Comparative Performance Analysis of Machine Learning Classifiers in Detection of Childhood Pneumonia Using Chest Radiographs

Sousa, Rafael Teixeira; Marques, Oge; Soares, Fabrízzio; Sene, Iwens I. G.; Oliveira, Leandro Luís Galdino de; Spoto, Edmundo Sérgio

doi:10.1016/j.procs.2013.05.444

Cited by 60 publications

(33 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…All features tested are based on texture and extracted in nine subspaces of Haar wavelet, like our previous paper [7]. We performed a 10-fold cross-validation test with each classifier.…”

Section: Methodsmentioning

confidence: 99%

“…In this work we use the features and dataset employed in previous studies [1] [6] [7], which have resulted in a full CAD system for pneumonia detection called PneumoCAD, which has been applied to assist in diagnostics, as well as to train and improve radiologists' expertise in childhood pneumonia detection using chest radiographs. PneumoCAD is currently in prototype stage.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluation of Classifiers to a Childhood Pneumonia Computer-Aided Diagnosis System

Sousa

2014

2014 IEEE 27th International Symposium on Computer-Based Medical Systems

Self Cite

View full text Add to dashboard Cite

This work extends PneumoCAD, a Computer-Aided Diagnosis system for detecting pneumonia in infants using radiographic images [1], with the aim of improving the system's accuracy and robustness. We implement and compare five contemporary machine learning classifiers, namely: Naïve Bayes, K-Nearest Neighbor (KNN), Support Vector Machines (SVM), Multi-Layer Perceptron (MLP) and Decision Tree, combined with three dimensionality reduction algorithms: Sequential Forward Selection (SFS), Principal Component Analysis (PCA) and Kernel Principal Component Analysis (KPCA). Current results demonstrate that Naïve Bayes classifier combined with KPCA produces the best overall results.

show abstract

“…All features tested are based on texture and extracted in nine subspaces of Haar wavelet, like our previous paper [7]. We performed a 10-fold cross-validation test with each classifier.…”

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Evaluation of Classifiers to a Childhood Pneumonia Computer-Aided Diagnosis System

Sousa

2014

2014 IEEE 27th International Symposium on Computer-Based Medical Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…These methods were able to achieve considerable results and had a comparatively low misclassification rate of~20% with kNN, and 10% with kNN and Haar Wavelet, despite the area-under-curve (AUC) value being comparatively less for receiver operating characteristic (ROC) curve. Later on, a comparative analysis of the algorithm had been conducted using Pneumo-CAD with Sequential Forward Elimination (SFE), Pneumo-CAD without SFE, and Support Vector Machine (which used SFE) gave accuracies of 66%, 70%, and 77% [19], respectively. This research also tested the Naïve Bayes algorithm and got an accuracy of 68% with it.…”

Section: Related Workmentioning

confidence: 99%

Detecting Pneumonia Using Convolutions and Dynamic Capsule Routing for Chest X-ray Images

Mittal

Kumar

Mittal³

et al. 2020

Sensors

119

View full text Add to dashboard Cite

An entity’s existence in an image can be depicted by the activity instantiation vector from a group of neurons (called capsule). Recently, multi-layered capsules, called CapsNet, have proven to be state-of-the-art for image classification tasks. This research utilizes the prowess of this algorithm to detect pneumonia from chest X-ray (CXR) images. Here, an entity in the CXR image can help determine if the patient (whose CXR is used) is suffering from pneumonia or not. A simple model of capsules (also known as Simple CapsNet) has provided results comparable to best Deep Learning models that had been used earlier. Subsequently, a combination of convolutions and capsules is used to obtain two models that outperform all models previously proposed. These models—Integration of convolutions with capsules (ICC) and Ensemble of convolutions with capsules (ECC)—detect pneumonia with a test accuracy of 95.33% and 95.90%, respectively. The latter model is studied in detail to obtain a variant called EnCC, where n = 3, 4, 8, 16. Here, the E4CC model works optimally and gives test accuracy of 96.36%. All these models had been trained, validated, and tested on 5857 images from Mendeley.

show abstract

“…Sousa [22] diagnoses childhood pneumonia by combining features dimension reduction using SFS, PCA and KPCA and classification algorithms. The best results have been achieved for KPCA: SVM 89%, kNN 93%, Naive Bayes 96%.…”

Section: Classificationmentioning

confidence: 99%

Advanced approaches to computer-aided detection of thoracic diseases on chest X-rays

Zakirov¹,

Kuleev²,

Vladimirov³

2015

ams

View full text Add to dashboard Cite

The research focus of the current paper are modern algorithms for solving the problem of automatic diagnostic of thorax diseases based on X-ray images. Special attention was paid to image preprocessing, classification together with calculation of features and their comparison in terms of efficiency. Approaches mentioned in this paper are used for development of new algorithm for automatic diagnostic of medical images.

show abstract

Comparative Performance Analysis of Machine Learning Classifiers in Detection of Childhood Pneumonia Using Chest Radiographs

Cited by 60 publications

References 8 publications

Evaluation of Classifiers to a Childhood Pneumonia Computer-Aided Diagnosis System

Evaluation of Classifiers to a Childhood Pneumonia Computer-Aided Diagnosis System

Detecting Pneumonia Using Convolutions and Dynamic Capsule Routing for Chest X-ray Images

Advanced approaches to computer-aided detection of thoracic diseases on chest X-rays

Contact Info

Product

Resources

About