A novel computer-aided diagnosis system for breast MRI based on feature selection and ensemble learning

Lü, Wei; Li, Zhe; Chu, Jinghui

doi:10.1016/j.compbiomed.2017.03.002

Cited by 57 publications

(25 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Studies have shown that performance of breast lesion diagnosis varied due to the intra- and inter-reader variability 13 , and only approximate one in four biopsies are malignant 27 . Thus, in order to help improve accuracy in classification between malignant and benign breast lesions, developing computer-aided diagnosis (CAD) schemes aiming to assist radiologists in their decision-making for better assessing risk of lesion malignancy has been attracting extensive research interest in medical imaging field for the last two decades 15,20 . Although CEDM is an emerging imaging modality, our recent pilot study demonstrated that classification results based on a machine learning classifier that fuses the computed quantitative image features extracted from CEDM images might provide complementary information to radiologists in particular to help reduce false-positive recalls 18 .…”

Section: Introductionmentioning

confidence: 99%

Classification of Breast Masses Using a Computer-Aided Diagnosis Scheme of Contrast Enhanced Digital Mammograms

et al. 2018

View full text Add to dashboard Cite

Contrast-enhanced digital mammography (CEDM) is a promising imaging modality in breast cancer diagnosis. This study aims to investigate how to optimally develop a computer-aided diagnosis (CAD) scheme of CEDM images to classify breast masses. A CEDM dataset of 111 patients was assembled, which includes 33 benign and 78 malignant cases. Each CEDM includes two types of images namely, low energy (LE) and dual-energy subtracted (DES) images. A CAD scheme was applied to segment mass regions depicting on LE and DES images separately. Optimal segmentation results generated from DES images were also mapped to LE images or vice versa. After computing image features, multilayer perceptron based machine learning classifiers that integrate with a correlation-based feature subset evaluator and leave-one-case-out cross-validation method were built to classify mass regions. When applying CAD to DES and LE images with original segmentation, areas under ROC curves (AUC) were 0.759 ± 0.053 and 0.753 ± 0.047, respectively. After mapping the mass regions optimally segmented on DES images to LE images, AUC significantly increased to 0.848 ± 0.038 (p < 0.01). Study demonstrated that DES images eliminated overlapping effect of dense breast tissue, which helps improve mass segmentation accuracy. The study demonstrated that applying a novel approach to optimally map mass region segmented from DES images to LE images enabled CAD to yield significantly improved performance.

show abstract

Section: Introductionmentioning

confidence: 99%

Classification of Breast Masses Using a Computer-Aided Diagnosis Scheme of Contrast Enhanced Digital Mammograms

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Furthermore, the segmented regions of interest (SROIs) in each image was marked by the expert radiologist. 20,41-43 In this work, an interactive segmentation approach has been used to segment the regions of interest for classification, where only initial seed points have to be entered manually. The marked SROIs are regions with most diagnostic information in textural form of the specific liver tissue class.…”

Section: Methodsmentioning

confidence: 99%

Preliminary Study of Chronic Liver Classification on Ultrasound Images Using an Ensemble Model

2018

View full text Add to dashboard Cite

Chronic liver diseases are fifth leading cause of fatality in developing countries. Their early diagnosis is extremely important for timely treatment and salvage life. To examine abnormalities of liver, ultrasound imaging is the most frequently used modality. However, the visual differentiation between chronic liver and cirrhosis, and presence of heptocellular carcinomas (HCC) evolved over cirrhotic liver is difficult, as they appear almost similar in ultrasound images. In this paper, to deal with this difficult visualization problem, a method has been developed for classifying four liver stages, that is, normal, chronic, cirrhosis, and HCC evolved over cirrhosis. The method is formulated with selected set of "handcrafted" texture features obtained after hierarchal feature fusion. These multiresolution and higher order features, which are able to characterize echotexture and roughness of liver surface, are extracted by using ranklet, gray-level difference matrix and gray-level co-occurrence matrix methods. Thereafter, these features are applied on proposed ensemble classifier that is designed with voting algorithm in conjunction with three classifiers, namely, k-nearest neighbor (k-NN), support vector machine (SVM), and rotation forest. The experiments are conducted to evaluate the (a) effectiveness of "handcrafted" texture features, (b) performance of proposed ensemble model, (c) effectiveness of proposed ensemble strategy, (d) performance of different classifiers, and (e) performance of proposed ensemble model based on Convolutional Neural Networks (CNN) features to differentiate four liver stages. These experiments are carried out on database of 754 segmented regions of interest formed by clinically acquired ultrasound images. The results show that classification accuracy of 96.6% is obtained by use of proposed classifier model.

show abstract

“…The J48 DT-inducing algorithm is an implementation of the well-known C4.5 DT in WEKA software [60]. The C4.5 tree which is known as one of top 10 data mining algorithms, chooses appropriate features and nodes based on the information gain ratios [61,62]. REP tree classi er is a fast DT learner that builds a tree based on information gain with entropy and prunes it using reducederror pruning [63].…”

Section: Decision Tree (Dt)mentioning

confidence: 99%

Development of a Fuzzy Model for Differentiating Peanut Plant from Broadleaf Weeds Using Image Features

Bakhshipour

Zareiforoush

2020

Preprint

View full text Add to dashboard Cite

A combination of decision tree (DT) and fuzzy logic techniques was used to develop a fuzzy model for differentiating peanut plant from weeds. Color features and wavelet-based texture features were extracted from images of peanut plant and its three common weeds. Two feature selection techniques namely Principal Component Analysis (PCA) and Correlation-based Feature Selection (CFS) were applied on input dataset and three Decision Trees (DTs) including J48, Random Tree (RT), and Reduced Error Pruning (REP) were used to distinguish between different plants. In all cases, the best overall classification accuracies were achieved when CFS-selected features were used as input data. The obtained accuracies of J48-CFS, REP-CFS, and RT-CFS trees for classification of the four plant categories namely peanut plant, Velvetleaf, False daisy, and Nicandra, were 80.83%, 80.00% and 79.17% respectively. Along with these almost low accuracies, the structures of the decision trees were complex making them unsuitable for developing a fuzzy inference system. The classifiers were also used for differentiating peanut plant from the group of weeds. The overall accuracies on training and testing datasets were respectively 95.56% and 93.75% for J48-CFS; 92.78% and 91.67% for REP-CFS; and 93.33% and 92.59% for RT-CFS DTs. The results showed that the J48-CFS and REP-CFS were the most appropriate models to set the membership functions and rules of the fuzzy classifier system. Based on the results, it can be concluded that the developed DT-based fuzzy logic model can be used effectively to discriminate weeds from peanut plant in the form of machine vision-based cultivating systems.

show abstract

A novel computer-aided diagnosis system for breast MRI based on feature selection and ensemble learning

Cited by 57 publications

References 33 publications

Classification of Breast Masses Using a Computer-Aided Diagnosis Scheme of Contrast Enhanced Digital Mammograms

Classification of Breast Masses Using a Computer-Aided Diagnosis Scheme of Contrast Enhanced Digital Mammograms

Preliminary Study of Chronic Liver Classification on Ultrasound Images Using an Ensemble Model

Development of a Fuzzy Model for Differentiating Peanut Plant from Broadleaf Weeds Using Image Features

Contact Info

Product

Resources

About