GMM y LDA aplicado a la detección de enfermedades pulmonares / GMM and LDA Applied to Lung Diseases Detection

El propósito de este artículo es presentar metodologías que pueden ser usadas para la valoración cuantitativa de los sonidos del pulmón, así como los indicadores de desórdenes respiratorios. En este contexto, se realizaron experimentos utilizando señales normales y anormales de la respiración (LS), las cuales fueron modeladas y evaluadas utilizando principalmente la base de datos RALE y señales de sujetos saludables y no saludables, logrando hasta un 98% de eficiencia. En la práctica médica la evaluación de enfermedades respiratorias involucra a la auscultación, pero la aplicación de métodos de análisis cuantitativos de señales podría mejorar estas valoraciones. En particular, se sugiere una metodología de evaluación acústica basada en representaciones de vectores acústicos MFCC (Coeficientes Cepstrales en Frecuencia Mel), GMM (Modelos Mezclados Gaussianos) y LDA (Análisis Discriminante Lineal). Estas técnicas podrían asistir en un análisis más amplio, identificación y diagnóstico de desórdenes pulmonares manifestados por sonidos respiratorios peculiares tales como sibilancias, crepitancias y asma, y distinguiéndolos de los sonidos respiratorios normales.

This study presents experimentally tested methods, which can be used for a quantitative assessment of respiratory sounds as the indicators of pulmonary disorders. In particular, conducted experiments considered both normal and abnormal lung sounds (LS). As a part of the RALE Database, signals were recorded from healthy subjects and those with respiratory disorders. Current medical practices including evaluation of respiratory diseases often involve qualitative and frequently subjective auscultation. However, the application of quantitative signal analysis methods could improve the assessments of these diseases. In particular, we utilized acoustic evaluation methodologies based on the MFCC (Mel frequency Cepstral Coefficients) acoustic vectors representation, GMM (Gaussian Mixed Models), and LDA (Linear Discriminant Analysis). To assure the validity of determined class models representing diagnostic classification, the LS signals were cross validated within sequential sets of respiratory cycles for a given subject as well as cross correlated within the specific groups of subjects representing particular conditions of normal or given class of abnormal pulmonary functions. Higher order MFCC vectors, including 9, 10 and 11 Gaussian mixtures, resulted in improved classification of the LS attributes, reached up to 98% of efficiency recognition. This documented automated classification of LS makes it suitable for a more efficient mass screening of respiratory disorders. In particular, the presence of peculiar sounds such as crackles and wheezes lead to more robust models thus reflecting the useful applicability of the presented diagnostic tool. These techniques can assist in broader analysis, identification, and diagnosis of pulmonary disorders manifested by peculiar auscultatory findings.

Method for automatic detection of wheezing in lung sounds

The present report describes the development of a technique for automatic wheezing recognition in digitally recorded lung sounds. This method is based on the extraction and processing of spectral information from the respiratory cycle and the use of these data for user feedback and automatic recognition. The respiratory cycle is first pre-processed, in order to normalize its spectral information, and its spectrogram is then computed. After this procedure, the spectrogram image is processed by a two-dimensional convolution filter and a half-threshold in order to increase the contrast and isolate its highest amplitude components, respectively. Thus, in order to generate more compressed data to automatic recognition, the spectral projection from the processed spectrogram is computed and stored as an array. The higher magnitude values of the array and its respective spectral values are then located and used as inputs to a multi-layer perceptron artificial neural network, which results an automatic indication about the presence of wheezes. For validation of the methodology, lung sounds recorded from three different repositories were used. The results show that the proposed technique achieves 84.82% accuracy in the detection of wheezing for an isolated respiratory cycle and 92.86% accuracy for the detection of wheezes when detection is carried out using groups of respiratory cycles obtained from the same person. Also, the system presents the original recorded sound and the post-processed spectrogram image for the user to draw his own conclusions from the data.

Sibilancias recurrentes en el niño / Recurrent wheezes in the child

Los objetivos de esta investigación fueron identificar algunas características y factores de riesgo en niños con sibilancias recurrentes, un decenio después del primer episodio de bronquiolitis aguda por virus sincitial respiratorio. Se realizó un estudio descriptivo retrospectivo, de tipo longitudinal, con 34 niños que se clasificaron en sibilantes transitorios y sibilantes persistentes. Se identificaron factores de riesgo y factores socioeconómicos y ambientales. Se realizaron pruebas cutáneas de sensibilidad con aeroalergenos. Se observaron sibilancias persistentes en 22 pacientes (64,7 por ciento) y sibilancias transitorias en 12 (35,3 por ciento). El antecedente atópico familiar màs frecuente fue el asma bronquial, con predominio en los niños con sibilancias persistentes (68,1 por ciento). El destete precoz constituyó el antecedente personal más frecuente. El hábito de fumar y la asistencia a círculos infantiles fueron los factores de riesgo socioeconómicos y ambientales que predominaron (63,6 por ciento y 59 por ciento, respectivamente), en especial en niños con sibilancias persistentes. Requirieron atención en la Unidad de Terapia Intensiva Polivalente 6 niños sibilantes persistentes (27,2 por ciento) y solo 1 niño (8,3 por ciento) sibilante transitorio necesitó admisión en dicha unidad. Las pruebas cutáneas de aeroalergenos fueron positivas en 7 niños sibilantes permanentes (31,8 por ciento) y en 1 de los sibilantes transitorios (8,3 por ciento). Las sibilancias recurrentes en el niño después de un primer episodio de bronquiolitis por virus sincitial respiratorio están generalmente asociadas a factores de riesgo, en especial al antecedente atópico familiar y al hábito de fumar. Es muy importante reconocer tempranamente a los niños con riesgo de presentar asma persistente atópica e intervenir terapéuticamente con el fin de evitar el deterioro de la función pulmonar.

The objectives of this research were to identify some characteristics and risk factors in children with recurrent wheezes 10 years after the first episode of acute bronchiolitis caused by respiratory syncytial virus. A descriptive retrospective longitudinal study was conducted among 34 children that were classified as transient wheezing and persistent wheezing. Risk, socioeconomic and environmental factors were identified. Skin sensitive tests with aeroallergens were performed. Persistent wheezes were observed in 22 patients (64,7 percent) and transient wheezes in 12 (35,3 percent). The most common atopic family history was bronchial asthma with predominance in children with persistent wheezes (68,1 percent). Early weaning was the most frequent personal history. Smoking and the attendance to day care centers were the prevailing socioeconomic and environmental risk factors (63,6 percent and 59 percent, respectively), specially in children with persistent wheezes. Six persistent wheezing children (27,2 percent) required attention at the Polivalent Intensive Care Unit, whereas only one transient wheezing child (8,3 percent) had to be admitted in that unit. The aeroallergen skin-prick tests yielded positive in 7 permanent wheezing children (31,8 percent) and in 1 transient wheezing child (8,3 percent). The recurrent wheezes in the child after a first episode of bronchiolitis due to respiratory syncytial virus are generally associated with risk factors, particularly to the atopic family history and smoking. It is very important to early identify the children at risk for presenting atopic persistent asthma and to take therapeutic measures in order to prevent the deterioration of the pulmonary function.