Crying is the only means of communication for a newborn baby with its surrounding environment, but it also provides significant information about the newborn’s health, emotions, and needs. The cries of newborn babies have long been known as a biomarker for the diagnosis of pathologies. However, to the best of our knowledge, exploring the discrimination of two pathology groups by means of cry signals is unprecedented. Therefore, this study aimed to identify septic newborns with Neonatal Respiratory Distress Syndrome (RDS) by employing the Machine Learning (ML) methods of Multilayer Perceptron (MLP) and Support Vector Machine (SVM). Furthermore, the cry signal was analyzed from the following two different perspectives: 1) the musical perspective by studying the spectral feature set of Harmonic Ratio (HR), and 2) the speech processing perspective using the short-term feature set of Gammatone Frequency Cepstral Coefficients (GFCCs). In order to assess the role of employing features from both short-term and spectral modalities in distinguishing the two pathology groups, they were fused in one feature set named the combined features. The hyperparameters (HPs) of the implemented ML approaches were fine-tuned to fit each experiment. Finally, by normalizing and fusing the features originating from the two modalities, the overall performance of the proposed design was improved across all evaluation measures, achieving accuracies of 92.49% and 95.3% by the MLP and SVM classifiers, respectively. The MLP classifier was outperformed in terms of all evaluation measures presented in this study, except for the Area Under Curve of Receiver Operator Characteristics (AUC-ROC), which signifies the ability of the proposed design in class separation. The achieved results highlighted the role of combining features from different levels and modalities for a more powerful analysis of the cry signals, as well as including a neural network (NN)-based classifier. Consequently, attaining a 95.3% accuracy for the separation of two entangled pathology groups of RDS and sepsis elucidated the promising potential for further studies with larger datasets and more pathology groups.

No abstract

This paper addresses the problem of automatic cry signal segmentation for the purposes of infant cry analysis. The main goal is to automatically detect expiratory and inspiratory phases from recorded cry signals. The approach used in this paper is made up of three stages: signal decomposition, features extraction, and classification. In the first stage, short-time Fourier transform, empirical mode decomposition (EMD), and wavelet packet transform have been considered. In the second stage, various set of features have been extracted, and in the third stage, two supervised learning methods, Gaussian mixture models and hidden Markov models, with four and five states, have been discussed as well. The main goal of this work is to investigate the EMD performance and to compare it with the other standard decomposition techniques. A combination of two and three intrinsic mode functions (IMFs) that resulted from EMD has been used to represent cry signal. The performance of nine different segmentation systems has been evaluated. The experiments for each system have been repeated several times with different training and testing datasets, randomly chosen using a 10-fold cross-validation procedure. The lowest global classification error rates of around 8.9% and 11.06% have been achieved using a Gaussian mixture models classifier and a hidden Markov models classifier, respectively. Among all IMF combinations, the winner combination is IMF3+IMF4+IMF5.

An analysis of newborn cry signals, either for the early diagnosis of neonatal health problems or to determine the category of a cry (e.g., pain, discomfort, birth cry, and fear), requires a primary and preliminary preprocessing step to quantify the important expiratory and inspiratory parts of the audio recordings of newborn cries. Data typically contain clean cries interspersed with sections of other sounds (generally, the sounds of speech, noise, or medical equipment) or silence. The purpose of signal segmentation is to differentiate the important acoustic parts of the cry recordings from the unimportant acoustic activities that compose the audio signals. This paper reports on our research to establish an automatic segmentation system for newborn cry recordings based on Hidden Markov Models using the HTK (Hidden Markov Model Toolkit). The system presented in this report is able to detect the two basic constituents of a cry, which are the audible expiratory and inspiratory parts, using a two-stage recognition architecture. The system is trained and tested on a real database collected from normal and pathological newborns. The experimental results indicate that the system yields accuracies of up to 83.79%. extracted from the cries and the health problems of the child [2][3][4][5]. Various studies are currently under way to devise a tool that analyzes cries automatically, to diagnose neonatal pathologies [6][7][8].We are involved in the design of an automatic system for early diagnosis, called the Newborn Cry-based Diagnostic System (NCDS), which can detect certain pathologies in newborns at an early stage. The implementation of this system requires a database containing hundreds of cry signals. The overwhelming problem that arises when working with such a database is the diversity of acoustic activities that compose the audio recordings, such as background noise, speech, the sound of medical equipment and silence. Such diversity could harm the analysis process, as the presence of any acoustic component other than the cry itself could result in the misclassification of pathologies by reducing the NCDS system performance. This is because the NCDS would decode every segment of the recording signal, whether it is part of a cry or not. In this case, unwanted segment insertion in essential crying segments would lengthen the process of classification unnecessarily and leave the system prone to error. An important subtask of the NCDS is the manipulation of the newborn cry sound, and what is needed to perform this subtask is a segmentation system. Until now, few works have been carried out in this area. In this paper, we propose an automatic segmentation module designed to isolate the audible expiration and inspiration parts of cry sounds to serve as a preprocessing step of our NCDS. The rest of this paper is organized as follows: Related work is presented in section 2. The HMM and the HTK are reviewed briefly in section 3. The training corpus and the testing corpus are described in section 4. In section 5, the architect...