Computer-aided classification of mammographic masses and normal tissue: linear discriminant analysis in texture feature space

Synergistic eigenanalysis of covariance and Hessian matrices for enhanced binary classification on health datasets

Covariance and Hessian matrices have been analyzed separately in the literature for classification problems. However, integrating these matrices has the potential to enhance their combined power in improving classification performance. We present a novel approach that combines the eigenanalysis of a covariance matrix evaluated on a training set with a Hessian matrix evaluated on a deep learning model to achieve optimal class separability in binary classification tasks. Our approach is substantiated by formal proofs that establish its capability to maximize between-class mean distance (the concept of separation) and minimize within-class variances (the concept of compactness), which together define the two linear discriminant analysis (LDA) criteria, particularly under ideal data conditions such as isotropy around class means and dominant leading eigenvalues. By projecting data into the combined space of the most relevant eigendirections from both matrices, we achieve optimal class separability as per these LDA criteria. Empirical validation across neural and health datasets consistently supports our theoretical framework and demonstrates that our method outperforms established methods. Our method stands out by addressing both separation and compactness criteria, unlike PCA and the Hessian method, which predominantly emphasize one criterion each. This comprehensive approach captures intricate patterns and relationships, enhancing classification performance. Furthermore, through the utilization of both LDA criteria, our method outperforms LDA itself by leveraging higher-dimensional feature spaces, in accordance with Cover's theorem, which favors linear separability in higher dimensions. Additionally, our approach sheds light on complex DNN decision-making, rendering them comprehensible within a 2D space.

A machine learning-based approach to identify peripheral artery disease using texture features from contrast-enhanced magnetic resonance imaging

Diagnosing and assessing the risk of peripheral artery disease (PAD) has long been a focal point for medical practitioners. The impaired blood circulation in PAD patients results in altered microvascular perfusion patterns in the calf muscles which is the primary location of intermittent claudication pain. Consequently, we hypothesized that changes in perfusion and increase in connective tissue could lead to alterations in the appearance or texture patterns of the skeletal calf muscles, as visualized with non-invasive imaging techniques. We designed an automatic pipeline for textural feature extraction from contrast-enhanced magnetic resonance imaging (CE-MRI) scans and used the texture features to train machine learning models to detect the heterogeneity in the muscle pattern among PAD patients and matched controls. CE-MRIs from 36 PAD patients and 20 matched controls were used for preparing training and testing data at a 7:3 ratio with cross-validation (CV) techniques. We employed feature arrangement and selection methods to optimize the number of features. The proposed method achieved a peak accuracy of 94.11% and a mean testing accuracy of 84.85% in a 2-class classification approach (controls vs. PAD). A three-class classification approach was performed to identify a high-risk PAD sub-group which yielded an average test accuracy of 83.23% (matched controls vs. PAD without diabetes vs. PAD with diabetes). Similarly, we obtained 78.60% average accuracy among matched controls, PAD treadmill exercise completers, and PAD exercise treadmill non-completers. Machine learning and imaging-based texture features may be of interest in the study of lower extremity ischemia.

Model-based deep CNN-regularized reconstruction for digital breast tomosynthesis with a task-based CNN image assessment approach

Objective. Digital breast tomosynthesis (DBT) is a quasi-three-dimensional breast imaging modality that improves breast cancer screening and diagnosis because it reduces fibroglandular tissue overlap compared with 2D mammography. However, DBT suffers from noise and blur problems that can lower the detectability of subtle signs of cancers such as microcalcifications (MCs). Our goal is to improve the image quality of DBT in terms of image noise and MC conspicuity.Approach. We proposed a model-based deep convolutional neural network (deep CNN or DCNN) regularized reconstruction (MDR) for DBT. It combined a model-based iterative reconstruction (MBIR) method that models the detector blur and correlated noise of the DBT system and the learning-based DCNN denoiser using the regularization-by-denoising framework. To facilitate the task-based image quality assessment, we also proposed two DCNN tools for image evaluation: a noise estimator (CNN-NE) trained to estimate the root-mean-square (RMS) noise of the images, and an MC classifier (CNN-MC) as a DCNN model observer to evaluate the detectability of clustered MCs in human subject DBTs.Main results. We demonstrated the efficacies of CNN-NE and CNN-MC on a set of physical phantom DBTs. The MDR method achieved low RMS noise and the highest detection area under the receiver operating characteristic curve (AUC) rankings evaluated by CNN-NE and CNN-MC among the reconstruction methods studied on an independent test set of human subject DBTs.Significance. The CNN-NE and CNN-MC may serve as a cost-effective surrogate for human observers to provide task-specific metrics for image quality comparisons. The proposed reconstruction method shows the promise of combining physics-based MBIR and learning-based DCNNs for DBT image reconstruction, which may potentially lead to lower dose and higher sensitivity and specificity for MC detection in breast cancer screening and diagnosis.

Skin-Interfaced Wearable Sweat Sensors for Precision Medicine

Wearable sensors hold great potential in empowering personalized health monitoring, predictive analytics, and timely intervention toward personalized healthcare. Advances in flexible electronics, materials science, and electrochemistry have spurred the development of wearable sweat sensors that enable the continuous and noninvasive screening of analytes indicative of health status. Existing major challenges in wearable sensors include: improving the sweat extraction and sweat sensing capabilities, improving the form factor of the wearable device for minimal discomfort and reliable measurements when worn, and understanding the clinical value of sweat analytes toward biomarker discovery. This review provides a comprehensive review of wearable sweat sensors and outlines state-of-the-art technologies and research that strive to bridge these gaps. The physiology of sweat, materials, biosensing mechanisms and advances, and approaches for sweat induction and sampling are introduced. Additionally, design considerations for the system-level development of wearable sweat sensing devices, spanning from strategies for prolonged sweat extraction to efficient powering of wearables, are discussed. Furthermore, the applications, data analytics, commercialization efforts, challenges, and prospects of wearable sweat sensors for precision medicine are discussed.

DLNLF-net: Denoised local and non-local deep features fusion network for malignancy characterization of hepatocellular carcinoma

INTRODUCTION

Hepatocellular carcinoma (HCC) is a primary liver cancer with high mortality rate. The degree of HCC malignancy is an important prognostic factor for predicting recurrence and survival after surgical resection or liver transplantation in clinical practice. Currently, deep features obtained from data-driven machine learning algorithms have demonstrated superior performance in characterising lesion features in medical imaging processing. However, previous convolutional neural network (CNN)-based studies on HCC lesion characterisation were based on traditional local deep features. The aim of this study was to propose a denoised local and non-local deep features fusion network (DLNLF-net) for grading HCC.

