Mitchell Silva

Cough sound description in relation to respiratory diseases in dairy calves.

Cough can be a biomarker in case of respiratory diseases. By monitoring and analyzing cough sounds through automatic devices, the farmer can obtain an early warning about a developing outbreak of respiratory infections. Cough sounds can be characterized by particular acoustic features (amplitude, frequency and duration) that are obtained by sound recording, labeling and analytic procedures. Based on these features, it might be possible to develop an automated cough recognition system. The aim of the study described in this paper is to investigate whether it is possible to discriminate cough sounds from other frequently occurring sounds in youngstock stables. Nasal swabs and blood were taken to identify the microbiological agents responsible for the respiratory problems. The collected cough sounds were compared to metal rack sounds, which are very common sounds in cattle farming, to identify acoustic differences between them. Results show that the length of cough sounds was significantly different from metal rack sounds (0.34 versus 0.37 s, P<0.05). Also, the analysis of amplitude and fundamental frequency showed significant differences between both types of sounds (resp. 0.21 and 0.18; 1326 and 3868 HZ). This indicates that it is possible discriminate cough sounds from other sounds and that cough sound can be used as a non-invasively diagnostic tool for respiratory diseases in youngstock groups.

Decoupling between fundamental frequency and energy envelope of neonate cries

BACKGROUND The presence of decoupling, i.e. the absence of coupling between fundamental frequency variation and energy envelope during phonetic crying, and its extent, reflects the degree of maturation of the central nervous system. AIM We hereby wanted to assess the existence and extent of decoupling in term neonates (neurodevelopmental relevance) and whether an association between decoupling and clinical pain expression could be unveiled (clinical relevance). STUDY DESIGN To assess decoupling in healthy term neonates during procedural pain, newborns were videotaped and crying was recorded during venous blood sampling. Besides acoustic analysis, pain expression was quantified based on the Modified Behavioral Pain Scale (MBPS). SUBJECTS 47 healthy term neonates underwent venous blood puncture at the 3rd day of life. OUTCOME MEASURES Beside the MBPS score, the correlation coefficients were calculated between the fundamental frequency variation and energy envelope of the cries. RESULTS Based on data collected in 47 healthy term neonates, correlation coefficients varied between 0.20 and 0.68. The degree of decoupling displayed extensive variability between the neonates and also in different cry bouts in a crying sequence within an individual neonate. A negative association was found between MBPS value and decoupling (r(2)=-0.12), the same as for the intra-subject variability although less extensive (r(2)=-0.02). CONCLUSION Decoupling only relates weakly with the amount of distress in 3day old newborns, even though a great intra-subject variability is present. This study suggests that there is no evidence of extensive decoupling as the newborn still has to fully develop the control of larynx and abdominal muscles.

Automated gait analysis in the open-field test for laboratory mice

In this article, an automated and accurate mouse observation method, based on a conventional test for motor function evaluation, is outlined. The proposed measurement technique was integrated in a regular open-field test, where the trajectory and locomotion of a free-moving mouse were measured simultaneously. The system setup consisted of a transparent cage and a camera placed below it with its lens pointing upward, allowing for images to be captured from underneath the cage while the mouse was walking on the transparent cage floor. Thus, additional information was obtained about the position of the limbs of the mice for gait reconstruction. In a first step, the camera was calibrated as soon as it was fixed in place. A linear calibration factor, relating distances in image coordinates to real-world dimensions, was determined. In a second step, the mouse was located and its body contour segmented from the image by subtracting a previously taken “background” image of the empty cage from the camera image. In a third step, the movement of the mouse was analyzed and its speed estimated from its location in the past few images. If the speed was above a 1-sec threshold, the mouse was recognized to be running, and the image was further processed for footprint recognition. In a fourth step, color filtering was applied within the recovered mouse region to measure the position of the mouse’s paws, which were visible in the image as small pink spots. Paws that were detected at the same location in a number of subsequent images were kept as footprints—that is, paws in contact with the cage floor. The footprints were classified by their position relative to the mouse’s outline as corresponding to the front left or right paw or the hind left or right paw. Finally, eight parameters were calculated from the footprint pattern to describe the locomotion of the mouse: right/left overlap, front/hind base, right/left front limb stride, and right/left hind limb stride. As an application, the system was tested using normal mice and mice displaying pentobarbital-induced ataxia. The footprint parameters measured using the proposed system showed differences of 10% to 20% between normal and ataxic mice.

