System and method for pain monitoring using a multidimensional analysis of physiological signals
Inventors
Zuckerman-Stark, Galit • Kliger, Mark
Assignees
Publication Number
US-9498138-B2
Publication Date
2016-11-22
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 relates to a system and method for pain monitoring by performing a multidimensional analysis of multiple physiological signals, enabling pain monitoring, classification, and identification across various states of consciousness including awake, semi-awake, and sedated individuals.
The invention addresses the problem that traditional pain monitoring has primarily focused on unconscious patients under anesthesia (Depth of Anesthesia monitoring), and pain and awareness are often difficult to distinguish in such states. Current methods are limited by reliance on individual physiological signals or a small number of parameters, often insufficient for awake patients.
The present invention overcomes these deficiencies by utilizing a plurality of physiological signals and parameters such as PPG, GSR, ECG, respiration, EMG, EEG, skin temperature, and others, extracting numerous features (a great plurality of features, GPF) from these signals to improve pain detection accuracy. It further applies advanced signal processing techniques including normalization, feature selection, dimensionality reduction, and classification using machine learning algorithms, optionally incorporating a priori data relating to patient history or conditions, thereby facilitating pain monitoring for patients in various consciousness states.
Claims Coverage
The patent includes claims directed to both a method and a system for monitoring pain using multiple physiological signals, focusing on the processing, feature extraction, transformation, and classification steps.
Obtaining multiple physiological signals including blood volume change and at least one selected signal
The method and system obtain at least two physiological signals comprising blood volume change and one or more signals selected from GSR, ECG, respiration, internal and skin temperature, EOG, pupil diameter, EEG, FEMG, EMG, EGG, partial pressure of carbon dioxide, and accelerometer readings.
Processing physiological signals to improve quality
The signals are processed to improve signal quality through synchronization, noise filtering, artifact reduction, or similar preprocessing techniques to form processed signals suitable for feature extraction.
Generating a feature vector from extracted features
A first vector is generated comprising at least three features extracted from the physiological signals, including features such as mean Peak amplitude, Peak mean amplitude, Peak standard deviation of amplitude, Trough amplitude, peak to peak intervals, and Peak-to-Peak High Frequency Power, among others.
Transforming the feature vector by normalization
The first feature vector is transformed into a second vector by applying normalization techniques, optionally including feature selection and dimensionality reduction, to enhance classification performance.
Monitoring pain by classification using ensemble methods
Pain status is monitored by applying a classification algorithm adapted to classify the second vector into a graduated scale representing pain severity, wherein the classification algorithm comprises an ensemble of classification and regression trees, including random forest classifiers or boosting frameworks.
Adaptation for various patient states and conditions
The classification algorithm and pain monitoring method are adapted for patients in different states of consciousness (unconscious, anesthetized, sedated, awake, semi-awake) and for pain experienced with particular diseases, stimuli, or medicaments.
Communication and display of pain monitoring results
The system further comprises modules for communicating pain status to caregivers, higher processing centers, or call centers and may include display modules to present pain classification results.
The claims cover a comprehensive method and system employing acquisition of multiple physiological signals, advanced signal processing including feature extraction and normalization, and sophisticated classification algorithms based on ensembles of trees to monitor and classify pain levels across various states of consciousness and conditions, with communication and display capabilities.
Stated Advantages
Improved pain and no-pain classification performance by using a combination of multiple physiological parameters rather than individual signals alone.
Capability to monitor pain in subjects exhibiting various states of consciousness including awake, semi-awake, and sedated individuals, overcoming limitations of prior art focused on unconscious patients under anesthesia.
Enhanced classification accuracy through employing a great plurality of features extracted from multiple physiological signals, combined with advanced feature selection, dimensionality reduction, and machine learning classifiers.
Flexibility to incorporate a priori data such as patient history, disease, stimulus, or medicament information, further improving classification efficiency.
Facilitation of communication and display of pain status to caregivers or external systems enabling real-time pain monitoring and potential integration with pain management protocols.
Documented Applications
Pain classification and monitoring in patients in various states of consciousness, including awake, partially sedated, sedated, semi-awake, and unconscious under anesthesia.
Pain monitoring in clinical contexts involving various diseases, stimuli, or medicaments, supporting personalized assessment.
Use in medical settings for monitoring and managing pain through detection, classification, and communication of pain levels to caregivers or control systems.
Potential integration with drug administration devices for automatic or semi-automatic pain medication delivery based on monitored pain status.
Interested in licensing this patent?