A Generic Deep Learning Based Cough Analysis System From Clinically Validated Samples for Point-of-Need Covid-19 Test and Severity Levels.

In an attempt to reduce the infection rate of the COrona VIrus Disease-19 (Covid-19) countries around the world have echoed the exigency for an economical, accessible, point-of-need diagnostic test to identify Covid-19 carriers so that they (individuals who test positive) can be advised to self isolate rather than the entire community. Availability of a quick turn-around time diagnostic test would essentially mean that life, in general, can return to normality-at-large. In this regards, studies concurrent in time with ours have investigated different respiratory sounds, including cough, to recognise potential Covid-19 carriers. However, these studies lack clinical control and rely on Internet users confirming their test results in a web questionnaire (crowdsourcing) thus rendering their analysis inadequate. We seek to evaluate the detection performance of a primary screening tool of Covid-19 solely based on the cough sound from 8,380 clinically validated samples with laboratory molecular-test (2,339 Covid-19 positive and 6,041 Covid-19 negative) under quantitative RT-PCR (qRT-PCR) from certified laboratories. All collected samples were clinically labelled, i.e., Covid-19 positive or negative, according to the results in addition to the disease severity based on the qRT-PCR threshold cycle (Ct) and lymphocytes count from the patients. Our proposed generic method is an algorithm based on Empirical Mode Decomposition (EMD) for cough sound detection with subsequent classification based on a tensor of audio sonographs and deep artificial neural network classifier with convolutional layers called 'DeepCough'. Two different versions of DeepCough based on the number of tensor dimensions, i.e., DeepCough2D and DeepCough3D, have been investigated. These methods have been deployed in a multi-platform prototype web-app 'CoughDetect'. Covid-19 recognition results rates achieved a promising AUC (Area Under Curve) of [Formula: see text] 98 . 80 % ± 0 . 83 % , sensitivity of [Formula: see text] 96 . 43 % ± 1 . 85 % , and specificity of [Formula: see text] 96 . 20 % ± 1 . 74 % and average AUC of [Formula: see text] 81 . 08 % ± 5 . 05 % for the recognition of three severity levels. Our proposed web tool as a point-of-need primary diagnostic test for Covid-19 facilitates the rapid detection of the infection. We believe it has the potential to significantly hamper the Covid-19 pandemic across the world.

A novel event-related fMRI supervoxels-based representation and its application to schizophrenia diagnosis.

BACKGROUND AND OBJECTIVE: The schizophrenia diagnosis represents a difficult task because of the confusing descriptions of symptoms given by the patient, their similarity among several disorders, the lower familiarity with genetic predisposition, and the probably inadequate response to the treatment. Neuro-biological markers of schizophrenia, as a quantitative relationship between the psychiatrist's reports and the biology of the brain, could be used. Functional Magnetic Resonance Imaging (fMRI) obtain the subject's performance in cognitive tasks and may find significant differences between the patient's data and controls. The input data of classifiers may imply alterations in diagnosis; therefore, it is essential to ensure an adequate representation to describe the entire dataset classified. METHODS: We propose a supervoxels-based representation calculated by two main steps: the short-range connectivity, supervoxels' generation using a Fuzzy Iterative Clustering algorithm, and the long-range connectivity, employing Detrended Cross-Correlation Analysis among supervoxels. The unrelated supervoxels, through a statistical test based on critical points calculated empirically, are removed. The remainder supervoxels are the input for feature selectors to extract the discriminative supervoxels. We implement support vector machine classifiers using the correlation coefficient of the significant supervoxels. The dataset of 1.5 Tesla was downloaded from the SchizConnect site, where the fMRI data, during an auditory oddball task, was acquired. We calculate the performance of the classifiers using a leave-one-out cross-validation and compute the area under the Receiver Operating Characteristic curve and a permutation test to ensure no bias in the classifiers. RESULTS: According to the permutation test, with p-values less than the significance level of 0.05, the classifiers extract discriminative class structure from data where no bias is shown. Our supervoxels-based representation gets the maximum values of sensitivity, specificity, and accuracy of 92.9%, 100%, and 96.4%, respectively. The discriminative brain regions, to discern among patients and controls, are extracted; these regions also are mentioned by the related works. CONCLUSIONS: The proposed representation, based on supervoxels, is a data-driven model that does not use predefined models of the signal nor pre-relocated brain regions of interest. The results are competitive against the related works, and the relevant supervoxels are related to the schizophrenia diagnosis.

Minimally invasive approach in radio distal end fractures, three different types of incisions.