METHODS

Gadolinium-diethylenetriaminepentaacetic-acid-enhanced magnetic resonance imaging data of 117 histopathologically proven HCCs were collected from 112 patients with resected HCC between October 2012 and October 2018. The proposed DLNLF-net primarily consists of three modules: feature denoising, non-local feature extraction, and bilinear kernel fusion. First, local feature maps were extracted from the original tumour images using convolution operations, followed by a feature denoising block to generate denoised local features. Simultaneously, a non-local feature extraction block was employed on the local feature maps to generate non-local features. Finally, the two generated features were fused using a bilinear kernel model to output the classification results. The dataset was divided into a training set (77 HCC images) and an independent test set (40 HCC images). Training and independent testing were repeated five times to reduce measurement errors. Accuracy, sensitivity, specificity, and area under the curve (AUC) values in the five repetitive tests were calculated to evaluate the performance of the proposed method.

RESULTS

Denoised local features (AUC 89.19%) and non-local features (AUC 88.28%) showed better performance than local features (AUC 86.21%) and global average pooling features (AUC 87.1%) that were derived from a CNN for malignancy characterisation of HCC. Furthermore, the proposed DLNFL-net yielded superior performance (AUC 94.89%) than a typical 3D CNN (AUC 86.21%), bilinear CNN (AUC 90.46%), recently proposed local and global diffusion method (AUC 93.94%), and convolutional block attention module method (AUC 93.62%) for malignancy characterisation of HCC.

CONCLUSION

The non-local operation demonstrated a better capability of yielding global representation, and feature denoising based on the non-local operation achieved performance gains for lesion characterisation. The proposed DLNLF-net, which integrates denoised local and non-local deep features, evidently outperforms conventional CNN-based methods in the malignancy characterisation of HCC.

Texture appearance model, a new model-based segmentation paradigm, application on the segmentation of lung nodule in the CT scan of the chest

Lung cancer causes more than one million deaths worldwide each year. Averages of 5-year survival rate of patients with Non-small cell lung cancer (NSCLC), which is the most common type of lung cancer, is 15%. Computer-Aided Detection (CAD) is a very important tool for identifying lung lesions in medical imaging. In general, the process line of a CAD system can be divided into four main stages: preprocessing, localization, feature extraction, and classification. As segmentation is required for localization in computer vision and medical image analysis, this step has become a major and challenging problem, and much research has been done on new segmentation techniques. In recent years, interest in model-based segmentation methods has increased, and the reason for this is even if some object information is lost, such gaps can be filled by using the previous information in the model. This paper proposed Texture Appearance Model (TAM), which is a new model-based method and segments all types of nodule areas accurately and efficiently, including juxta-pleural nodules, without separating the lung from the surrounding area in a CT scan of the lung. In this method, Texture Representation of Image (TRI) is obtained using tissue texture feature extraction and feature selection algorithms. The proposed method has been evaluated in 85 nodules of the dataset, received from the Iranian hospital, in which the ground-truth annotation by physicians and CT imaging data were provided. The results show that the proposed algorithm has an encouraging performance for distinguishing different types of nodules, including pleural, cavity and non-solid nodules, achieving an average dice similarity coefficient (DSC) of 84.75%.

Development of computer-aided model to differentiate COVID-19 from pulmonary edema in lung CT scan: EDECOVID-net

The efforts made to prevent the spread of COVID-19 face specific challenges in diagnosing COVID-19 patients and differentiating them from patients with pulmonary edema. Although systemically administered pulmonary vasodilators and acetazolamide are of great benefit for treating pulmonary edema, they should not be used to treat COVID-19 as they carry the risk of several adverse consequences, including worsening the matching of ventilation and perfusion, impaired carbon dioxide transport, systemic hypotension, and increased work of breathing. This study proposes a machine learning-based method (EDECOVID-net) that automatically differentiates the COVID-19 symptoms from pulmonary edema in lung CT scans using radiomic features. To the best of our knowledge, EDECOVID-net is the first method to differentiate COVID-19 from pulmonary edema and a helpful tool for diagnosing COVID-19 at early stages. The EDECOVID-net has been proposed as a new machine learning-based method with some advantages, such as having simple structure and few mathematical calculations. In total, 13 717 imaging patches, including 5759 COVID-19 and 7958 edema images, were extracted using a CT incision by a specialist radiologist. The EDECOVID-net can distinguish the patients with COVID-19 from those with pulmonary edema with an accuracy of 0.98. In addition, the accuracy of the EDECOVID-net algorithm is compared with other machine learning methods, such as VGG-16 (Acc = 0.94), VGG-19 (Acc = 0.96), Xception (Acc = 0.95), ResNet101 (Acc = 0.97), and DenseNet20l (Acc = 0.97).

Validation of eyes-closed resting alpha amplitude predicting neurofeedback learning of upregulation alpha activity

Neurofeedback training (NFT) enables users to learn self-control of EEG activity of interest and then to create many benefits on cognitive function. A considerable number of nonresponders who fail to achieve successful NFT have often been reported in the within-session prediction. This study aimed to investigate successful EEG NFT of upregulation alpha activity in terms of trainability, independence, and between-session predictability validation. Forty-six participants completed 12 training sessions. Spectrotemporal analysis revealed the upregulation success on brain activity of 8-12 Hz exclusively to demonstrate trainability and independence of alpha NFT. Three learning indices of between-session changes exhibited significant correlations with eyes-closed resting state (ECRS) alpha amplitude before the training exclusively. Through a stepwise linear discriminant analysis, the prediction model of ECRS's alpha frequency band amplitude exhibited the best accuracy (89.1%) validation regarding the learning index of increased alpha amplitude on average. This study performed a systematic analysis on NFT success, the performance of the 3 between-session learning indices, and the validation of ECRS alpha activity for responder prediction. The findings would assist researchers in obtaining insight into the training efficacy of individuals and then attempting to adapt an efficient strategy in NFT success.

