Inês Domingues

INbreast: toward a full-field digital mammographic database.

RATIONALE AND OBJECTIVES Computer-aided detection and diagnosis (CAD) systems have been developed in the past two decades to assist radiologists in the detection and diagnosis of lesions seen on breast imaging exams, thus providing a second opinion. Mammographic databases play an important role in the development of algorithms aiming at the detection and diagnosis of mammary lesions. However, available databases often do not take into consideration all the requirements needed for research and study purposes. This article aims to present and detail a new mammographic database. MATERIALS AND METHODS Images were acquired at a breast center located in a university hospital (Centro Hospitalar de S. João [CHSJ], Breast Centre, Porto) with the permission of the Portuguese National Committee of Data Protection and Hospitals Ethics Committee. MammoNovation Siemens full-field digital mammography, with a solid-state detector of amorphous selenium was used. RESULTS The new database-INbreast-has a total of 115 cases (410 images) from which 90 cases are from women with both breasts affected (four images per case) and 25 cases are from mastectomy patients (two images per case). Several types of lesions (masses, calcifications, asymmetries, and distortions) were included. Accurate contours made by specialists are also provided in XML format. CONCLUSION The strengths of the actually presented database-INbreast-relies on the fact that it was built with full-field digital mammograms (in opposition to digitized mammograms), it presents a wide variability of cases, and is made publicly available together with precise annotations. We believe that this database can be a reference for future works centered or related to breast cancer imaging.

Assessment of a novel mass detection algorithm in mammograms

CONTEXT Mammography is the most effective procedure for an early detection of the breast abnormalities. Masses are a type of abnormality, which are very difficult to be visually detected on mammograms. AIMS In this paper an efficient method for detection of masses in mammograms is implemented. SETTINGS AND DESIGN The proposed mass detector consists of two major steps. In the first step, several suspicious regions are extracted from the mammograms using an adaptive thresholding technique. In the second step, false positives originating by the previous stage are reduced by a machine learning approach. MATERIALS AND METHODS All modules of the mass detector were assessed on mini-MIAS database. In addition, the algorithm was tested on INBreast database for more validation. RESULTS According to FROC analysis, our mass detection algorithm outperforms other competing methods. CONCLUSIONS We should not just insist on sensitivity in the segmentation phase because if we forgot FP rate, and our goal was just higher sensitivity, then the learning algorithm would be biased more toward false positives and the sensitivity would decrease dramatically in the false positive reduction phase. Therefore, we should consider the mass detection problem as a cost sensitive problem because misclassification costs are not the same in this type of problems.

Closed Shortest Path in the Original Coordinates with an Application to Breast Cancer

Breast cancer is one of the most mediated malignant diseases, because of its high incidence and prevalence, but principally due to its physical and psychological invasiveness. The study of this disease using computer science tools resorts often to the image segmentation operation. Image segmentation, although having been extensively studied, is still an open problem. Shortest path algorithms are extensively used to tackle this problem. There are, however, applications where the starting and ending positions of the shortest path need to be constrained, defining a closed contour enclosing a previously detected seed. Mass and calcification segmentation in mammograms and areola segmentation in digital images are two particular examples of interest within the field of breast cancer research. Usually the closed contour computation is addressed by transforming the image into polar coordinates, where the closed contour is transformed into an open contour between two opposite margins. In this work, after illustrating some of the limitations of this approach, we show how to compute the closed contour in the original coordinate space. After defining a directed acyclic graph appropriate for this task, we address the main difficulty in operating in the original coordinate space. Since small paths collapsing in the seed point are naturally favored, we modulate the cost of the edges to counterbalance this bias. A thorough evaluation is conducted with datasets from the breast cancer field. The algorithm is shown to be fast and reliable and suffers no loss in resolution.

Pectoral muscle detection in mammograms based on polar coordinates and the shortest path

The automatic detection and segmentation of the pectoral muscle in the medio-lateral oblique view of mammograms is essential for further analysis of breast anormalies. However, it is still a very difficult task since the sizes, shapes and intensity contrasts of pectoral muscles change greatly from image to image. In this paper, an algorithm based on the shortest path on a graph is proposed to automatically detect the pectoral muscle contour. To overcome the difficulties of searching for the path between a lateral and the top margins of the image, this is first transformed, using polar coordinates. In the transformed image, the muscle boundary in amongst the shortest paths between the top and the bottom rows. A comprehensive comparison with manually-drawn contours reveals the strength of the proposed method.

Pectoral muscle detection in mammograms based on the shortest path with endpoints learnt by SVMs

Automatic pectoral muscle removal on medio-lateral oblique view of mammogram is an essential step for many mammographic processing algorithms. However, the wide variability in the position of the muscle contour, together with the similarity between in muscle and breast tissues makes the detection a difficult task. In this paper, we propose a two step procedure to detect the muscle contour. In a first step, the endpoints of the contour are predicted with a pair of support vector regression models; one model is trained to predict the intersection point of the contour with the top row while the other is designed for the prediction of the endpoint of the contour on the left column. Next, the muscle contour is computed as the shortest path between the two endpoints. A comprehensive comparison with manually-drawn contours reveals the strength of the proposed method.

