Assessment of Postoperative Pain Through Facial Expressions Using Facial Recognition Software

Assessment of Postoperative Pain Through Facial Expressions Using Facial Recognition Software: a Prospective Observational Feasibility Study

Proper management of postoperative pain is an ongoing medical challenge. Inadequate treatment of pain is associated with significantly worse patient outcomes. However, as pain is a subjective experience accurate assessment is difficult.

Commonly used methods for pain assessment include the use of self-reports from patients, or observers assessments.

However, both techniques are subjective to bias. Therefore, automatic assessment of pain based on objective data would enable individualized patient care, optimize provided anesthesia treatment and analgesic regimes.

While research has shown that facial expressions are valid indicators of pain levels, to date research has yet to yield a reliable clinical tool which can be easily implemented in clinical practice.

In this pilot study we intend to assess the feasibility, of facial expression analysis by using machine learning models of artificial intelligence (AI) to accurately predict pain levels of patients experienced in the immediate post operative period.

This pilot trial will take place in two stages:

First stage will include development of an AI algorithm that correlates facial recognition with pain levels.

Second stage will include validation of the algorithm by comparison of to standard pain assessment modalities.

In the first stage each assessment of facial expressions will be filmed in a 30 second segment and will be followed by an immediate pain assessment using two modalities, first will be pain score assessed by an anesthesiologist attending the patient at that moment, second will be VAS assessment by the participant patient. Three objective parameters: heart rate, blood pressure and respiratory rate will be recorded simultaneously from the automated record keeping system used in every patient in the recovery room (post anesthesia care unit-PACU).

These assessments will take place at different time intervals according to the investigator's decision, throughout the participant's staying in the post anesthesia care unit.

After completion of the first stage, the second stage of the study will be done in the same manner as described above regarding patients enrollment. Pain assessment will be done by VAS and physician assessment as described above but this time will be correlated with pain assessment by the algorithm developed in the first stage of the study.

Study Overview

• Status: Recruiting
• Conditions: Pain, Postoperative
• Intervention / Treatment: Other: Filming facial expressions

Detailed Description

After consenting to participation patients will undergo an explanation on pain assessment using VAS and then will proceed with surgery, inclusion in the study will not affect anesthesia or surgery management in any way.

Study participation will take place in the PACU. Upon admission to the PACU unit, all study participants' facial expressions will be videoed by a camera placed in front of the patient's bed.

The facial expressions will be filmed in 30 second segments. A pain assessment will be measured immediately following filming of each segment using two modalities:

• Pain score assessed by an attending anesthesiologist assigned to the study team.
• VAS assessment by the patient.
• Three objective parameters: heart rate, blood pressure and respiratory rate will be recorded simultaneously from the automated record keeping system used in every patient in the PACU The quantity of segments filmed for each of the participants will be decided by the investigator taking into account participant's cooperation level and VAS levels.

In order to engineer an accurate predictive model the dataset will also include participants reporting a VAS of 0- experiencing no pain.

Data Management:

Following data collection, the data will be forwarded in a coded manner, according to Clalit's data security regulations, to Third Eye systems a facial recognition software company.

For first stage Third Eye systems will analyze and process the data using AI and machine learning models and develop an algorithm that can predict pain level by watching facial expressions.

After completion of the first stage, the second stage of the study will be done in the same manner as described above regarding patients enrollment. Pain assessment will be done by VAS and physician assessment as described above but this time will be correlated with pain assessment by the algorithm developed in the first stage of the study.

This feasibility study is pilot study to examine whether there is a positive correlation, on a relatively small sample size analysis, using simple resources and limited data to perform this study.

In the event that a positive hypothesis can be confirmed, a second stage observational study with a large sample size and an increased data source will be investigated.

• Study Type: Observational
• Enrollment (Anticipated): 200

Contacts and Locations

Study Locations

• Israel
  • Petach Tikva, Israel, 49100
    • Recruiting
    • Rabin Medical Center/Beilinson Campus

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 18 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

patients above 18 presenting for an elective surgery at Beilinson Hospital

Description

Inclusion Criteria:

All patients above 18 presenting for an elective surgery at Beilinson Hospital following obtaining written informed consents form with the ability to comply with the study requirements will be included in our study.

Exclusion Criteria:

• Patients under the age of 18
• Patients unable to sign an informed consent
• Patients with a history of psychiatric disease.

Study Plan

How is the study designed?

Design Details

• Observational Models: Cohort
• Time Perspectives: Prospective

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort

Intervention / Treatment

Study group; Patients will be requested to sign an informed consent in which they will agree to have their face filmed in the post-anesthesia care unit. The facial expressions will be filmed in 30 second segments. A pain assessment will be measured immediately following filming of each segment using two modalities: Pain score assessed by an attending anesthesiologist assigned to the study team.; VAS assessment by the patient. Following data collection, the data will be forwarded in a coded manner, according to Clalit's data security regulations, to Third Eye systems a facial recognition software company. Third Eye systems will analyze and process the data using AI and machine learning models and develop an algorithm that can predict pain level by watching facial expressions.

Other: Filming facial expressions; Study participants' facial expressions will be videoed by a camera placed in front of the patient's bed during their stay in the post anesthesia care unit.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
visual analog scale of pain (VAS)	vas scores of pain from 0-10 a lower number reflecting a lower measure of pain the study will try to correlate VAS scores with automatic facial expression , accurately reflecting the pain intensity measurements of patients in the immediate post-operative period during duration in the PACU ( post anesthesia care unit).	stay during pacu up to 6 hours

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
automatic analysis of facial expressions and observer pain assessment scores.	Our secondary endpoint is to examine whether automatic analysis of facial expressions correspond with observer pain assessment scores.	stay during pacu up to 6 hours

Collaborators and Investigators

• Sponsor: Rabin Medical Center
• Collaborators: Third eye systems

Study record dates

Study Major Dates

• Study Start (Actual): July 25, 2021
• Primary Completion (Anticipated): December 1, 2022
• Study Completion (Anticipated): July 1, 2023

Study Registration Dates

• First Submitted: May 30, 2021
• First Submitted That Met QC Criteria: May 30, 2021
• First Posted (Actual): June 4, 2021

Study Record Updates

• Last Update Posted (Actual): May 3, 2022
• Last Update Submitted That Met QC Criteria: May 2, 2022
• Last Verified: May 1, 2021

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Pathologic Processes; Postoperative Complications; Pain; Neurologic Manifestations; Pain, Postoperative
• Other Study ID Numbers: 0778-20

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: Undecided

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No