Related Experiment Video
Updated: Jun 6, 2026
Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli
Published on: April 5, 2019
Refining multiple artificial intelligence strategies for automatic pain assessment investigations (RUGGI Study): A
Marco Cascella1, Alfonso Maria Ponsiglione1, Vittorio Santoriello1
1From the Anesthesia, Intensive Care and Pain Medicine, Department of Medicine, Surgery and Dentistry 'Scuola Medica Salernitana', University of Salerno, Baronissi (MC, OP), Department of Electrical Engineering and Information Technology, University of Naples 'Federico II', Naples (AMP, VS, MR, FA) and Oncology, Department of Medicine, Surgery and Dentistry, University of Salerno, Baronissi, Salerno, Italy (FS, SP).
Background:
Chronic pain is a complex, multidimensional condition that severely impairs patients' quality of life. As conventional techniques for evaluating pain are based on subjective self-reporting, these approaches have crucial drawbacks, especially for individuals with communication difficulties. Artificial intelligence (AI) provides the opportunity to complement subjective self-reports through multimodal, data-informed analysis to enhance real-time pain assessment and care.
Objective:
To develop, calibrate and validate AI models for automatic pain assessment (APA) in adult patients by merging physiological, behavioural and clinical data and, consequently, complement patient-reported information and support more personalised and effective pain management.
Design:
Prospective, single-centre, noninterventional study.
Setting:
University of Salerno Hospital, Italy.
Patients And Participants:
Adult patients (>18 years) with chronic primary or secondary pain (oncologic and nononcologic), able to provide their informed consent. The main exclusion criteria are severe psychiatric or cognitive disorders and treatment with psychotropic medications.
Primary Outcome Measures:
Predictive performance of AI models (sensitivity, specificity, area under the receiver operating characteristic curve, AUC-ROC) for automatic pain assessment based on collected multimodal data.
Secondary Outcomes:
Quality-of-life evaluation, analgesic treatment monitoring, development and analysis of a multidimensional dataset for APA and identification of correlations between clinical and physiological variables.
Results:
N/A (study ongoing).
Conclusions:
This study will provide essential data for developing and validating integrated AI tools for objective, multidimensional pain assessment, with potential future clinical and therapeutic applications.
Trial Registration:
ClinicalTrials.gov Identifier: NCT07038434.