Standardization in Quantitative Imaging: A Multicenter Comparison of Radiomic Features from Different Software Packages on Digital Reference Objects and Patient Data Sets

Radiomic features are being increasingly studied for clinical applications. We aimed to assess the agreement among radiomic features when computed by several groups by using different software packages under very tightly controlled conditions, which included standardized feature definitions and common image data sets. Ten sites (9 from the NCI's Quantitative Imaging Network] positron emission tomography–computed tomography working group plus one site from outside that group) participated in this project. Nine common quantitative imaging features were selected for comparison including features that describe morphology, intensity, shape, and texture. The common image data sets were: three 3D digital reference objects (DROs) and 10 patient image scans from the Lung Image Database Consortium data set using a specific lesion in each scan. Each object (DRO or lesion) was accompanied by an already-defined volume of interest, from which the features were calculated. Feature values for each object (DRO or lesion) were reported. The coefficient of variation (CV), expressed as a percentage, was calculated across software packages for each feature on each object. Thirteen sets of results were obtained for the DROs and patient data sets. Five of the 9 features showed excellent agreement with CV < 1%; 1 feature had moderate agreement (CV < 10%), and 3 features had larger variations (CV ≥ 10%) even after attempts at harmonization of feature calculations. This work highlights the value of feature definition standardization as well as the need to further clarify definitions for some features.

Risks of feature leakage and sample size dependencies in deep feature extraction for breast mass classification

PURPOSE

Transfer learning is commonly used in deep learning for medical imaging to alleviate the problem of limited available data. In this work, we studied the risk of feature leakage and its dependence on sample size when using pretrained deep convolutional neural network (DCNN) as feature extractor for classification breast masses in mammography.

METHODS

Feature leakage occurs when the training set is used for feature selection and classifier modeling while the cost function is guided by the validation performance or informed by the test performance. The high-dimensional feature space extracted from pretrained DCNN suffers from the curse of dimensionality; feature subsets that can provide excessively optimistic performance can be found for the validation set or test set if the latter is allowed for unlimited reuse during algorithm development. We designed a simulation study to examine feature leakage when using DCNN as feature extractor for mass classification in mammography. Four thousand five hundred and seventy-seven unique mass lesions were partitioned by patient into three sets: 3222 for training, 508 for validation, and 847 for independent testing. Three pretrained DCNNs, AlexNet, GoogLeNet, and VGG16, were first compared using a training set in fourfold cross validation and one was selected as the feature extractor. To assess generalization errors, the independent test set was sequestered as truly unseen cases. A training set of a range of sizes from 10% to 75% was simulated by random drawing from the available training set in addition to 100% of the training set. Three commonly used feature classifiers, the linear discriminant, the support vector machine, and the random forest were evaluated. A sequential feature selection method was used to find feature subsets that could achieve high classification performance in terms of the area under the receiver operating characteristic curve (AUC) in the validation set. The extent of feature leakage and the impact of training set size were analyzed by comparison to the performance in the unseen test set.

RESULTS

All three classifiers showed large generalization error between the validation set and the independent sequestered test set at all sample sizes. The generalization error decreased as the sample size increased. At 100% of the sample size, one classifier achieved an AUC as high as 0.91 on the validation set while the corresponding performance on the unseen test set only reached an AUC of 0.72.

CONCLUSIONS

Our results demonstrate that large generalization errors can occur in AI tools due to feature leakage. Without evaluation on unseen test cases, optimistically biased performance may be reported inadvertently, and can lead to unrealistic expectations and reduce confidence for clinical implementation.

Comparison of segmentation-free and segmentation-dependent computer-aided diagnosis of breast masses on a public mammography dataset

PURPOSE

To compare machine learning methods for classifying mass lesions on mammography images that use predefined image features computed over lesion segmentations to those that leverage segmentation-free representation learning on a standard, public evaluation dataset.

METHODS

We apply several classification algorithms to the public Curated Breast Imaging Subset of the Digital Database for Screening Mammography (CBIS-DDSM), in which each image contains a mass lesion. Segmentation-free representation learning techniques for classifying lesions as benign or malignant include both a Bag-of-Visual-Words (BoVW) method and a Convolutional Neural Network (CNN). We compare classification performance of these techniques to that obtained using two different segmentation-dependent approaches from the literature that rely on specific combinations of end classifiers (e.g. linear discriminant analysis, neural networks) and predefined features computed over the lesion segmentation (e.g. spiculation measure, morphological characteristics, intensity metrics).

RESULTS

We report area under the receiver operating characteristic curve (AZ) values for malignancy classification on CBIS-DDSM for each technique. We find average AZ values of 0.73 for a segmentation-free BoVW method, 0.86 for a segmentation-free CNN method, 0.75 for a segmentation-dependent linear discriminant analysis of Rubber-Band Straightening Transform features, and 0.58 for a hybrid rule-based neural network classification using a small number of hand-designed features.

CONCLUSIONS

We find that malignancy classification performance on the CBIS-DDSM dataset using segmentation-free BoVW features is comparable to that of the best segmentation-dependent methods we study, but also observe that a common segmentation-free CNN model substantially and significantly outperforms each of these (p < 0.05). These results reinforce recent findings suggesting that representation learning techniques such as BoVW and CNNs are advantageous for mammogram analysis because they do not require lesion segmentation, the quality and specific characteristics of which can vary substantially across datasets. We further observe that segmentation-dependent methods achieve performance levels on CBIS-DDSM inferior to those achieved on the original evaluation datasets reported in the literature. Each of these findings reinforces the need for standardization of datasets, segmentation techniques, and model implementations in performance assessments of automated classifiers for medical imaging.

Deep-Learning-Based Computer-Aided Systems for Breast Cancer Imaging: A Critical Review

This paper provides a critical review of the literature on deep learning applications in breast tumor diagnosis using ultrasound and mammography images. It also summarizes recent advances in computer-aided diagnosis/detection (CAD) systems, which make use of new deep learning methods to automatically recognize breast images and improve the accuracy of diagnoses made by radiologists. This review is based upon published literature in the past decade (January 2010–January 2020), where we obtained around 250 research articles, and after an eligibility process, 59 articles were presented in more detail. The main findings in the classification process revealed that new DL-CAD methods are useful and effective screening tools for breast cancer, thus reducing the need for manual feature extraction. The breast tumor research community can utilize this survey as a basis for their current and future studies.