Ordinal Data Classification Using Kernel Discriminant Analysis: A Comparison of Three Approaches

Ordinal data classification (ODC) has a wide range of applications in areas where human evaluation plays an important role, ranging from psychology and medicine to information retrieval. In ODC the output variable has a natural order; however, there is not a precise notion of the distance between classes. The recently proposed method for ordinal data, Kernel Discriminant Learning Ordinal Regression (KDLOR), is based on Linear Discriminant Analysis (LDA), a simple tool for classification. KDLOR brings LDA to the forefront in the ODC held, motivating further research. This paper compares three LDA based algorithms for ODC. The first method uses the generic framework of Frank and Hall for ODC instantiated with a kernel version of LDA. Similarly, the second method is based on the also generic Data Replication framework for ODC instantiated with the same kernel version of LDA. Both the Frank and Hall and Data Replication methods address the ODC problem by the use of a base binary classifier. Finally, the third method under comparison is KDLOR. The experiments are carried out on synthetic and real datasets. A comparison between the performances of the three systems is made based on tstatistics. The performance and running time complexity of the methods do not support any advantage of KDLOR over the other two methods.

Normal breast identification in screening mammography: A study on 18 000 images

Through the years, several CAD systems have been developed to help radiologists in the hard task of detecting signs of cancer in the numerous screening mammograms. A more recent trend includes the development of pre-CAD systems aiming at identifying normal mammograms instead of detecting suspicious ones. Normal breasts are screened-out from the process, leaving radiologists more time to focus on more difficult cases. In this work, a new approach for the identification of normal breasts is presented. Considering that even breasts with malignant findings are mostly constituted by normal tissue, the breast area is divided into blocks which are then compared pairwise. If all blocks are very similar, the breast is labelled as normal, and as suspicious otherwise. Features characterizing the pairwise block similarity and characterizing the intra-block pixel distribution are used to design a predictive method based on machine learning techniques. The proposed solution was applied on a real world screening setting composed by nearly 18000 mammograms. Results are similar to the more complex state of the art approaches by correctly identifying more than 20% of the normal mammograms. These results suggest the usefulness of the relative comparison instead of the absolute classification. When properly used, simple statistics can suffice to distinguish the clearly normal breasts.

Max-Coupled Learning: Application to Breast Cancer

In the predictive modeling tasks, a clear distinction is often made between learning problems that are supervised or unsupervised, the first involving only labeled data (training patterns with known category labels) while the latter involving only unlabeled data. There is a growing interest in a hybrid setting, called semi-supervised learning, in semi-supervised classification, the labels of only a small portion of the training data set are available. The unlabeled data, instead of being discarded, are also used in the learning process. Motivated by a breast cancer application, in this work we address a new learning task, in-between classification and semi-supervised classification. Each example is described using two different feature sets, not necessarily both observed for a given example. If a single view is observed, then the class is only due to that feature set, if both views are present the observed class label is the maximum of the two values corresponding to the individual views. We propose new learning methodologies adapted to this learning paradigm and experimentally compare them with baseline methods from the conventional supervised and unsupervised settings. The experimental results verify the usefulness of the proposed approaches.

Using Bayesian Surprise to Detect Calcifications in Mammogram Images

Breast Cancer is still a serious health threat to women, both physically and psychologically. Fortunately, treatments involving complete breast removal are rarely needed today, as better treatment options are available. Mammography can show changes in the breast up to two years before a physician can feel them. Computer-aided detection and diagnosis is considered to be one of the most promising approaches that may improve the efficiency of mammography. Furthermore, there is a strong correlation between the presence of calcifications and the occurrence of breast cancer. In this paper we present a new technique to detect calcifications in mammogram images. The main objective is to support radiologists with automatic detection methods applied to medical images. Motivated by the fact that calcifications, when compared to the rest of the image, exhibit irregular characteristics, a technique based on Bayesian surprise is used. Tests were performed using INBreast, a recent fully annotated database, composed of full field digital mammograms. Comparison both with a recently proposed state of the art method and other common image techniques showed the superiority of our method. False positives are, however, still an issue and further studies focused on their reduction while maintaining a high sensitivity are planned.

Max-Ordinal Learning

In predictive modeling tasks, knowledge about the training examples is neither fully complete nor totally incomplete. Unlike semisupervised learning, where one either has perfect knowledge about the label of the point or is completely ignorant about it, here we address a setting where, for each example, we only possess partial information about the label. Each example is described using two (or more) different feature sets or views, where neither are necessarily observed for a given example. If a single view is observed, then the class is only due to that feature set; if more views are present, the observed class label is the maximum of the values corresponding to the individual views. After formalizing this new learning concept, we propose two new learning methodologies that are adapted to this learning paradigm. We also compare their instantiation in experiments with different base models and with conventional methods. The experimental results made both on real and synthetic data sets verify the usefulness of the proposed approaches.