NCT03957967

Acute Pain Assessment Using Facial Expression Analysis

Development of a Digital Solution for Acute Pain Assessment Using Automartized Facial Expression Analysis

• Status: Completed
• Phase: —
• Study type: Observational
• Enrollment: 1,000 (actual)
• Sponsor: Centre Hospitalier Universitaire de Nice · Academic / Other
• Sex: All
• Age: 18 Years
• Healthy volunteers: Not accepted

Summary

Managing pain, which affects 20-50% of the population, is a major issue in daily clinical practice. Evaluation of pain intensity is essential to adapt treatment but as it mainly relies on self-report, this assessment is difficult or impossible in non-communicating patients. In these cases, pain can only be evaluated by medical staff by the observation of pain-related characteristics like facial expression of pain (FEP). However, recognition of FEP is subjective, time-consuming and subject to multiple biases frequently leading to underestimation of pain and consequently under-treatment. Some of these biases could be solved by the use of facial recognition technology, allowing objective, automated and time-saving pain assessment. DEF-I aims to address technical issues and achieve the development of facial expression recognition digital tool able to evaluate severe acute pain in clinical practice, with high validity and utility by improving the quality of the images to be analyzed, by studying larger samples of patients, data and images, in order to correlate more efficiently the pain intensity felt by a patient with the expression of his face. The main objective of this study is to verify whether it is possible to quantitatively correlate the intensity of acute postoperative pain felt by a patient with his facial expression. The secondary objective is to define a reliable computer algorithm that qualitatively correlates the type of acute postoperative pain experienced by a patient with his facial expression.

Conditions

• Acute Pre and Post Operative Pain

Interventions

Type	Name	Description
OTHER	video taken	Patients enrolled will be operated in our institution. Facial expression will be collected before and after surgery, and pain intensity will be collected at the same time. Facial expression will be analyzed using Facial Action Coding System (FACS) .This step will seek to confirm that the subset of Action Units (AUs) defined in previous studies is correlated with the presence and intensity of an acute pain or to identify a different original subset of facial AUs better correlated to pain intensity. . The developed algorithm (or model) here should be able to correlate facial AUs to pain intensity reported by the patients on the numerical rating scale. The developed model at this stage will be then validated on an additional sample of patients.

Timeline

• Start date: 2019-05-31
• Primary completion: 2019-05-31
• Completion: 2019-11-30
• First posted: 2019-05-21
• Last updated: 2020-03-16

Locations

1 site across 1 country: France