Prediction of Human Inhibition Brain Function with Inter-Subject and Intra-Subject Variability

The stop signal task has been used to quantify the human inhibitory control. The inter-subject and intra-subject variability was investigated under the inhibition of human response with a realistic environmental scenario. In present study, we used a battleground scenario where a sniper-scope picture was the background, a target picture was a go signal, and a nontarget picture was a stop signal. The task instructions were to respond on the target image and inhibit the response if a nontarget image appeared. This scenario produced a threatening situation and endorsed the evaluation of how subject's response inhibition manifests in a real situation. In this study, 32 channels of electroencephalography (EEG) signals were collected from 20 participants during successful stop (response inhibition) and failed stop (response) trials. These EEG signals were used to predict two possible outcomes: successful stop or failed stop. The inter-subject variability (between-subjects) and intra-subject variability (within-subjects) affect the performance of participants in the classification system. The EEG signals of successful stop versus failed stop trials were classified using quadratic discriminant analysis (QDA) and linear discriminant analysis (LDA) (i.e., parametric) and K-nearest neighbor classifier (KNNC) and Parzen density-based (PARZEN) (i.e., nonparametric) under inter- and intra-subject variability. The EEG activities were found to increase during response inhibition in the frontal cortex (F3 and F4), presupplementary motor area (C3 and C4), parietal lobe (P3 and P4), and occipital (O1 and O2) lobe. Therefore, power spectral density (PSD) of EEG signals (1-50Hz) in F3, F4, C3, C4, P3, P4, O1, and O2 electrodes were measured in successful stop and failed stop trials. The PSD of the EEG signals was used as the feature input for the classifiers. Our proposed method shows an intra-subject classification accuracy of 97.61% for subject 15 with QDA classifier in C3 (left motor cortex) and an overall inter-subject classification accuracy of 71.66% ± 9.81% with the KNNC classifier in F3 (left frontal lobe). These results display how inter-subject and intra-subject variability affects the performance of the classification system. These findings can be used effectively to improve the psychopathology of attention deficit hyperactivity disorder (ADHD), obsessive-compulsive disorder (OCD), schizophrenia, and suicidality.

Generalization error analysis for deep convolutional neural network with transfer learning in breast cancer diagnosis

Deep convolutional neural network (DCNN), now popularly called artificial intelligence (AI), has shown the potential to improve over previous computer-assisted tools in medical imaging developed in the past decades. A DCNN has millions of free parameters that need to be trained, but the training sample set is limited in size for most medical imaging tasks so that transfer learning is typically used. Automatic data mining may be an efficient way to enlarge the collected data set but the data can be noisy such as incorrect labels or even a wrong type of image. In this work we studied the generalization error of DCNN with transfer learning in medical imaging for the task of classifying malignant and benign masses on mammograms. With a finite available data set, we simulated a training set containing corrupted data or noisy labels. The balance between learning and memorization of the DCNN was manipulated by varying the proportion of corrupted data in the training set. The generalization error of DCNN was analyzed by the area under the receiver operating characteristic curve for the training and test sets and the weight changes after transfer learning. The study demonstrates that the transfer learning strategy of DCNN for such tasks needs to be designed properly, taking into consideration the constraints of the available training set having limited size and quality for the classification task at hand, to minimize memorization and improve generalizability.

Methods Used in Computer-Aided Diagnosis for Breast Cancer Detection Using Mammograms: A Review

According to the American Cancer Society's forecasts for 2019, there will be about 268,600 new cases in the United States with invasive breast cancer in women, about 62,930 new noninvasive cases, and about 41,760 death cases from breast cancer. As a result, there is a high demand for breast imaging specialists as indicated in a recent report for the Institute of Medicine and National Research Council. One way to meet this demand is through developing Computer-Aided Diagnosis (CAD) systems for breast cancer detection and diagnosis using mammograms. This study aims to review recent advancements and developments in CAD systems for breast cancer detection and diagnosis using mammograms and to give an overview of the methods used in its steps starting from preprocessing and enhancement step and ending in classification step. The current level of performance for the CAD systems is encouraging but not enough to make CAD systems standalone detection and diagnose clinical systems. Unless the performance of CAD systems enhanced dramatically from its current level by enhancing the existing methods, exploiting new promising methods in pattern recognition like data augmentation in deep learning and exploiting the advances in computational power of computers, CAD systems will continue to be a second opinion clinical procedure.

Data-mining Techniques for Image-based Plant Phenotypic Traits Identification and Classification

Statistical data-mining (DM) and machine learning (ML) are promising tools to assist in the analysis of complex dataset. In recent decades, in the precision of agricultural development, plant phenomics study is crucial for high-throughput phenotyping of local crop cultivars. Therefore, integrated or a new analytical approach is needed to deal with these phenomics data. We proposed a statistical framework for the analysis of phenomics data by integrating DM and ML methods. The most popular supervised ML methods; Linear Discriminant Analysis (LDA), Random Forest (RF), Support Vector Machine with linear (SVM-l) and radial basis (SVM-r) kernel are used for classification/prediction plant status (stress/non-stress) to validate our proposed approach. Several simulated and real plant phenotype datasets were analyzed. The results described the significant contribution of the features (selected by our proposed approach) throughout the analysis. In this study, we showed that the proposed approach removed phenotype data analysis complexity, reduced computational time of ML algorithms, and increased prediction accuracy.

Breast Cancer Detection and Diagnosis Using Mammographic Data: Systematic Review

BACKGROUND

Machine learning (ML) has become a vital part of medical imaging research. ML methods have evolved over the years from manual seeded inputs to automatic initializations. The advancements in the field of ML have led to more intelligent and self-reliant computer-aided diagnosis (CAD) systems, as the learning ability of ML methods has been constantly improving. More and more automated methods are emerging with deep feature learning and representations. Recent advancements of ML with deeper and extensive representation approaches, commonly known as deep learning (DL) approaches, have made a very significant impact on improving the diagnostics capabilities of the CAD systems.

OBJECTIVE

This review aimed to survey both traditional ML and DL literature with particular application for breast cancer diagnosis. The review also provided a brief insight into some well-known DL networks.

METHODS

In this paper, we present an overview of ML and DL techniques with particular application for breast cancer. Specifically, we search the PubMed, Google Scholar, MEDLINE, ScienceDirect, Springer, and Web of Science databases and retrieve the studies in DL for the past 5 years that have used multiview mammogram datasets.

RESULTS

The analysis of traditional ML reveals the limited usage of the methods, whereas the DL methods have great potential for implementation in clinical analysis and improve the diagnostic capability of existing CAD systems.

CONCLUSIONS

From the literature, it can be found that heterogeneous breast densities make masses more challenging to detect and classify compared with calcifications. The traditional ML methods present confined approaches limited to either particular density type or datasets. Although the DL methods show promising improvements in breast cancer diagnosis, there are still issues of data scarcity and computational cost, which have been overcome to a significant extent by applying data augmentation and improved computational power of DL algorithms.

