Multimodal Platform for Pain Monitoring in Physiotherapy

BMBF Funded Project PainMonit

Our Approach

The core objective of the pain monitoring tool is to provide therapists with assistance during the selection of therapeutic exercises in an appropriate level of intensity for an optimal therapy progress. The analysis of pain perception is of great importance in numerous medical applications. For example, in physiotherapeutic treatments pain, or more precisely the pain threshold, not only affects the course or the result of the treatment, but also shapes the structure and composition of the exercises from the very beginning. The quantitative assessment of one’s own pain is traditionally based on self-assessment with questionnaires. However, for patients who are not able to communicate their pain (objectively), this method is not an option. There are, however, other ways to assess experience pain for instance via physiological and audiovisual parameters.

Experimental setup: Exmanier’s point of view

About the Project

PainMonit aims to develop a sensor-based tool for the holistic monitoring of pain. The pain data is collected via two central project modules and integrated and analyzed on a central platform. The first module comprises the behavioral pain measurement which includes the development of a recording and analysis system for the specification of behavioral movement data using wearable-based sensors. The second module covers a video-based pain measurement and consists of developing an automated camera tool for pain recognition. It captures mimic micro-expression and head movement indicators during the execution of physiotherapy treatment exercises.

Overview Heat Stimuli – Exemplary Temperature Curve with Alternating Stimuli and Pauses

Work Packages, Insights and Outcomes

Ixp is responsible for the development of a software module to monitor pain based on facial expression– and head movement–related indicators. As a basis for machine learning and deep learning models, a video pain database with annotated pain states will be created. In addition, ixp is responsible for assessing the user as well as technical requirements for the software module and handles the implementation of the ELSI design guidelines within the project. Further, ixp administers the final system evaluation of the multimodal PainMonit demonstrator in order to verify reliability, acceptance, usability and accuracy of the classification.

Exemplary course of facial expressions during the different pain states

Face and Facial Landmarks Detection. Regions of Interest: Eye Detection (Eye Closing), Mouth Points (Distances) and Nasolabial Folds

Related Projects