This paper presents an automated procedure for acquisition and analysis of BallistoCardioGraphy (BCG) traces. A tri-axial accelerometer and a microcontroller unit are used to record heart-induced recoil forces generated from a lying subject. The problem of BCG J-peak annotation is split into two sub-tasks: candidates extraction, based on a detection signal, and actual annotation, guided by subject-specific search windows. Such procedure is derived from an automatic calibration, which is carried out with no need of concurrent ElectroCardioGram (ECG) or user intervention. The algorithm also implements post-annotation checks for refinement of annotation, which effectively reduces the number of missed J-peaks. The impact of each algorithm phase is analyzed, assessing statistical significance of each step; finally, performance is optimized in a data-driven fashion. Results show that the proposed methodology is able to achieve high sensitivity and precision (the median score is 98.9% and 98.1%, respectively) in J-peak detection. The quality of J-peaks timing annotation is further demonstrated by a very low discrepancy between BCG and ECG HR estimates. Over all population, the standard deviation of such error was found to be approximately 6.56 ms, whereas the Mean Absolute Error just 4.7 ms (i.e. ≈1.18;Ts, where Ts = 4 ms is the sampling period). Such scores, indeed, improve over recent related literature.

This paper presents an unsupervised methodology to analyze SeismoCardioGram (SCG) signals. Starting from raw accelerometric data, heartbeat complexes are extracted and annotated, using a two-step procedure. An unsupervised calibration procedure is added to better adapt to different user patterns. Results show that the performance scores achieved by the proposed methodology improve over related literature: on average, 98.5% sensitivity and 98.6% precision are achieved in beat detection, whereas RMS (Root Mean Square) error in heartbeat interval estimation is as low as 4.6 ms. This allows SCG heartbeat complexes to be reliably extracted. Then, the morphological information of such waveforms is further processed by means of a modular Convolutional Variational AutoEncoder network, aiming at extracting compressed, meaningful representation. After unsupervised training, the VAE network is able to recognize different signal morphologies, associating each user to its specific patterns with high accuracy, as indicated by specific performance metrics (including adjusted random and mutual information score, completeness, and homogeneity). Finally, a Linear Model is used to interpret the results of clustering in the learned latent space, highlighting the impact of different VAE architectural parameters (i.e., number of stacked convolutional units and dimension of latent space).

This paper presents a fully automated procedure for acquiring and analyzing seismocardiographic (SCG) traces from an Inertial Measurement Unit (IMU) placed over a subject’s sternum. An automated calibration procedure allows for straightforward adaption to different subjects. Calibration is performed once per subject, exploiting ECG (electrocardiogram) markers; relevant patterns and parameters are automatically extracted and used for successive SCG processing, which does not require concurrent ECG information any longer. Annotation of SCG traces is performed in two steps: in the first one, a suitably engineered signal is derived from SCG and used as coarse heartbeat detector; then, annotation can be performed by comparing the prototype extracted at calibration time with segments of SCG data, near to the detected beats. The proposed methodology is validated by direct comparison with ECG, adopted as gold-standard. In particular, three main metrics are taken into account: sensitivity (i.e. percentage of correctly identified heartbeats, compared to ECG), precision (i.e. impact of false positives on truly detected beats) and R2 (i.e. linearity between beat-to-beat measurements as computed by ECG and SCG). Results show satisfactory performance, more than adequate to continuous, long-term monitoring: overall, approximately 90% of heartbeats are correctly detected, on average, with minimal false positives (≈1%). Linearity between ECG and SCG-computed beat-to-beat intervals is extremely high (R2 > 0.95, on average), indicating good agreement between the two measurement methods. These results suggest SCG can be used as a reliable, contactless measure of heart-related parameters.

This article presents an unsupervised, automated procedure for the analysis of SeismoCardioGram (SCG) signals. SCG is a measure of chest vibrations, induced by the mechanical activity of the heart, that allows to extract relevant parameters, including Heart Rate (HR) and HR Variability (HRV). An initial self-calibration is performed, solely based on SCG traces, yielding a suitable heartbeat template (personalized for each subject). Then, beat detection and timing annotation are performed in two steps: at first, candidate beats are identified and validated, by means of suitably defined detection signals; then, precise timing annotation is achieved by best aligning such candidate beats to the previously extracted template. The algorithm has been validated on two separate datasets, featuring different acquisition setups: the first one is the publicly available CEBS database, reporting SCG signals from subjects lying in supine position, whereas the second one was acquired using a custom setup, involving sitting subjects. Results show good sensitivity and precision scores (98.5%, 98.6% for the CEBS database, and 99.1%, 97.9% for the Custom one, respectively). Also, comparison with ECG gold-standard is given, showing good agreement between beat-to-beat intervals computed from SCG and the ECG gold-standard: on average, R2 scores of 99.3% and 98.4% are achieved on CEBS and Custom datasets, respectively. Furthermore, a low RMS Error is achieved on the CEBS and Custom dataset, amounting to 4.6 ms and 6.2 ms, respectively (i.e. 2.3 Ts and 3.1 Ts, where Ts is the sampling period): such results well compare to related literature. Validation on two different datasets indicates the robustness of the proposed methodology.

This work presents a methodology to analyze and segment both seismocardiogram (SCG) and ballistocardiogram (BCG) signals in a unified fashion. An unsupervised approach is followed to extract a template of SCG/BCG heartbeats, which is then used to fine-tune temporal waveform annotation. Rigorous performance assessment is conducted in terms of sensitivity, precision, Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) of annotation. The methodology is tested on four independent datasets, covering different measurement setups and time resolutions. A wide application range is therefore explored, which better characterizes the robustness and generality of the method with respect to a single dataset. Overall, sensitivity and precision scores are uniform across all datasets ( $p>0.05$ from the Kruskal–Wallis test): the average sensitivity among datasets is 98.7%, with 98.2% precision. On the other hand, a slight yet significant difference in RMSE and MAE scores was found ( $p<0.01$ ) in favor of datasets with higher sampling frequency. The best RMSE scores for SCG and BCG are 4.5 and 4.8 ms, respectively; similarly, the best MAE scores are 3.3 and 3.6 ms. The results were compared to relevant recent literature and are found to improve both detection performance and temporal annotation errors.

The lack of physical exercise is among the most relevant factors in developing health issues, and strategies to incentivize active lifestyles are key to preventing these issues. The PLEINAIR project developed a framework for creating outdoor park equipment, exploiting the IoT paradigm to build “Outdoor Smart Objects” (OSO) for making physical activity more appealing and rewarding to a broad range of users, regardless of their age and fitness. This paper presents the design and implementation of a prominent demonstrator of the OSO concept, consisting of a smart, sensitive flooring, based on anti-trauma floors commonly found in kids playgrounds. The floor is equipped with pressure sensors (piezoresistors) and visual feedback (LED-strips), to offer an enhanced, interactive and personalized user experience. OSOs exploit distributed intelligence and are connected to the Cloud infrastructure by using a MQTT protocol; apps have then been developed for interacting with the PLEINAIR system. Although simple in its general concept, several challenges must be faced, related to the application range (which called for high pressure sensitivity) and the scalability of the approach (requiring to implement a hierarchical system architecture). Some prototypes were fabricated and tested in a public environment, providing positive feedback to both the technical design and the concept validation.