Breast Cancer Diagnosis in Digital Breast Tomosynthesis: Effects of Training Sample Size on Multi-Stage Transfer Learning Using Deep Neural Nets

In this paper, we developed a deep convolutional neural network (CNN) for the classification of malignant and benign masses in digital breast tomosynthesis (DBT) using a multi-stage transfer learning approach that utilized data from similar auxiliary domains for intermediate-stage fine-tuning. Breast imaging data from DBT, digitized screen-film mammography, and digital mammography totaling 4039 unique regions of interest (1797 malignant and 2242 benign) were collected. Using cross validation, we selected the best transfer network from six transfer networks by varying the level up to which the convolutional layers were frozen. In a single-stage transfer learning approach, knowledge from CNN trained on the ImageNet data was fine-tuned directly with the DBT data. In a multi-stage transfer learning approach, knowledge learned from ImageNet was first fine-tuned with the mammography data and then fine-tuned with the DBT data. Two transfer networks were compared for the second-stage transfer learning by freezing most of the CNN structures versus freezing only the first convolutional layer. We studied the dependence of the classification performance on training sample size for various transfer learning and fine-tuning schemes by varying the training data from 1% to 100% of the available sets. The area under the receiver operating characteristic curve (AUC) was used as a performance measure. The view-based AUC on the test set for single-stage transfer learning was 0.85 ± 0.05 and improved significantly (p <; 0.05$ ) to 0.91 ± 0.03 for multi-stage learning. This paper demonstrated that, when the training sample size from the target domain is limited, an additional stage of transfer learning using data from a similar auxiliary domain is advantageous.

Gray-level invariant Haralick texture features

Haralick texture features are common texture descriptors in image analysis. To compute the Haralick features, the image gray-levels are reduced, a process called quantization. The resulting features depend heavily on the quantization step, so Haralick features are not reproducible unless the same quantization is performed. The aim of this work was to develop Haralick features that are invariant to the number of quantization gray-levels. By redefining the gray-level co-occurrence matrix (GLCM) as a discretized probability density function, it becomes asymptotically invariant to the quantization. The invariant and original features were compared using logistic regression classification to separate two classes based on the texture features. Classifiers trained on the invariant features showed higher accuracies, and had similar performance when training and test images had very different quantizations. In conclusion, using the invariant Haralick features, an image pattern will give the same texture feature values independent of image quantization.

Density estimation of grey-level co-occurrence matrices for image texture analysis

The Haralick texture features are common in the image analysis literature, partly because of their simplicity and because their values can be interpreted. It was recently observed that the Haralick texture features are very sensitive to the size of the GLCM that was used to compute them, which led to a new formulation that is invariant to the GLCM size. However, these new features still depend on the sample size used to compute the GLCM, i.e. the size of the input image region-of-interest (ROI). The purpose of this work was to investigate the performance of density estimation methods for approximating the GLCM and subsequently the corresponding invariant features. Three density estimation methods were evaluated, namely a piece-wise constant distribution, the Parzen-windows method, and the Gaussian mixture model. The methods were evaluated on 29 different image textures and 20 invariant Haralick texture features as well as a wide range of different ROI sizes. The results indicate that there are two types of features: those that have a clear minimum error for a particular GLCM size for each ROI size, and those whose error decreases monotonically with increased GLCM size. For the first type of features, the Gaussian mixture model gave the smallest errors, and in particular for small ROI sizes (less than about [Formula: see text]). In conclusion, the Gaussian mixture model is the preferred method for the first type of features (in particular for small ROIs). For the second type of features, simply using a large GLCM size is preferred.

Radiomics based detection and characterization of suspicious lesions on full field digital mammograms

BACKGROUND AND OBJECTIVE

Early detection is the important key to reduce breast cancer mortality rate. Detecting the mammographic abnormality as a subtle sign of breast cancer is essential for the proper diagnosis and treatment. The aim of this preliminary study is to develop algorithms which detect suspicious lesions and characterize them to reduce the diagnostic errors regarding false positives and false negatives.

METHODS

The proposed hybrid mechanism detects suspicious lesions automatically using connected component labeling and adaptive fuzzy region growing algorithm. A novel neighboring pixel selection algorithm reduces the computational complexity of the seeded region growing algorithm used to finalize lesion contours. These lesions are characterized using radiomic features and then classified as benign mass or malignant tumor using k-NN and SVM classifiers. Two datasets of 460 full field digital mammograms (FFDM) utilized in this clinical study consists of 210 images with malignant tumors, 30 with benign masses and 220 normal breast images that are validated by radiologists expert in mammography.

RESULTS

The qualitative assessment of segmentation results by the expert radiologists shows 91.67% sensitivity and 58.33% specificity. The effects of seven geometric and 48 textural features on classification accuracy, false positives per image (FPsI), sensitivity and specificity are studied separately and together. The features together achieved the sensitivity of 84.44% and 85.56%, specificity of 91.11% and 91.67% with FPsI of 0.54 and 0.55 using k-NN and SVM classifiers respectively on local dataset.

CONCLUSIONS

The overall breast cancer detection performance of proposed scheme after combining geometric and textural features with both classifiers is improved in terms of sensitivity, specificity, and FPsI.

Eyes-Closed Resting EEG Predicts the Learning of Alpha Down-Regulation in Neurofeedback Training

Neurofeedback training, which enables the trainee to learn self-control of the EEG activity of interest based on online feedback, has demonstrated benefits on cognitive and behavioral performance. Nevertheless, as a core mechanism of neurofeedback, learning of EEG regulation (i.e., EEG learning) has not been well understood. Moreover, a substantial number of non-learners who fail to achieve successful EEG learning have often been reported. This study investigated the EEG learning in alpha down-regulation neurofeedback, aiming to better understand the alpha learning and to early predict learner/non-learner. Twenty-nine participants received neurofeedback training to down-regulate alpha in two days, while eight of them were identified as non-learners who failed to reduce their alpha within sessions. Through a stepwise linear discriminant analysis, a prediction model was built based on participant's eyes-closed resting EEG activities in broad frequency bands including lower alpha, theta, sigma and beta 1 measured before training, which was validated in predicting learners/non-learners. The findings would assist in the early identification of the individuals who would not likely reduce their alpha during neurofeedback.

