Abstract. Although Fourier and Wavelet Transform have been widely used for texture classification methods in medical images, the discrimination performance of FDCT has not been investigated so far in respect to breast cancer detection. Ιn this paper, three multi-resolution transforms, namely the Discrete Wavelet Transform (DWT), the Stationary Wavelet Transform (SWT) and the Fast Discrete Curvelet Transform (FDCT) were comparatively assessed with respect to their ability to discriminate between malignant and benign breast tumors in Dynamic Contrast-Enhanced Magnetic Resonance Images (DCE-MRI). The mean and entropy of the detail sub-images for each decomposition scheme were used as texture features, which were subsequently fed as input into several classifiers. FDCT features fed to a Linear Discriminant Analysis (LDA) classifier produced the highest overall classification performance (93,18 % Accuracy).

Ιn this paper, a multiresolution approach is proposed for texture characterization of breast tumors in dynamic contrast-enhanced magnetic resonance images. The decomposition scheme represented by the stationary wavelet transform (SWT) is investigated in terms of its' ability to discriminate between malignant and benign tumors. The mean and entropy of the detail subimages produced for the specific decomposition scheme are used as texture features. The extracted features are subsequently provided into a linear classifier in a leave-one-out cross-validation setting. The experimental results for the proposed features exhibit high performance, when compared to the existing approaches, with the classification accuracy approaching 0.91.

Objectives: The purpose of this article is to review the technical and radiological aspects of MagSeed® localisation, to assess its accuracy based on post-localisation mammograms and excision specimen X-rays and to discuss the radiological experience of our institutions. Methods: Two-year data were collected retrospectively from three NHS boards from the West of Scotland. A total of 309 MagSeeds® were inserted under mammographic or ultrasonographic guidance in 300 women with unifocal, multifocal and/or bilateral breast lesions at the day of surgery or up to 30 days prior to it. Radiological review of post-localisation mammograms and intraoperative specimen X-rays as well as a review of the surgical outcomes were performed to assess the accuracy and efficacy of the method. Our experience relating to the technique’s strengths and downsides were also noted. Results: The MagSeeds® were inserted on average 7.2 days before surgery. The localisation technique was straight forward for the radiologists. In 99% of the cases, the MagSeed® was successfully deployed and 100% of the successfully localised lesions were excised at surgery. There was no difference in the accuracy of the localisation whether this was mammographically or ultrasonographically guided. On post-localisation mammograms, the MagSeed® was radiologically accurately positioned in 97.3% of the cases. No delayed MagSeed® migration was observed. On the specimen X-rays, the lesion was centrally positioned in 45.1%, eccentric within more than 1 mm from the margin in 35.7% and in 14.8% it was at the specimen’s margin. The re-excision rate was 18.3%. Conclusion: The MagSeed® is an accurate and reliable localisation method in breast conserving surgery with good surgical outcomes. Advances in knowledge: To our knowledge, the radiological aspects of MagSeed® localisation have not been widely described in peer-reviewed journals thus far.

The authors propose a method for breast dynamic contrast enhanced‐magnetic resonance imaging classification by combining radiomic texture analysis with a hybrid adaptive neuro‐fuzzy inference system (ANFIS)‐particle swarm optimization (PSO) classifier. The fast discrete curvelet transform is utilized as a decomposition scheme in multiple scales. The mean and entropy features extracted from the produced scheme are used as texture descriptors. Principal component analysis (PCA) involves reduction of the dimensionality of the initial feature set. The transformed feature vector is subsequently introduced to a hybrid ANFIS‐PSO classifier. The average overall classification power of the proposed hybrid ANFIS‐PSO classifier is comparatively assessed to that obtained using several classifiers (ANFIS, linear discriminant analysis, Naïve Bayes, artificial neural networks, random forest and support vector machine) by using the 70 training‐30 testing data ratio. The comparison performed highlights the superiority of the proposed methodology, thus underlying the potential of ANFIS‐PSO for the breast cancer diagnosis with a classification accuracy of 94%.

No abstract