System and method for pain monitoring using a multidimensional analysis of physiological signals
Inventors
Zuckerman-Stark, Galit • Kliger, Mark
Assignees
Publication Number
US-8512240-B1
Publication Date
2013-08-20
Expiration Date
2028-11-13
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
The present invention is for a method and system for pain classification and monitoring optionally in a subject that is an awake, semi-awake or sedated.
Core Innovation
The invention provides a system and method for pain monitoring, classification, and identification in individuals exhibiting various states of consciousness, including awake, semi-awake, sedated, and unconscious states. This system is achieved by performing a multidimensional analysis of a plurality of physiological signals and parameters to determine pain levels.
The background problem addressed is that prior art pain monitoring and Depth of Anesthesia (DOA) systems are limited in that they primarily focus on patients in unconscious states, rely on analysis of individual physiological signals, or are dependent heavily on the number and type of physiological variables used. Traditional methods fail to provide accurate pain monitoring for patients with varying levels of consciousness and do not adequately combine multiple physiological signals for improved pain classification performance.
The invention overcomes these deficiencies by utilizing a great plurality of physiological features (GPF), extracted from at least three or more physiological signals, including Photoplethysmograph (PPG), Galvanic Skin Response (GSR), and optionally skin temperature, ECG, respiration, EMG, and EEG/FEMG. The system comprises modules for signal acquisition, preprocessing, feature extraction, feature selection and dimensionality reduction, and classification using machine learning techniques. This method provides pain classification into at least two classes with optional multi-class and scalable scoring.
Claims Coverage
The patent contains several independent claims, each detailing inventive features related to a system and method for pain monitoring based on physiological signals.
Pain monitoring method using PPG and GSR physiological signals
A method comprising obtaining Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) physiological signals, processing to improve signal quality, extracting at least three features selected from PPG mean Peak amplitude, PPG peak to peak time intervals, PPG high frequency power, GSR amplitude, and GSR peak to peak time intervals, transforming these into a normalized feature vector, and applying a classification algorithm comprising an ensemble of classification and regression trees to classify pain level on a graduated scale.
Pain monitoring system with acquisition and processing modules
A system comprising a signal acquisition module measuring at least PPG and GSR physiological signals and a processing module configured to improve signal quality, extract the specified features, normalize the feature vector, and apply a classification algorithm based on ensemble of classification and regression trees to classify pain on a graduated scale.
Optional inclusion of additional physiological signals and a priori data
The method and system optionally include additional physiological signals such as ECG, blood pressure, respiration, temperature, EOG, pupil diameter, EEG, EMG, EGG, LDV, capnograph, accelerometer readings, and may process a priori data including environmental and patient parameters, disease, stimulus, and medicament for improved pain classification.
Adaptation to states of consciousness and pain classification contexts
The method and system are adapted to classify pain in patients in various states of consciousness (unconscious, anesthetized, sedated, awake, semi-awake) and for pain associated with particular diseases, stimuli, or medicaments, with specific features selected accordingly from the physiological signals.
Use of random forest or boosting frameworks for classification
Classification algorithms include use of random forest classification or boosting frameworks within the ensemble of classification and regression trees for improved pain level detection and accuracy.
Communication and display of pain monitoring results
The method and system include communicating monitored pain status to receiving units such as higher processing centers, caregivers, call centers, and optionally providing a display module for pain level presentation.
The claims cover methods and systems that perform multidimensional analysis of physiological signals, primarily PPG and GSR, combined with feature extraction, normalization, and classification using ensemble methods like random forest or boosting, optionally expanded to additional signals, a priori data, various consciousness states, and communicative outputs for accurate pain monitoring and classification.
Stated Advantages
Improved pain and no-pain classification performance by utilizing a combination of multiple physiological signals rather than relying on single signal analysis.
Capability to monitor pain in subjects across various states of consciousness including awake, semi-awake, sedated, and unconscious.
Flexibility to classify pain into multiple classes with scalable pain scoring correlated to subjective pain scales like Visual Analog Scale (VAS).
Enhanced classification efficiency by inclusion of a priori data such as patient history, disease, and medicament information alongside physiological features.
Provision of a modular system that can be implemented as local, semi-remote, or fully remote configurations facilitating varied clinical scenarios.
Documented Applications
Pain classification and monitoring for subjects in various states of consciousness including awake, sedated, semi-awake and unconscious patients.
Pain monitoring related to specific pain stimuli such as cold pressor test and heat pain test in clinical trials.
Pain monitoring and classification in contexts involving various diseases, stimuli, and medicaments.
Use in clinical environments for continuous, non-invasive monitoring of pain to guide pain management and medicament delivery.
Remote and telemedicine applications using systems configured for remote physiological signal acquisition, processing, and real-time pain status communication.
Interested in licensing this patent?