Label-free quantitative evaluation of breast tissue using Spatial Light Interference Microscopy (SLIM)

Breast cancer is the most common type of cancer among women worldwide. The standard histopathology of breast tissue, the primary means of disease diagnosis, involves manual microscopic examination of stained tissue by a pathologist. Because this method relies on qualitative information, it can result in inter-observer variation. Furthermore, for difficult cases the pathologist often needs additional markers of malignancy to help in making a diagnosis, a need that can potentially be met by novel microscopy methods. We present a quantitative method for label-free breast tissue evaluation using Spatial Light Interference Microscopy (SLIM). By extracting tissue markers of malignancy based on the nanostructure revealed by the optical path-length, our method provides an objective, label-free and potentially automatable method for breast histopathology. We demonstrated our method by imaging a tissue microarray consisting of 68 different subjects −34 with malignant and 34 with benign tissues. Three-fold cross validation results showed a sensitivity of 94% and specificity of 85% for detecting cancer. Our disease signatures represent intrinsic physical attributes of the sample, independent of staining quality, facilitating classification through machine learning packages since our images do not vary from scan to scan or instrument to instrument.

Multi-task transfer learning deep convolutional neural network: application to computer-aided diagnosis of breast cancer on mammograms

Transfer learning in deep convolutional neural networks (DCNNs) is an important step in its application to medical imaging tasks. We propose a multi-task transfer learning DCNN with the aim of translating the 'knowledge' learned from non-medical images to medical diagnostic tasks through supervised training and increasing the generalization capabilities of DCNNs by simultaneously learning auxiliary tasks. We studied this approach in an important application: classification of malignant and benign breast masses. With Institutional Review Board (IRB) approval, digitized screen-film mammograms (SFMs) and digital mammograms (DMs) were collected from our patient files and additional SFMs were obtained from the Digital Database for Screening Mammography. The data set consisted of 2242 views with 2454 masses (1057 malignant, 1397 benign). In single-task transfer learning, the DCNN was trained and tested on SFMs. In multi-task transfer learning, SFMs and DMs were used to train the DCNN, which was then tested on SFMs. N-fold cross-validation with the training set was used for training and parameter optimization. On the independent test set, the multi-task transfer learning DCNN was found to have significantly (p  =  0.007) higher performance compared to the single-task transfer learning DCNN. This study demonstrates that multi-task transfer learning may be an effective approach for training DCNN in medical imaging applications when training samples from a single modality are limited.

Urinary bladder cancer staging in CT urography using machine learning

PURPOSE

To evaluate the feasibility of using an objective computer-aided system to assess bladder cancer stage in CT Urography (CTU).

MATERIALS AND METHODS

A dataset consisting of 84 bladder cancer lesions from 76 CTU cases was used to develop the computerized system for bladder cancer staging based on machine learning approaches. The cases were grouped into two classes based on pathological stage ≥ T2 or below T2, which is the decision threshold for neoadjuvant chemotherapy treatment clinically. There were 43 cancers below stage T2 and 41 cancers at stage T2 or above. All 84 lesions were automatically segmented using our previously developed auto-initialized cascaded level sets (AI-CALS) method. Morphological and texture features were extracted. The features were divided into subspaces of morphological features only, texture features only, and a combined set of both morphological and texture features. The dataset was split into Set 1 and Set 2 for two-fold cross-validation. Stepwise feature selection was used to select the most effective features. A linear discriminant analysis (LDA), a neural network (NN), a support vector machine (SVM), and a random forest (RAF) classifier were used to combine the features into a single score. The classification accuracy of the four classifiers was compared using the area under the receiver operating characteristic (ROC) curve (Az ).

RESULTS

Based on the texture features only, the LDA classifier achieved a test Az of 0.91 on Set 1 and a test Az of 0.88 on Set 2. The test Az of the NN classifier for Set 1 and Set 2 were 0.89 and 0.92, respectively. The SVM classifier achieved test Az of 0.91 on Set 1 and test Az of 0.89 on Set 2. The test Az of the RAF classifier for Set 1 and Set 2 was 0.89 and 0.97, respectively. The morphological features alone, the texture features alone, and the combined feature set achieved comparable classification performance.

CONCLUSION

The predictive model developed in this study shows promise as a classification tool for stratifying bladder cancer into two staging categories: greater than or equal to stage T2 and below stage T2.

Urinary bladder cancer staging in CT urography using machine learning

PURPOSE

To evaluate the feasibility of using an objective computer-aided system to assess bladder cancer stage in CT Urography (CTU).

MATERIALS AND METHODS

A dataset consisting of 84 bladder cancer lesions from 76 CTU cases was used to develop the computerized system for bladder cancer staging based on machine learning approaches. The cases were grouped into two classes based on pathological stage ≥ T2 or below T2, which is the decision threshold for neoadjuvant chemotherapy treatment clinically. There were 43 cancers below stage T2 and 41 cancers at stage T2 or above. All 84 lesions were automatically segmented using our previously developed auto-initialized cascaded level sets (AI-CALS) method. Morphological and texture features were extracted. The features were divided into subspaces of morphological features only, texture features only, and a combined set of both morphological and texture features. The dataset was split into Set 1 and Set 2 for two-fold cross-validation. Stepwise feature selection was used to select the most effective features. A linear discriminant analysis (LDA), a neural network (NN), a support vector machine (SVM), and a random forest (RAF) classifier were used to combine the features into a single score. The classification accuracy of the four classifiers was compared using the area under the receiver operating characteristic (ROC) curve (A_z ).

RESULTS

Based on the texture features only, the LDA classifier achieved a test A_z of 0.91 on Set 1 and a test A_z of 0.88 on Set 2. The test A_z of the NN classifier for Set 1 and Set 2 were 0.89 and 0.92, respectively. The SVM classifier achieved test A_z of 0.91 on Set 1 and test A_z of 0.89 on Set 2. The test A_z of the RAF classifier for Set 1 and Set 2 was 0.89 and 0.97, respectively. The morphological features alone, the texture features alone, and the combined feature set achieved comparable classification performance.

CONCLUSION

The predictive model developed in this study shows promise as a classification tool for stratifying bladder cancer into two staging categories: greater than or equal to stage T2 and below stage T2.

A computer-aided diagnosis (CAD) scheme for pretreatment prediction of pathological response to neoadjuvant therapy using dynamic contrast-enhanced MRI texture features

OBJECTIVE

To assess whether a computer-aided, diagnosis (CAD) system can predict pathological Complete Response (pCR) to neoadjuvant chemotherapy (NAC) prior to treatment using texture features.

METHODS

Response to treatment of 44 patients was defined according to the histopatology of resected tumour and extracted axillary nodes in two ways: (a) pCR+ (Smith's Grade = 5) vs pCR- (Smith's Grade < 5); (b) pCRN+ (pCR+ and absence of residual lymph node metastases) vs pCRN - . A CAD system was developed to: (i) segment the breasts; (ii) register the DCE-MRI sequence; (iii) detect the lesion and (iv) extract 27 3D texture features. The role of individual texture features, multiparametric models and Bayesian classifiers in predicting patients' response to NAC were evaluated.