Two techniques are described to treat distal radius fractures: conventional (Henry approach) and minimally invasive plate osteosynthesis. The latter technique has been described by different authors such as Imatani et al. and Zenke et al. This was a descriptive retrospective study, analyzing 26 adult patients with unstable distal radius fracture, extra-articular type A or partially intra-articular type B according to AO. The approaches used were: (i) single longitudinal palmar incision; (ii) double T-incision (horizontal and vertical) and (iii) double position II. Ages were between 21 and 78 years. Most affected hand was the right. Most common fracture was 23B2 (AO classification). In total, 84.6% of the patients did not present complications. According to the functional evaluation of the wrist by the Mayo Clinic, 31% showed excellent results, 42% showed good results, 27% showed fair results. The techniques had satisfactory results for the osteosynthesis with more aesthetic and less invasive approach.

Analysis of facial expressions in parkinson's disease through video-based automatic methods.

BACKGROUND: The automatic analysis of facial expressions is an evolving field that finds several clinical applications. One of these applications is the study of facial bradykinesia in Parkinson's disease (PD), which is a major motor sign of this neurodegenerative illness. Facial bradykinesia consists in the reduction/loss of facial movements and emotional facial expressions called hypomimia. NEW METHOD: In this work we propose an automatic method for studying facial expressions in PD patients relying on video-based METHODS: 17 Parkinsonian patients and 17 healthy control subjects were asked to show basic facial expressions, upon request of the clinician and after the imitation of a visual cue on a screen. Through an existing face tracker, the Euclidean distance of the facial model from a neutral baseline was computed in order to quantify the changes in facial expressivity during the tasks. Moreover, an automatic facial expressions recognition algorithm was trained in order to study how PD expressions differed from the standard expressions. RESULTS: Results show that control subjects reported on average higher distances than PD patients along the tasks. COMPARISON WITH EXISTING METHODS: This confirms that control subjects show larger movements during both posed and imitated facial expressions. Moreover, our results demonstrate that anger and disgust are the two most impaired expressions in PD patients. CONCLUSIONS: Contactless video-based systems can be important techniques for analyzing facial expressions also in rehabilitation, in particular speech therapy, where patients could get a definite advantage from a real-time feedback about the proper facial expressions/movements to perform.

Application of Pattern Recognition Techniques to the Classification of Full-Term and Preterm Infant Cry.

OBJECTIVES: Scientific and clinical advances in perinatology and neonatology have enhanced the chances of survival of preterm and very low weight neonates. Infant cry analysis is a suitable noninvasive complementary tool to assess the neurologic state of infants particularly important in the case of preterm neonates. This article aims at exploiting differences between full-term and preterm infant cry with robust automatic acoustical analysis and data mining techniques. STUDY DESIGN: Twenty-two acoustical parameters are estimated in more than 3000 cry units from cry recordings of 28 full-term and 10 preterm newborns. METHODS: Feature extraction is performed through the BioVoice dedicated software tool, developed at the Biomedical Engineering Lab, University of Firenze, Italy. Classification and pattern recognition is based on genetic algorithms for the selection of the best attributes. Training is performed comparing four classifiers: Logistic Curve, Multilayer Perceptron, Support Vector Machine, and Random Forest and three different testing options: full training set, 10-fold cross-validation, and 66% split. RESULTS: Results show that the best feature set is made up by 10 parameters capable to assess differences between preterm and full-term newborns with about 87% of accuracy. Best results are obtained with the Random Forest method (receiver operating characteristic area, 0.94). CONCLUSIONS: These 10 cry features might convey important additional information to assist the clinical specialist in the diagnosis and follow-up of possible delays or disorders in the neurologic development due to premature birth in this extremely vulnerable population of patients. The proposed approach is a first step toward an automatic infant cry recognition system for fast and proper identification of risk in preterm babies.

Gyroscope-driven mouse pointer with an EMOTIV® EEG headset and data analysis based on Empirical Mode Decomposition.

This paper presents a project on the development of a cursor control emulating the typical operations of a computer-mouse, using gyroscope and eye-blinking electromyographic signals which are obtained through a commercial 16-electrode wireless headset, recently released by Emotiv. The cursor position is controlled using information from a gyroscope included in the headset. The clicks are generated through the user's blinking with an adequate detection procedure based on the spectral-like technique called Empirical Mode Decomposition (EMD). EMD is proposed as a simple and quick computational tool, yet effective, aimed to artifact reduction from head movements as well as a method to detect blinking signals for mouse control. Kalman filter is used as state estimator for mouse position control and jitter removal. The detection rate obtained in average was 94.9%. Experimental setup and some obtained results are presented.

Spectrographic cry analysis in newborns with profound hearing loss and perinatal high-risk newborns.