Time-series analysis for online recognition and localization of sick pig (Sus scrofa) cough sounds

This paper considers the online localization of sick animals in pig houses. It presents an automated online recognition and localization procedure for sick pig cough sounds. The instantaneous energy of the signal is initially used to detect and extract individual sounds from a continuous recording and their duration is used as a preclassifier. Autoregression (AR) analysis is then employed to calculate an estimate of the sound signal, and the parameters of the estimated signal are subsequently evaluated to identify the sick cough sounds. It is shown that the distribution of just three AR parameters provides an adequate classifier for sick pig coughs. A localization technique based on the time difference of arrival is evaluated on field data and is shown that it is of acceptable accuracy for this particular application. The algorithm is applied on continuous recordings from a pig house to evaluate its effectiveness. The correct identification ratio ranged from 73% (27% false positive identifications) to 93% (7% false positive identifications) depending on the position of the microphone that was used for the recording. Although the false negative identifications are about 50% it is shown that this accuracy can be enough for the purpose of this tool. Finally, it is suggested that the presented application can be used to online monitor the welfare in a pig house, and provide early diagnosis of a cough hazard and faster treatment of sick animals.

Assessment of Pain Expression in Infant Cry Signals Using Empirical Mode Decomposition

BACKGROUND The presence of decoupling, i.e. the absence of coupling between fundamental frequency variation and intensity contour during phonetic crying, and its extent, reflects the degree of maturation of the central nervous system. OBJECTIVES The aim of this work was to evaluate whether Empirical Mode Decomposition (EMD) is a suitable technique for analyzing infant cries. We hereby wanted to assess the existence and extent of decoupling in term neonates and whether an association between decoupling (derived from EMD) and clinical pain expression could be unveiled. METHODS To assess decoupling in healthy term neonates during procedural pain, 24 newborns were videotaped and crying was recorded during venous blood sampling. Besides acoustic analysis, pain expression was quantified based on the Modified Behavioral Pain Scale (MBPS). Fundamental frequency and the intensity contour of the cry signals were extracted by applying the EMD to the data, and the correlation between the two was studied. RESULTS Based on data collected in healthy term neonates, correlation coefficients varied between 0.39 and 0.83. The degree of decoupling displayed extended variability between the neonates and also in different cry bouts in a crying sequence within an individual neonate. CONCLUSION When considering the individual ratio between the mean correlation of cry bouts during a crying sequence and their standard deviation, there seems to be a positive trend with increasing MBPS value. This might indicate that higher stressed subjects have less consistency in the investigated acoustic cry features, concluding that EMD has potential in the assessment of infant cry analysis.

Analysis of Cough Sounds for Diagnosis of Respiratory Infections in Intensive Pig Farming

Respiratory diseases are widespread causes of mortality and loss of productivity in intensive pig farming. Cough is one of the symptoms and a central element in screening and diagnosis of common illnesses caused, for example, by Pasteurella multocida or Actinobacillus pleuropneumoniae (App). The aim of this research is to compare the acoustic features of cough sounds originating from the mentioned infections and non-infectious cough sounds provoked by inhalation of citric acid by means of labeling and sound analysis. The acoustic parameters investigated are peak frequency and duration of the cough signals. The differences resulting from the sound analysis confirm the variability in acoustical parameters according to health status or disease in the animals. In infections, there is a change in the status of the respiratory system; consequently, infectious coughs are different from non-infectious coughs. The duration of single infectious coughs is considerably different among the types of cough analyzed, which are: non-infectious coughs, App coughs, and P. multocida coughs. Frequency analysis of single coughs allows a more general classification between non-infectious and infectious coughs. Acoustics parameters can be used in an algorithm-based alarm system to automatically identify cough sounds and provide farmers an early warning about the health status of their herds.

