This work shows the behavioral and heart condition monitoring of the patients during the COVID-19 infection. As the COVID-19 infection spreads very easily through contact, this work introduces a noninvasive technique to monitor such patients. The heart condition monitoring can be done through the analysis of acousticalcardiogram (ACG) as well as the behavioral changes observed through the Formant analysis. The speech of the patient is recorded, digitized, and its analysis is done using the PRAAT software. The required information from the speech samples is extracted and subjected to the analysis. The spectrogram of each utterance is plotted, and its first and second formants are analysed to form the vowel triangle. These vowel triangle provides behavioral monitoring. The monitoring of heart condition can be done using ACG. The Formant analysis of utterance of 12 Hindi consonants provides the assessment technique of the patient heart condition.
TopIntroduction
Since speech reflects essential information regarding human emotion, it can be extracted in the form of speech features from the recorded speech samples by applying various speech processing techniques. Multiple analysis approaches are applied to these retrieved speech features. Acoustic vowel triangle analysis can be used to detect and predict a patient's emotions and behavior. Recently, special attention has been brought to speech as a modality to automatically deduct information on human emotions and behavioral prediction (A. Deshpande et.al.,2014; G.Fant, 1960; J.Benesty,2008). Emotion detection and classification play an essential role in user-friendly human-machine interaction. Human speech is influenced by a variety of factors, including posture. As a result, a human can be said to communicate using his breathing system, bodily muscles, and head. Speech is produced as a result of physical and vocal efforts. Speech processing techniques are being developed to provide a non-invasive, less expensive, and speedier method for assessing and diagnosing physiological conditions and behavioral abnormalities during COVID-19 infection.
The required information from the speech samples is extracted and subjected to the analysis. The spectrogram of each utterance is plotted and its first and second formants are analyzed to form the vowel triangle (Berstein A. et.al.,1985; E. Dmitrity et.al., 2009, L.Rabiner et.al., 1978; R.F. Orlikoff et.al., 1989). Various parameters like Euclidean distance, area of the vowel triangle, length of the base of the triangle, size of the vowel triangle, inclination of the triangle is used to predict emotional state of the persons. The stress, fear, anger, happiness, irritation is some of the common emotions which vary day by day in the COVID patients. Speech samples of some COVID patients has been recorded regularly for fourteen days daily and subjected to speech signal analysis. From these recorded speech samples, the F1 and F2 formants of the three particular Hindi vowels “aa”, “ee” and “oo” have been determined to plot their vowel triangles (L.Rabiner et.al., 1978; R.F. Orlikoff et.al., 1989). These vowel triangles provide the behavioral information of patients.
Medical experts usually use an electrocardiogram (ECG) to monitor the heart condition. This ECG has various important parameters like heart rate, RR-duration or RR-cycle, PQ, QR, PQR, QRS duration, which are used to predict various heart diseases (K. Charles Friedberg,1966). The pattern of the various peaks is also used to diagnose many heart diseases. The patient's heart condition can be monitored using an acoustical cardiogram (ACG). Acoustical cardiogram (ACG) is a tool that can be an alternative to an electrocardiogram (ECG) for some cases. The speech signal processing techniques are used to extract the required information from the recorded speech sample of the COVID patients. The Formant analysis of utterance of twelve Hindi consonants' ka', 'kha', 'ga', 'gha', 'anga', 'cha', 'chha', 'ja', 'jha', 'yan', 'ta', 'tha' is used to determine the acoustical-cardiogram. Almost all heart parameters can be obtained using an acoustical cardiogram (ACG), in addition to those calculated using a regular electrocardiogram (ECG). This acoustical cardiogram (ACG) is used to determine the patient's cardiac status (A. Deshpande et.al.,2014; Nader Salari,2020). ACG can be used to forecast a variety of cardiac problems.