BACKGROUND: Crying is the first neurophysiological demonstration of the newborn. The acoustic analysis of crying episodes can provide useful information in the early diagnosis of several pathologies. METHODS: We carried out a spectrographic cry analysis of 40 infants with several diseases such as neonatal asphyxia, breathing disorders, deafness and neurological disorders. Statistical t-test was used in order to compare means of fundamental frequencies (F0) of cries among pathologies, as well as contingency tables for qualitative variables and melodic form. RESULTS: No significant differences of F0 were found by applying the t-test among pathologies. Time of inspiration in breathing disorders is higher compared with other groups. Infants with deafness and neurological disorders present poor melodic forms. Qualitative analysis revealed more often glottal roll and vibrato. CONCLUSIONS: Cry analysis is clinically useful for early detection of diseases in newborns as a result of alterations in the central nervous system that are reflected in phonoarticulatory and respiratory functions for sound emission.

Análisis del llanto en niños hipoacúsicos y normoyentes de 0 a 2 años de edad

resumen está disponible en el texto completo

Abstract: Infant crying is a complex phenomenon that implies several functions: breathing, action of laryngeal and supralaryngeal muscles under the control of the neurovegetative systems of the brainstem, and the limbic system, and the association of cortical areas and the cerebellum. Although it is a communication system different to babbling and language, it is related with the future development of phonation. Cry analysis provides information about the neuro-physiologic and psychological states of newborns and the identification of perinatal abnormalities. It is necessary to discuss the subject extensively because there are new data on situations such as laringomalacia, congenital hypothyroidism, deafness and sleep apnea that seem to be associated to infantile crying behaviors. Infant cries can be analyzed as behavioral conditions (hunger, anger and pain cries) allows knowing of mother-child relationship or the effect under diverse cultural conditions, such as stress, emo-tional deprivation or illness. A spectrographic analysis of the cries may identify several characteristics: threshold, latency, duration of phonation, maximum and minimum of the fundamental frequency (F0), occurrence and maximum pitch of shift, gliding, melody, biphonation, bifurcation, noise concentration, quality of the voice, double harmonic break, glottal plosives, vibratos, melody types, F0 stability and inspiratory stridor. To date, it has not been possible to establish alteration patterns. The best studied variables are F0, its harmonics and the duration of each emission; it is accepted that F0 varies between 400 and 600 Hz, during 1.4 ± 0.6s. Under such approaches, diverse alterations and risk factors have been studied: congenital alterations, malnutrition, sudden death, maternal exposition to drugs, prematurely born babies or perinatal asphyxia and disturbances of the central nervous system. Authors have reported F0 equal or less than 300 Hz in cases of sudden death or with high frequencies, near the 1000 Hz in the Cri du chat syndrome, perinatal asphyxia and other cases who died suddenly. During the cry, there is an increase of intra-abdominal pressure, heart rate and blood pressure, reduction of oxygen saturation, increase of the intra cranial-pressure, beginning of stress reactions, depletion of the energy anf oxygen reserves, such as the found in the Valsalva's maneuver. Every event of prolonged cries implies alteration of the breathing control like a Hering-Breuer reflex. Considering that some authors have proposed early vocalizations are a good predictor of deafness, in a previous paper we reported the characteristics of the cry of 20 deaf neonates. However, we were not able to demonstrate differences when comparing them with normal hearing neonates and infants, using only parametric methods. Still, we decided to go further and investgate the quality of infant cries of deaf neonates and infants. Material and methods. Twenty zero-to two-year old cases were studied; they were deaf children of both sexes; all cases were included in a follow-up program on the Human Communication Department of the National Institute of Perinatology of Mexico and were compared with 20 normal hearing children. We re-corded Brain Stem Evoked Auditory Responses (BEAR) and cry recording using a digital Sony recorder during the physical exploration. We analyzed the frequency (Hz) and duration of the espiratory cries, the duration of inspiration between two cry emissions and the characteristics of the spectrogram. Quantitative analysis. The usual estimates of means and standard variation were obtained and they were compared with one way analysis of variance. We organized typologies of frequency by means of cluster techniques (Ward method). The distribution of the duration of the periods of crying and silence was explored with a contingency tables. Qualitative analysis. Two standardized observers visually analyzed all the cries to determine any variation of F0 and of harmonic frequencies. Whenever a variation of F0 was observed, we obtained maximum and minimum frequencies, as well as average duration of each cry emission. The procedure was validated by means of the graphic comparison with a Fouries analysis. Results. Mean duration of cries in the deaf group was 0.5845 ± 0.6150 s (range 0.08-5.2 s), while in the group of normal hearing cases was 0.5387 ± 0.2631 (range 0.06-1.75 s). From the deaf group, five cases had very prolonged duration of cries, without statistical significance. The mean duration of the inspiration was 0.3962 ± 0.2326, with a range of 0.06 to 1.75 in the deaf group and of 0.4083 ± 0.1854, with a range of 0.21 at 0.96, in the controls, without difference among groups. There was no correlation between the time of espiratory cry and that of the inspiration. Three cry topologies were organized: one of shorter duration (mean 0.30 s), with 111 spectrograms, an intermediate one (mean 0.73) with 85 spectrograms and one of prolonged duration (mean 4.5 s) with spectrograms of three cases. Three topologies of the inspiratory period were obtained: one of short periods (mean 0.33 s), with 171 spectrograms, one of intermediate duration (mean 0.80 s) with 18 spectrograms and one of prolonged duration (mean 1.60 s) with three cases. There were no statistical differences of tipologies between the deaf groups and normal hearing cases. On the qualitative analysis of cries, we came across several variations which are interpreted as abnormalities: vibratos, poor melodic control, loss of fundamental frequencies, harmonic limited production, plosives, gliding, bi phonation, and a loss of intensity at end of cry emissions. These changes were also observed on the control cases, but only in a very limited number. Discussion. Cry spectrogram analysis are non invasive indicators of the neonate's neurophysiologic organization. Although cry duration varies in healthy newborns, the accepted variation for a normal range is 1.1 to 2.8 s, with standard deviations around 0.6 s. Consistent differences have not been demonstrated between risk and control groups. However, abnormal cases such as Down syndrome or severe asphyxia have very short cries, whereas on the Cri du chat syndrome the duration of cries is prolonged. Extended cries imply cardiac and respiratory risks which have been associated with later outcomes as development retardation and sudden death. There are also some questions to solve, such as the regulation and control of cry, starting from breathing mechanisms or from a sensorial afferent, mediated by hearing. The deaf infants are constituted in a study model, considering that the auditory afference is suppressed and the control of the cry is restricted to the breathing environment. In the studied spectrograms, the duration of the cry was within reported normal limits by other authors, inasmuch in the normal hearing control cases as in the deaf, except the dissident cases, but without these reaching statistical significance. Further research of brainstem function is needed for the abnormal cases with prolonged cry periods, since such cries are interpreted as an alteration of the breathing reflexes of Hering-Breuer, which might have a pathological meaning in the sense of the sob's spasm or even more severe risk factors as sleep apnea and even sudden death. The qualitative analysis in the deaf individuals demonstrated a poor quality and unstable character of melodic control, with a smaller number of harmonics. The deaf cases lost the relationship between the fundamental frequencies and their harmonics, mainly because of the participation of supraglottic structures that modulate pitch and due to the poorness of melodic control, either for monotony or due to the impossibility of returning to a normal pattern, following variations such as vibrato, plosives or noise concentration. In the cases of prolonged cries, starting from the third second, the sound intensity tends to diminish and the harmonics are lost, perhaps due to a decrease of the subglottal pressure of phonation. This finding supports the auditory control of crying related to breathing mechanisms. Conclusions. In preliminary terms, by means of the melodic analysis of the spectrograms, differences are demonstrated be-tween the cries of the deaf and of the normal hearing cases. The increase of the complexity of the melody of the cry, or their poverty, are indicative of the neuromuscular function and they may support the evaluation of phonation before language development. The study of the spectrograms of deaf individuals does not constitute an element for the detection or for diagnosis since, to date, estimators of sensibility or of specificity have not been established, but they constitute a support for its integral evaluation, with the possibility of evaluating and of improving therapeutic rehabilitation.