RESULTS

A cross-validated Bayesian classifier fed with 6 features was able to predict pCR with a specificity of 72% and a sensitivity of 67%. Conversely, 2 features were used by the Bayesian classifier to predict pCRN, obtaining a sensitivity of 69% and a specificity of 61%.

CONCLUSION

A CAD scheme, that extracts texture features from an automatically segmented 3D mask of the tumour, could predict pathological response to NAC. Additional research should be performed to validate these promising results on a larger cohort of patients and using different classification strategies. Advances in knowledge: This is the first study assessing the role of an automatic CAD system in predicting the pathological response to NAC before treatment. Fully automatic methods represent the backbone of standardized analysis and may help in timely managing patients candidate to NAC.

Mass detection in digital breast tomosynthesis: Deep convolutional neural network with transfer learning from mammography

PURPOSE

Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms.

METHODS

A data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were collected with IRB approval. The mass of interest on the images was marked by an experienced breast radiologist as reference standard. The data set was partitioned into a training set (2282 mammograms with 2461 masses and 230 DBT views with 228 masses) and an independent test set (94 DBT views with 89 masses). For DCNN training, the region of interest (ROI) containing the mass (true positive) was extracted from each image. False positive (FP) ROIs were identified at prescreening by their previously developed CAD systems. After data augmentation, a total of 45 072 mammographic ROIs and 37 450 DBT ROIs were obtained. Data normalization and reduction of non-uniformity in the ROIs across heterogeneous data was achieved using a background correction method applied to each ROI. A DCNN with four convolutional layers and three fully connected (FC) layers was first trained on the mammography data. Jittering and dropout techniques were used to reduce overfitting. After training with the mammographic ROIs, all weights in the first three convolutional layers were frozen, and only the last convolution layer and the FC layers were randomly initialized again and trained using the DBT training ROIs. The authors compared the performances of two CAD systems for mass detection in DBT: one used the DCNN-based approach and the other used their previously developed feature-based approach for FP reduction. The prescreening stage was identical in both systems, passing the same set of mass candidates to the FP reduction stage. For the feature-based CAD system, 3D clustering and active contour method was used for segmentation; morphological, gray level, and texture features were extracted and merged with a linear discriminant classifier to score the detected masses. For the DCNN-based CAD system, ROIs from five consecutive slices centered at each candidate were passed through the trained DCNN and a mass likelihood score was generated. The performances of the CAD systems were evaluated using free-response ROC curves and the performance difference was analyzed using a non-parametric method.

RESULTS

Before transfer learning, the DCNN trained only on mammograms with an AUC of 0.99 classified DBT masses with an AUC of 0.81 in the DBT training set. After transfer learning with DBT, the AUC improved to 0.90. For breast-based CAD detection in the test set, the sensitivity for the feature-based and the DCNN-based CAD systems was 83% and 91%, respectively, at 1 FP/DBT volume. The difference between the performances for the two systems was statistically significant (p-value < 0.05).

CONCLUSIONS

The image patterns learned from the mammograms were transferred to the mass detection on DBT slices through the DCNN. This study demonstrated that large data sets collected from mammography are useful for developing new CAD systems for DBT, alleviating the problem and effort of collecting entirely new large data sets for the new modality.

Texture features on T2-weighted magnetic resonance imaging: new potential biomarkers for prostate cancer aggressiveness

To explore contrast (C) and homogeneity (H) gray-level co-occurrence matrix texture features on T2-weighted (T2w) Magnetic Resonance (MR) images and apparent diffusion coefficient (ADC) maps for predicting prostate cancer (PCa) aggressiveness, and to compare them with traditional ADC metrics for differentiating low- from intermediate/high-grade PCas. The local Ethics Committee approved this prospective study of 93 patients (median age, 65 years), who underwent 1.5 T multiparametric endorectal MR imaging before prostatectomy. Clinically significant (volume ≥0.5 ml) peripheral tumours were outlined on histological sections, contoured on T2w and ADC images, and their pathological Gleason Score (pGS) was recorded. C, H, and traditional ADC metrics (mean, median, 10th and 25th percentile) were calculated on the largest lesion slice, and correlated with the pGS through the Spearman correlation coefficient. The area under the receiver operating characteristic curve (AUC) assessed how parameters differentiate pGS = 6 from pGS ≥ 7. The dataset included 49 clinically significant PCas with a balanced distribution of pGS. The Spearman ρ and AUC values on ADC were: -0.489, 0.823 (mean); -0.522, 0.821 (median); -0.569, 0.854 (10th percentile); -0.556, 0.854 (25th percentile); -0.386, 0.871 (C); 0.533, 0.923 (H); while on T2w they were: -0.654, 0.945 (C); 0.645, 0.962 (H). AUC of H on ADC and T2w, and C on T2w were significantly higher than that of the mean ADC (p = 0.05). H and C calculated on T2w images outperform ADC parameters in correlating with pGS and differentiating low- from intermediate/high-risk PCas, supporting the role of T2w MR imaging in assessing PCa biological aggressiveness.

A computer-aided diagnosis system for dynamic contrast-enhanced MR images based on level set segmentation and ReliefF feature selection

This study established a fully automated computer-aided diagnosis (CAD) system for the classification of malignant and benign masses via breast magnetic resonance imaging (BMRI). A breast segmentation method consisting of a preprocessing step to identify the air-breast interfacing boundary and curve fitting for chest wall line (CWL) segmentation was included in the proposed CAD system. The Chan-Vese (CV) model level set (LS) segmentation method was adopted to segment breast mass and demonstrated sufficiently good segmentation performance. The support vector machine (SVM) classifier with ReliefF feature selection was used to merge the extracted morphological and texture features into a classification score. The accuracy, sensitivity, and specificity measurements for the leave-half-case-out resampling method were 92.3%, 98.2%, and 76.2%, respectively. For the leave-one-case-out resampling method, the measurements were 90.0%, 98.7%, and 73.8%, respectively.

