Classification of lung sounds using convolutional neural networks

Özkan Kılıç,Bahar Kurt,Sevgi Saryal,Murat Aykanat

doi:10.1186/s13640-017-0213-2

Abstract

In the field of medicine, with the introduction of computer systems that can collect and analyze massive amounts of data, many non-invasive diagnostic methods are being developed for a variety of conditions. In this study, our aim is to develop a non-invasive method of classifying respiratory sounds that are recorded by an electronic stethoscope and the audio recording software that uses various machine learning algorithms.In order to store respiratory sounds on a computer, we developed a cost-effective and easy-to-use electronic stethoscope that can be used with any device. Using this device, we recorded 17,930 lung sounds from 1630 subjects.We employed two types of machine learning algorithms; mel frequency cepstral coefficient (MFCC) features in a support vector machine (SVM) and spectrogram images in the convolutional neural network (CNN). Since using MFCC features with a SVM algorithm is a generally accepted classification method for audio, we utilized its results to benchmark the CNN algorithm. We prepared four data sets for each CNN and SVM algorithm to classify respiratory audio: (1) healthy versus pathological classification; (2) rale, rhonchus, and normal sound classification; (3) singular respiratory sound type classification; and (4) audio type classification with all sound types. Accuracy results of the experiments were; (1) CNN 86%, SVM 86%, (2) CNN 76%, SVM 75%, (3) CNN 80%, SVM 80%, and (4) CNN 62%, SVM 62%, respectively.As a result, we found out that spectrogram image classification with CNN algorithm works as well as the SVM algorithm, and given the large amount of data, CNN and SVM machine learning algorithms can accurately classify and pre-diagnose respiratory audio.

Highlights

That is why we developed our electronic stethoscope with as much sound isolation as possible and selected our recording device carefully
The hardware-software system was used to collect a dataset of respiratory sounds to train support vector machine (SVM) and convolutional neural network (CNN) machine learning algorithms for the automated analysis and diagnosis
Since mel frequency cepstral coefficient (MFCC) features combined with SVM is a generally accepted practice for audio classification, we used it as a benchmark for our CNN algorithm

Summary

Introduction

Most of the time, it is very hard to spot these patterns, especially if the data is very large. Data collected from the environment is usually non-linear, so we cannot use traditional methods to find patterns or create mathematical models. Various technologies, such as expert systems, have been used to attempt to solve this problem. For critical systems, the error rate for the decision was too high [1]. The latest technology that is attempting to solve this problem is machine learning. Various successful algorithms were developed and with the deep learning algorithms, error rate became close to Research in this area attempts to make better representations and create models to learn these representations from large-scale unlabeled data [2]. Some of the representations are inspired by advances in neuroscience and are loosely based on interpretation of information processing and communication patterns in a nervous system, such as neural coding which attempts to define a relationship between the stimulus and the neuronal responses and the relationship among the electrical activities of the neurons in the brain [3, 4]

Objectives

Methods

Results

Conclusion