Acoustic Hatch Monitor for Egg Incubation: Detection of Internal Pipping in an Industrial Incubator

During the final phase of incubation, continuous recording and analysis of embryo sounds allow estimations about the percentage of hatched chickens. The objective of this study was to predict the time at which all eggs in an industrial incubator had passed the stage of internal pipping (IP100%). Chicken vocalization starts around day 19 when lung respiration commences. The amount of chicken vocalization increases during the hatching process, resulting in an increase of sound energy in the frequency range of 2500 to 3300 Hz. Using a standard microphone mounted inside the incubator, continuous sound recordings were processed to monitor the sound intensity during the hatching process. The method discussed in this article was assessed in an industrial-scale hatch setter (19,200 eggs) and repeated five times. The results show that the sound intensity signal results in a specific pattern that holds information about the stage of hatching. It was shown for five trials that the time at which automatic IP100% detection occurred was within ±3 h of the manually observed time for IP100%. The results imply that real-time sound recording and analysis during incubation allow prediction of the hatch outcome in industrial incubators. Such an acoustic monitoring system might provide industrial users with valuable information for hatch management.

A tool for labeling individual sounds from continuous recordings: An application to bio-acoustics

Bio-acoustics is a common tool for studying health, welfare and behavior of living organisms. Current practice consists of manually extracting and labelling individual sounds out of continuous recordings. This paper presents an application tool for speeding up bio-acoustical analysis. In particular, a semi-automatic tool has been developed that enables the extraction of individual sounds from continuous recordings. Although the approach can be applied to any kind of recordings, in the present study, the focus is on extracting individual pig sounds from recordings within a pig stable. More specifically, the sound extraction is based on the amplitude of the sound signal. The energy envelope of the signal is automatically extracted using the Hilbert transform on the discrete time signals. Subsequently, the individual sounds are presented to the user for manual labelling. The front-end provides the user with tools to adjust the length of the individual sounds since it is understood that in various cases a longer or shorter part of the signal may have been identified as an individual sound. The user also has an overview of the complete recording, the number of identified sounds, and the spread of the sounds within the continuous recording. Finally, the user is allowed to label using a predefined gamut of sounds or adding a sound category that has not been foreseen, making it a flexible tool. The labeling tool also allows the integration of an automatic sound classifier in the future and can be used for both training and/or validating algorithms under development. At a later stage the tool can be connected either online or offline to a web-based database that can allow for a collection of labeled sounds to be created.

Inspired fraction of carbon dioxide in oxygen supply to chronic pulmonary disease

Hypoxemic patients with chronic obstructive pulmonary disease (COPD) are at risk of carbon dioxide (CO(2)) retention during oxygen therapy and hypercapnia in COPD is associated with an ominous prognosis. Rebreathing with oxygen mask is possible in practice and possibly affects CO(2) retention due to an increased inspired fraction of CO(2). Its effects on arterial partial pressure of CO(2) during oxygen supply have, to the best of our knowledge, never been studied. We measured the inspired fraction of CO(2) in eighteen non-hypoxemic stable COPD patients with a capnograph during a 5 min trial with two different modes of oxygen supply (oxygen mask without reservoir bag and nasal prongs, respectively at a flow of 10 l/min and 2l/min). We found no significant increase in inspiratory CO(2) concentration. These findings suggest that inspired fraction of CO(2) does not increase markedly during controlled oxygen therapy.

Sound Localisation in Practice: An Application in Localisation of Sick Animals in Commercial Piggeries

Vasileios Exadaktylos1, Mitchell Silva1, Sara Ferrari2, Marcella Guarino2 and Daniel Berckmans1 1Measure, Model and Manage Bioresponses (M3-BIORES), Department of Biosystems, Katholieke Universiteit Leuven, Kasteelpark Arenberg 30, 3001 Heverlee 2Department of Veterinarian Sciences and Technologies for Food Safety, Faculty of Veterinary Medicine, University of Milano, Via Celoria 10, 20133 Milan 1Belgium 2Italy