Classifying infant cry patterns by the Genetic Selection of a Fuzzy Model

Rosales-Pérez, Alejandro; Reyes-García, Carlos A.; González, Jesús A.; Reyes-Galaviz, Orion F.; Orlandi, Silvia

doi:10.1016/j.bspc.2014.10.002

Cited by 50 publications

(33 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Abnormalities were searched in the cries of newborns with multiple or severe problems during 12 A c c e p t e d M a n u s c r i p t the neonatal period, such as low birth weight, respiratory symptoms, jaundice, apnea, but also infants subsequently victims of presumed sudden infant death syndrome (Golub and Corwin, 1982). In the 2000s, normal and pathological cries began to be automatically labeled thanks to a wide variety of machine learning approaches in the context of deafness (Orozco-García and Reyes-García, 2003;Suaste-Rivas et al, 2004;Rosales-Pérez et al, 2015), hypoxia-based Central Nervous System (CNS) diseases (Ortiz et al, 2004), cleft palate (Lederman et al, 2008) and asphyxia (Suaste-Rivas et al, 2004;Hariharan et al, 2011;Rosales-Pérez et al, 2015).…”

Section: Clinical Applicationsmentioning

confidence: 99%

“…Among them, we can cite the Baby Chillanto cry signals database, property of the Instituto Nacional de Astrofisica Optica y Electronica (Mexico). It was composed by five types of annotated cries (pain, hunger, normal, asphyxia and deaf) and on which relied the work of Reyes-Garcia group (Orozco-García and Reyes-García, 2003;Ortiz et al, 2004;Suaste-Rivas et al, 2004;Reyes-Galaviz et al, 2004;Galaviz and García, 2005;Reyes-Galaviz et al, 2005;Barajas-Montiel and Reyes-García, 2006;Hariharan et al, 2011;Verduzco-Mendoza et al, 2012;Díaz et al, 2012;Rosales-Pérez et al, 2015;Wahid et al, 2016). Cohort size varied from few infants (Lederman et al, 2008;Yamamoto et al, 2013) to more than three hundred babies (Várallyay, 2007) or more than 120 premature babies in (Stevens et al, 1994).…”

Section: Audio Acquisitionmentioning

confidence: 99%

“…3.2.5. Automatic cry classification Automated cry classification have been performed after a manual (Grunau et al, 1990;Fuller, 1991;Johnston et al, 1993;Stevens et al, 1994;Fuller and Conner, 1995;Goberman and Robb, 1999;Orozco-García and Reyes-García, 2003;Bellieni et al, 2004;Suaste-Rivas et al, 2004;Ortiz et al, 2004;Reyes-Galaviz et al, 2004Galaviz and García, 2005;Barajas-Montiel and Reyes-García, 2006;Lederman et al, 2008;Hariharan et al, 2011;Verduzco-Mendoza et al, 2012;Shinya et al, 2014;Rosales-Pérez et al, 2015;Borysiak et al, 2016;Wahid et al, 2016) or an automated (Manfredi et al, 2009;Orlandi et al, 2015Orlandi et al, , 2016 segmentation step. First studies used classical statistical approaches such as Student T-test (Manfredi et al, 2009;Verduzco-Mendoza et al, 2012), ANOVA (Grunau et al, 1990;Fuller, 1991;Goberman and Robb, 1999;Shinya et al, 2014) and MANOVA (Johnston et al, 1993;Stevens et al, 1994;Fuller and Conner, 1995) or regression (Bellieni et al, 2004;Borysiak et al, 2016).…”