[Cry analysis of hypoacoustic children and normal hearing children]. / Análisis del llanto del niño hipoacúsico y del niño normo-oyente.

OBJECTIVE: We analyzed the spectrograms of the cry of deaf children and compared the results with those of normal children. MATERIAL AND METHODS: Twenty deaf children (0-2 years of age, both sexes) and 20 normal hearing children of both sexes within the same age range were studied. The deaf cases were selected from patients with high-risk pregnancies who are followed up at the National Institute of Perinatology (INPer), Department of Human Communication. Brainstem evoked potential responses were carried out in order to corroborate hearing loss in the affected group and to verify normal hearing in the other group. During physical examination in the Human Communication Clinic of the Inper, the cry was recorded with a Fisher tape recorder and MK2 unidirectional microphone. Sonograms were analyzed in the "Laboratorio de Bioacústica" of the Institute of the Human Communication (CNR) using the Software Cool Edit 96. We registered the fundamental frequency and 6 harmonic F1 at F6 (Hz). The cry frequencies of the deaf children were compared and of the normal hearing children using one-way analysis of variance. A typology of frequencies was organized by means of cluster analysis by the method of Ward and was compared with the study groups by means of contingency analysis (chi square). RESULTS: No significant difference was demonstrated in the fundamental frequency or in the harmonic by means of the one-way analysis. With cluster analysis, 8 types were obtained that were not different when compared with the study groups. The hypothesis was rejected with the obtained data. CONCLUSIONS: Upon confirmation of the obtained results, analysis of the spectrogram of deaf children at early ages does not have the ability to identify the problem and thus cannot be used as a method for early detection of deafness.