Wavelet Convolution Neural Network for Classification of Spiculated Findings in Mammograms

Jasionowska, Magdalena; Gacek, Aleksandra

doi:10.1007/978-3-030-23762-2_18

Cited by 7 publications

(5 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The AUC of the CNN was 0.87, which was higher than the radiologists' mean AUC (0.84), although the difference was not significant. On the other hand, the studies discussed in 40 , 41 circumvent the use of deep learning by adopting wavelet decomposition.…”

Section: Related Workmentioning

confidence: 99%

A novel wavelet decomposition and transformation convolutional neural network with data augmentation for breast cancer detection using digital mammogram

Oyelade

Ezugwu

2022

Sci Rep

View full text Add to dashboard Cite

Research in deep learning (DL) has continued to provide significant solutions to the challenges of detecting breast cancer in digital images. Image preprocessing methods and architecture enhancement techniques have been proposed to improve the performance of DL models such as convolutional neural networks (CNNs). For instance, the wavelet decomposition function has been used for image feature extraction in CNNs due to its strong compactness. Additionally, CNN architectures have been optimized to improve the process of feature detection to support the classification process. However, these approaches still lack completeness, as no mechanism exists to discriminate features to be enhanced and features to be eliminated for feature enhancement. More so, no studies have approached the use of wavelet transform to restructure CNN architectures to improve the detection of discriminant features in digital mammography for increased classification accuracy. Therefore, this study addresses these problems through wavelet-CNN-wavelet architecture. The approach presented in this paper combines seam carving and wavelet decomposition algorithms for image preprocessing to find discriminative features. These features are passed as input to a CNN-wavelet structure that uses the new wavelet transformation function proposed in this paper. The CNN-wavelet architecture applied layers of wavelet transform and reduced feature maps to obtain features suggestive of abnormalities that support the classification process. Meanwhile, we synthesized image samples with architectural distortion using a generative adversarial network (GAN) model to argue for their training datasets' insufficiency. Experimentation of the proposed method was carried out using DDSM + CBIS and MIAS datasets. The results obtained showed that the new method improved the classification accuracy and lowered the loss function values. The study's findings demonstrate the usefulness of the wavelet transform function in restructuring CNN architectures for performance enhancement in detecting abnormalities leading to breast cancer in digital mammography.

show abstract

Section: Related Workmentioning

confidence: 99%

A novel wavelet decomposition and transformation convolutional neural network with data augmentation for breast cancer detection using digital mammogram

Oyelade

Ezugwu

2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…The performance comparison of proposed framework for both CBIS-DDSM and MIAS databases with other existing framework (CNN based texture feature based) is shown in Table IV. For CBIS-DDSM database, the proposed framework has given better performance than the CNN based approaches [13], [36], [38], [39], [42]. In [36], authors used bag of visual words with CNN for both segment-free and segment dependent based classification and the segment-free based method has shown better performance.…”

Section: Performance Comparison Of Proposed Frameworkmentioning

confidence: 99%

“…(feature like local binary pattern, global image descriptor, and histogram of oriented grading) and deep features from DenseNet and VGG. In[42] authors have used DWT (level-2) and extracted features from detail subband images. Finally, CNN is used for classification.…”

mentioning

confidence: 99%

Automatic Detection of Benign and Malignant Masses in Mammogram Using 2D-Fourier-BesselIntrinsic Band Functions and Improved Feature Space

Chaudhary¹,

Pachori²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Most of these studies are completely dependent on segmentation and hand-crafted features, which are not optimal. In some studies, wavelet coefficients and features extracted from wavelet coefficients have been used to detect breast abnormalities in mammograms (5,6). In these methods, it is necessary to define hand-crafted features to detect and classify mass regions in mammograms; therefore, the performance of these methods is limited.…”

Section: Introductionmentioning

confidence: 99%

An Efficient Method for Automated Breast Mass Segmentation and Classification in Digital Mammograms

2021

View full text Add to dashboard Cite

Background: Automatic detection and classification of breast masses in mammograms are still challenging tasks. Today, computer-aided diagnosis (CAD) systems are being developed to assist radiologists in interpreting mammograms. Objectives: This study aimed to provide a novel method for automatic segmentation and classification of masses in mammograms to help radiologists make an accurate diagnosis. Materials and Methods: For an efficient mass diagnosis in mammograms, we proposed an automatic scheme to perform both mass detection and classification. First, a combination of several image enhancement algorithms, including contrast-limited adaptive histogram equalization (CLAHE), guided imaging, and median ﬁltering, was investigated to enhance the visual features of breast area and increase the accuracy of segmentation outcomes. Second, the density of discrete wavelet coefficient density (DDWCs), based on the quincunx lifting scheme (QLS), was proposed to find suspicious mass regions or regions of interest (ROIs). Finally, mass lesions that appeared in the mammogram were classiﬁed into four categories of benign, probably benign, malignant, and probably malignant, based on the morphological shape. The proposed method was evaluated among 1593 images from the Curated Breast Imaging Subset-Digital Database for Screening Mammography (CBIS-DDSM) dataset. Results: The experimental results revealed that the suspected region localization had 100% sensitivity, with a mean of 6.4 ± 4.5 false positive (FP) detections per image. Moreover, the results showed an overall accuracy of 85.9% and an area under the curve (AUC) of 0.901 for the mass classification algorithm. Conclusion: The present results showed the comparable performance of our proposed method to that of the state-of-the-art methods.

show abstract

Wavelet Convolution Neural Network for Classification of Spiculated Findings in Mammograms

Cited by 7 publications

References 15 publications

A novel wavelet decomposition and transformation convolutional neural network with data augmentation for breast cancer detection using digital mammogram

A novel wavelet decomposition and transformation convolutional neural network with data augmentation for breast cancer detection using digital mammogram

Automatic Detection of Benign and Malignant Masses in Mammogram Using 2D-Fourier-BesselIntrinsic Band Functions and Improved Feature Space

An Efficient Method for Automated Breast Mass Segmentation and Classification in Digital Mammograms

Contact Info

Product

Resources

About