A method to determine the mammographic regions that show early changes due to the development of breast cancer

It is well understood nowadays that changes in the mammographic parenchymal pattern are an indicator of a risk of breast cancer and we have developed a statistical method that estimates the mammogram regions where the parenchymal changes, due to breast cancer, occur. This region of interest is computed from a score map by utilising the anatomical breast coordinate system developed in our previous work. The method also makes an automatic scale selection to avoid overfitting while the region estimates are computed by a nested cross-validation scheme. In this way, it is possible to recover those mammogram regions that show a significant difference in classification scores between the cancer and the control group. Our experiments suggested that the most significant mammogram region is the region behind the nipple and that can be justified by previous findings from other research groups. This result was conducted on the basis of the cross-validation experiments on independent training, validation and testing sets from the case-control study of 490 women, of which 245 women were diagnosed with breast cancer within a period of 2-4 years after the baseline mammograms. We additionally generalised the estimated region to another, mini-MIAS study and showed that the transferred region estimate gives at least a similar classification result when compared to the case where the whole breast region is used. In all, by following our method, one most likely improves both preclinical and follow-up breast cancer screening, but a larger study population will be required to test this hypothesis.

Changes in primary breast cancer heterogeneity may augment midtreatment MR imaging assessment of response to neoadjuvant chemotherapy

PURPOSE

To evaluate whether changes in magnetic resonance (MR) imaging heterogeneity may aid assessment for pathologic complete response (pCR) to neoadjuvant chemotherapy (NACT) in primary breast cancer and to compare pCR with standard Response Evaluation Criteria in Solid Tumors response.

MATERIALS AND METHODS

Institutional review board approval, with waiver of informed consent, was obtained for this retrospective analysis of 36 consecutive female patients, with unilateral unifocal primary breast cancer larger than 2 cm in diameter who were receiving sequential anthracycline-taxane NACT between October 2008 and October 2012. T2- and T1-weighted dynamic contrast material-enhanced MR imaging was performed before, at midtreatment (after three cycles), and after NACT. Changes in tumor entropy (irregularity) and uniformity (gray-level distribution) were determined before and after MR image filtration (for different-sized features). Entropy and uniformity for pathologic complete responders and nonresponders were compared by using the Mann-Whitney U test and receiver operating characteristic analysis.

RESULTS

With NACT, there was an increase in uniformity and a decrease in entropy on T2-weighted and contrast-enhanced subtracted T1-weighted MR images for all filters (uniformity: 23.45% and 22.62%; entropy: -19.15% and -19.26%, respectively). There were eight complete pathologic responders. An area under the curve of 0.84 for T2-weighted MR imaging entropy and uniformity (P = .004 and .003) and 0.66 for size (P = .183) for pCR was found, giving a sensitivity and specificity of 87.5% and 82.1% for entropy and 87.5% and 78.6% for uniformity compared with 50% and 82.1%, respectively, for tumor size change for association with pCR.

CONCLUSION

Tumors become more homogeneous with treatment. An increase in T2-weighted MR imaging uniformity and a decrease in T2-weighted MR imaging entropy following NACT may provide an earlier indication of pCR than tumor size change.

Mass type-specific sparse representation for mass classification in computer-aided detection on mammograms

Background

Breast cancer is the leading cause of both incidence and mortality in women population. For this reason, much research effort has been devoted to develop Computer-Aided Detection (CAD) systems for early detection of the breast cancers on mammograms. In this paper, we propose a new and novel dictionary configuration underpinning sparse representation based classification (SRC). The key idea of the proposed algorithm is to improve the sparsity in terms of mass margins for the purpose of improving classification performance in CAD systems.

Methods

The aim of the proposed SRC framework is to construct separate dictionaries according to the types of mass margins. The underlying idea behind our method is that the separated dictionaries can enhance the sparsity of mass class (true-positive), leading to an improved performance for differentiating mammographic masses from normal tissues (false-positive). When a mass sample is given for classification, the sparse solutions based on corresponding dictionaries are separately solved and combined at score level. Experiments have been performed on both database (DB) named as Digital Database for Screening Mammography (DDSM) and clinical Full Field Digital Mammogram (FFDM) DBs. In our experiments, sparsity concentration in the true class (SCTC) and area under the Receiver operating characteristic (ROC) curve (AUC) were measured for the comparison between the proposed method and a conventional single dictionary based approach. In addition, a support vector machine (SVM) was used for comparing our method with state-of-the-arts classifier extensively used for mass classification.

Results

Comparing with the conventional single dictionary configuration, the proposed approach is able to improve SCTC of up to 13.9% and 23.6% on DDSM and FFDM DBs, respectively. Moreover, the proposed method is able to improve AUC with 8.2% and 22.1% on DDSM and FFDM DBs, respectively. Comparing to SVM classifier, the proposed method improves AUC with 2.9% and 11.6% on DDSM and FFDM DBs, respectively.

Conclusions

The proposed dictionary configuration is found to well improve the sparsity of dictionaries, resulting in an enhanced classification performance. Moreover, the results show that the proposed method is better than conventional SVM classifier for classifying breast masses subject to various margins from normal tissues.

Computerized segmentation and characterization of breast lesions in dynamic contrast-enhanced MR images using fuzzy c-means clustering and snake algorithm

This paper presents a novel two-step approach that incorporates fuzzy c-means (FCMs) clustering and gradient vector flow (GVF) snake algorithm for lesions contour segmentation on breast magnetic resonance imaging (BMRI). Manual delineation of the lesions by expert MR radiologists was taken as a reference standard in evaluating the computerized segmentation approach. The proposed algorithm was also compared with the FCMs clustering based method. With a database of 60 mass-like lesions (22 benign and 38 malignant cases), the proposed method demonstrated sufficiently good segmentation performance. The morphological and texture features were extracted and used to classify the benign and malignant lesions based on the proposed computerized segmentation contour and radiologists' delineation, respectively. Features extracted by the computerized characterization method were employed to differentiate the lesions with an area under the receiver-operating characteristic curve (AUC) of 0.968, in comparison with an AUC of 0.914 based on the features extracted from radiologists' delineation. The proposed method in current study can assist radiologists to delineate and characterize BMRI lesion, such as quantifying morphological and texture features and improving the objectivity and efficiency of BMRI interpretation with a certain clinical value.

Semi-automated segmentation and classification of digital breast tomosynthesis reconstructed images

Digital breast tomosynthesis (DBT) is a limited-angle tomographic x-ray imaging technique that reduces the effect of tissue superposition observed in planar mammography. An integrated imaging platform that combines DBT with near infrared spectroscopy (NIRS) to provide co-registered anatomical and functional imaging is under development. Incorporation of anatomic priors can benefit NIRS reconstruction. In this work, we provide a segmentation and classification method to extract potential lesions, as well as adipose, fibroglandular, muscle and skin tissue in reconstructed DBT images that serve as anatomic priors during NIRS reconstruction. The method may also be adaptable for estimating tumor volume, breast glandular content, and for extracting lesion features for potential application to computer aided detection and diagnosis.