A Spectral-Based Approach for BCG Signal Content Classification

Abstract

This paper has two objectives: the first is to generate two binary flags to indicate useful frames permitting the measurement of cardiac and respiratory rates from Ballistocardiogram (BCG) signals—in fact, human body activities during measurements can disturb the BCG signal content, leading to difficulties in vital sign measurement; the second objective is to achieve refined BCG signal segmentation according to these activities. The proposed framework makes use of two approaches: an unsupervised classification based on the Gaussian Mixture Model (GMM) and a supervised classification based on K-Nearest Neighbors (KNN). Both of these approaches consider two spectral features, namely the Spectral Flatness Measure (SFM) and Spectral Centroid (SC), determined during the feature extraction step. Unsupervised classification is used to explore the content of the BCG signals, justifying the existence of different classes and permitting the definition of useful hyper-parameters for effective segmentation. In contrast, the considered supervised classification approach aims to determine if the BCG signal content allows the measurement of the heart rate (HR) and the respiratory rate (RR) or not. Furthermore, two levels of supervised classification are used to classify human-body activities into many realistic classes from the BCG signal (e.g., coughing, holding breath, air expiration, movement, et al.). The first one considers frame-by-frame classification, while the second one, aiming to boost the segmentation performance, transforms the frame-by-frame SFM and SC features into temporal series which track the temporal variation of the measures of the BCG signal. The proposed approach constitutes a novelty in this field and represents a powerful method to segment BCG signals according to human body activities, resulting in an accuracy of 94.6%.

New End-to-End Strategy Based on DeepLabv3+ Semantic Segmentation for Human Head Detection

“…Deep learning models have seen significant success in a wide range of applications in recent years. In fact, different deep learning methods have been introduced for image and motion segmentation [10][11][12][13], detection [14][15][16], tracking [17,18] and classification [19,20]. Due to the success of deep learning methods, CNN models have also been used in the literature to provide relevant information on the number of people present on the stage.…”

Heart rate estimation from ballistocardiogram signals processing via low-cost telemedicine architectures: a comparative performance evaluation

“…Therefore, rising interest and novel instrumentation are bringing to the development of novel signal processing techniques focused on both BCG signal modeling (58) and features extraction-Deep Learning, see e.g. (59).…”

Investigating Cardiorespiratory Interaction Using Ballistocardiography and Seismocardiography—A Narrative Review

“…The only way to measure BCG in the early studies was to measure body displacement by a suspended bed that was free to move [ 3 , 4 ] or a table strongly coupled to its zero position [ 59 ]. In the last few decades, with the revival of the BCG technology [ 8 ], these large BCG sensors have been reduced in size to the point where they could be placed on or under a mattress [ 68 , 69 , 70 , 71 , 72 ], on a wheelchair [ 73 ], on the skin as a wearable device [ 17 , 74 , 75 ], or even record body movements in a contactless way [ 76 ] (see Table 2 ). Each of these approaches has its own advantages and can be chosen depending on the desired function.…”

“…To evaluate sleep-related disorders (see Section 5.1.1 ) more comfortably than with traditional polysomnography, BCG has been shown to have a great potential, as it does not disrupt a subject’s normal sleep patterns to gather data [ 8 ]. Bed-based BCG can benefit from different kinds of sensors, from static charge sensitive beds [ 78 , 79 , 80 ] to pressure pads or pressure sensors [ 70 , 81 , 82 ], force sensors [ 69 , 83 , 84 ], loadcells in bed legs [ 85 , 86 , 87 , 88 , 89 ], fiber optic sensors [ 71 , 72 , 90 ], or electromagnetic film sensors [ 68 , 91 ]. These systems can offer comfort to the user, as well as the opportunity to easily install the systems at home [ 30 ].…”

“…Their results showed 96.37% accuracy, 92.46% sensitivity, and 98.11% specificity for 20 healthy subjects. In another study, Ben Nasr et al [ 71 ] classified the BCG signals in phases of normal breathing, coughing, post-coughing, breath hold, expiration, movement, and others by a Gaussian mixture model and K-nearest-neighbor methods. For this purpose, they extracted spectral flatness based on Wiener entropy and spectral centroid.…”