Section: Automatic Cry Segmentationmentioning

confidence: 99%

“…Different numbers of cry classes were considered according to the clinical target: two classes (normal vs abnormal (Orozco-García and Reyes-García, 2003;Ortiz et al, 2004;Lederman et al, 2008;Hariharan et al, 2011;Rosales-Pérez et al, 2015) or preterm vs fullterm (Orlandi et al, , 2016), three classes (normal, hypo acoustic and asphyxia (Reyes- Suaste-Rivas et al, 2004;Galaviz and García, 2005;Reyes-Galaviz et al, 2005;Barajas-Montiel and Reyes-García, 2006), hunger, pain and sleep (Chang and Li, 2016) or hunger, pain and no-pain-no-hunger (Barajas-Montiel and Reyes-García, 2006)) and five classes (pain, asphyxia, hunger, deaf and normal (Wahid et al, 2016)). A wide variety of machine learning approaches has been evaluated, regrouping classical methods such as SVM (Barajas-Montiel and Reyes-García, 2006;Orlandi et al, 2016), KNN (Rosales-Pérez et al, 2015), Random Forest (RF) (Rosales-Pérez et al, 2015;Orlandi et al, 2015Orlandi et al, , 2016, HMM (Lederman et al, 2008) or Neural Networks (Schönweiler et al, 1996;Orozco-García and Reyes-García, 2003;Suaste-Rivas et al, 2004;Ortiz et al, 2004;Reyes-Galaviz et al, 2004Galaviz and García, 2005;Hariharan et al, 2011;Wahid et al, 2016). Classification results were efficient since some studies reached results above 95% (Orozco-García and Reyes-García, 2003;Reyes-Galaviz et al, 2004Galaviz and García, 2005;Barajas-Montiel and Reyes-García, 2006;Hariharan et al, 2011).…”

Section: Automatic Cry Segmentationmentioning

confidence: 99%

See 3 more Smart Citations

Video and audio processing in paediatrics: a review

Cabon¹,

Porée²,

Simon³

et al. 2019

Physiol. Meas.

View full text Add to dashboard Cite

Objective: Video and sound acquisition and processing technologies have seen great improvements in recent decades, with many applications in the biomedical area. The aim of this paper is to review the overall state of the art of advances within these topics in paediatrics and to evaluate their potential application for monitoring in the neonatal intensive care unit (NICU). Approach: For this purpose, more than 150 papers dealing with video and audio processing were reviewed. For both topics, clinical applications are described according to the considered cohorts—full-term newborns, infants and toddlers or preterm newborns. Then, processing methods are presented, in terms of data acquisition, feature extraction and characterization. Main results: The paper first focuses on the exploitation of video recordings; these began to be automatically processed in the 2000s and we show that they have mainly been used to characterize infant motion. Other applications, including respiration and heart rate estimation and facial analysis, are also presented. Audio processing is then reviewed, with a focus on the analysis of crying. The first studies in this field focused on induced-pain cries and the newest ones deal with spontaneous cries; the analyses are mainly based on frequency features. Then, some papers dealing with non-cry signals are also discussed. Significance: Finally, we show that even if recent improvements in digital video and signal processing allow for increased automation of processing, the context of the NICU makes a fully automated analysis of long recordings problematic. A few proposals for overcoming some of the limitations are given.

show abstract

Section: Clinical Applicationsmentioning

confidence: 99%

Section: Audio Acquisitionmentioning

confidence: 99%

Section: Automatic Cry Segmentationmentioning

confidence: 99%

Section: Automatic Cry Segmentationmentioning

confidence: 99%

See 2 more Smart Citations

Video and audio processing in paediatrics: a review

Cabon¹,

Porée²,

Simon³

et al. 2019

Physiol. Meas.

View full text Add to dashboard Cite

show abstract

“…Many researchers collected their own data including our Speech Research Group at DA-IICT [8], [9]. Few studies used Baby Chillanto infant cry database [10] while most of the recent studies used their own corpus. The early work includes analysis of different features for infant cry signals [11].…”

Section: Introductionmentioning

confidence: 99%

Auditory Filterbank Learning Using ConvRBM for Infant Cry Classification

Sailor

Patil

2018

Interspeech 2018

View full text Add to dashboard Cite

The infant cry classification is a socially-relevant problem where the task is to classify the normal vs. pathological cry signals. Since the cry signals are very different from the speech signals in terms of temporal and spectral content, there is a need for better feature representation for infant cry signals. In this paper, we propose to use unsupervised auditory filterbank learning using Convolutional Restricted Boltzmann Machine (ConvRBM). Analysis of the subband filters shows that most of the subband filters are Fourier-like basis functions. The infant cry classification experiments were performed on the two databases, namely, DA-IICT Cry and Baby Chillanto. The experimental results show that the proposed features perform better than the standard Mel Frequency Cepstral Coefficients (MFCC) using various statistically meaningful performance measures. In particular, our proposed ConvRBM-based features obtained an absolute improvement of 2 % and 0.58 % in the classification accuracy on the DA-IICT Cry and the Baby Chillanto database, respectively.

show abstract