophrenia, psychosis, major depression)
Injuries or loss sensitivity to their forearms or hands
Pregnant women or in post-partum period (<1 year)

Participants with chronic pain

Inclusion Criteria:

18-79 years old
Chronic pain (chronic pain is defined by any regular pain for more than 6 months)
Able to provide consent

Exclusion Criteria:

Cardiovascular disease (arrhythmia, cerebrovascular accident, infarction...)
Raynaud syndrome
Severe psychiatric disease (dementia, schizophrenia, psychosis, major depression)
Injuries or loss sensitivity to their forearms or hands
Pregnant women or in post-partum period (<1 year)
Chronic pain caused by cancer or migraine