Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals
This paper presents a novel computational algorithm to estimate blood volume decompensation state based on machine learning (ML) analysis of multi-modal wearable-compatible physiological signals. To the best of our knowledge, our algorithm may be the first of its kind which can not only discriminate...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/4/1336 |
| _version_ | 1827652853277130752 |
|---|---|
| author | Yekanth Ram Chalumuri Jacob P. Kimball Azin Mousavi Jonathan S. Zia Christopher Rolfes Jesse D. Parreira Omer T. Inan Jin-Oh Hahn |
| author_facet | Yekanth Ram Chalumuri Jacob P. Kimball Azin Mousavi Jonathan S. Zia Christopher Rolfes Jesse D. Parreira Omer T. Inan Jin-Oh Hahn |
| author_sort | Yekanth Ram Chalumuri |
| collection | DOAJ |
| description | This paper presents a novel computational algorithm to estimate blood volume decompensation state based on machine learning (ML) analysis of multi-modal wearable-compatible physiological signals. To the best of our knowledge, our algorithm may be the first of its kind which can not only discriminate normovolemia from hypovolemia but also classify hypovolemia into absolute hypovolemia and relative hypovolemia. We realized our blood volume classification algorithm by (i) extracting a multitude of features from multi-modal physiological signals including the electrocardiogram (ECG), the seismocardiogram (SCG), the ballistocardiogram (BCG), and the photoplethysmogram (PPG), (ii) constructing two ML classifiers using the features, one to classify normovolemia vs. hypovolemia and the other to classify hypovolemia into absolute hypovolemia and relative hypovolemia, and (iii) sequentially integrating the two to enable multi-class classification (normovolemia, absolute hypovolemia, and relative hypovolemia). We developed the blood volume decompensation state classification algorithm using the experimental data collected from six animals undergoing normovolemia, relative hypovolemia, and absolute hypovolemia challenges. Leave-one-subject-out analysis showed that our classification algorithm achieved an F1 score and accuracy of (i) 0.93 and 0.89 in classifying normovolemia vs. hypovolemia, (ii) 0.88 and 0.89 in classifying hypovolemia into absolute hypovolemia and relative hypovolemia, and (iii) 0.77 and 0.81 in classifying the overall blood volume decompensation state. The analysis of the features embedded in the ML classifiers indicated that many features are physiologically plausible, and that multi-modal SCG-BCG fusion may play an important role in achieving good blood volume classification efficacy. Our work may complement existing computational algorithms to estimate blood volume compensatory reserve as a potential decision-support tool to provide guidance on context-sensitive hypovolemia therapeutic strategy. |
| first_indexed | 2024-03-09T21:06:59Z |
| format | Article |
| id | doaj.art-3da8c59a9e7c4547a9655b0d4f3702e1 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T21:06:59Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-3da8c59a9e7c4547a9655b0d4f3702e12023-11-23T21:57:56ZengMDPI AGSensors1424-82202022-02-01224133610.3390/s22041336Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological SignalsYekanth Ram Chalumuri0Jacob P. Kimball1Azin Mousavi2Jonathan S. Zia3Christopher Rolfes4Jesse D. Parreira5Omer T. Inan6Jin-Oh Hahn7Department of Mechanical Engineering, University of Maryland, College Park, MD 20742, USADepartment of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA 30308, USADepartment of Mechanical Engineering, University of Maryland, College Park, MD 20742, USADepartment of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA 30308, USAGlobal Center for Medical Innovation, Translational Training and Testing Laboratories, Inc. (T3 Labs), Atlanta, GA 30313, USADepartment of Mechanical Engineering, University of Maryland, College Park, MD 20742, USADepartment of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA 30308, USADepartment of Mechanical Engineering, University of Maryland, College Park, MD 20742, USAThis paper presents a novel computational algorithm to estimate blood volume decompensation state based on machine learning (ML) analysis of multi-modal wearable-compatible physiological signals. To the best of our knowledge, our algorithm may be the first of its kind which can not only discriminate normovolemia from hypovolemia but also classify hypovolemia into absolute hypovolemia and relative hypovolemia. We realized our blood volume classification algorithm by (i) extracting a multitude of features from multi-modal physiological signals including the electrocardiogram (ECG), the seismocardiogram (SCG), the ballistocardiogram (BCG), and the photoplethysmogram (PPG), (ii) constructing two ML classifiers using the features, one to classify normovolemia vs. hypovolemia and the other to classify hypovolemia into absolute hypovolemia and relative hypovolemia, and (iii) sequentially integrating the two to enable multi-class classification (normovolemia, absolute hypovolemia, and relative hypovolemia). We developed the blood volume decompensation state classification algorithm using the experimental data collected from six animals undergoing normovolemia, relative hypovolemia, and absolute hypovolemia challenges. Leave-one-subject-out analysis showed that our classification algorithm achieved an F1 score and accuracy of (i) 0.93 and 0.89 in classifying normovolemia vs. hypovolemia, (ii) 0.88 and 0.89 in classifying hypovolemia into absolute hypovolemia and relative hypovolemia, and (iii) 0.77 and 0.81 in classifying the overall blood volume decompensation state. The analysis of the features embedded in the ML classifiers indicated that many features are physiologically plausible, and that multi-modal SCG-BCG fusion may play an important role in achieving good blood volume classification efficacy. Our work may complement existing computational algorithms to estimate blood volume compensatory reserve as a potential decision-support tool to provide guidance on context-sensitive hypovolemia therapeutic strategy.https://www.mdpi.com/1424-8220/22/4/1336hypovolemiablood volumemachine learningseismocardiogramballistocardiogramwearables |
| spellingShingle | Yekanth Ram Chalumuri Jacob P. Kimball Azin Mousavi Jonathan S. Zia Christopher Rolfes Jesse D. Parreira Omer T. Inan Jin-Oh Hahn Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals Sensors hypovolemia blood volume machine learning seismocardiogram ballistocardiogram wearables |
| title | Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals |
| title_full | Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals |
| title_fullStr | Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals |
| title_full_unstemmed | Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals |
| title_short | Classification of Blood Volume Decompensation State via Machine Learning Analysis of Multi-Modal Wearable-Compatible Physiological Signals |
| title_sort | classification of blood volume decompensation state via machine learning analysis of multi modal wearable compatible physiological signals |
| topic | hypovolemia blood volume machine learning seismocardiogram ballistocardiogram wearables |
| url | https://www.mdpi.com/1424-8220/22/4/1336 |
| work_keys_str_mv | AT yekanthramchalumuri classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT jacobpkimball classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT azinmousavi classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT jonathanszia classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT christopherrolfes classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT jessedparreira classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT omertinan classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals AT jinohhahn classificationofbloodvolumedecompensationstateviamachinelearninganalysisofmultimodalwearablecompatiblephysiologicalsignals |