Browsing by Subject "Computer aided diagnosis"

Now showing 1 - 6 of 6

Open Access
2-D adaptive prediction based Gaussianity tests in microcalcification detection
(SPIE, 1998-01) Gürcan, M. Nafi; Yardımcı, Yasemin; Çetin, A. Enis
With increasing use of Picture Archiving and Communication Systems (PACS), Computer-aided Diagnosis (CAD) methods will be more widely utilized. In this paper, we develop a CAD method for the detection of microcalcification clusters in mammograms, which are an early sign of breast cancer. The method we propose makes use of two-dimensional (2-D) adaptive filtering and a Gaussianity test recently developed by Ojeda et al. for causal invertible time series. The first step of this test is adaptive linear prediction. It is assumed that the prediction error sequence has a Gaussian distribution as the mammogram images do not contain sharp edges. Since microcalcifications appear as isolated bright spots, the prediction error sequence contains large outliers around microcalcification locations. The second step of the algorithm is the computation of a test statistic from the prediction error values to determine whether the samples are from a Gaussian distribution. The Gaussianity test is applied over small, overlapping square regions. The regions, in which the Gaussianity test fails, are marked as suspicious regions. Experimental results obtained from a mammogram database are presented.
Open Access
Detection of microcalcifications in mammograms using nonlinear subband decomposition and outlier labeling
(SPIE, 1997-02) Gürcan, M. Nafi; Yardımcı, Yasemin C.; Çetin, A. Enis; Ansari, R.
Computer-aided diagnosis will be an important feature of the next generation picture archiving and communication systems. In this paper, computer-aided detection of microcalcifications in mammograms using a nonlinear subband decomposition and outlier labeling is examined. The mammogram image is first decomposed into subimages using a nonlinear subband decomposition filter bank. A suitably identified subimage is divided into overlapping square regions in which skewness and kurtosis as measures of the asymmetry and impulsiveness of the distribution are estimated. A region with high positive skewness and kurtosis is marked as a region of interest. Finally, an outlier labeling method is used to find the locations of microcalcifications in these regions. Simulation studies are presented.
Open Access
Expert advice ensemble for thyroid disease diagnosis
(IEEE, 2017) Qureshi, Muhammad Anjum; Ekşioğlu, Kubilay
Thyroid gland influences the metabolic processes of human body due to the fact that it produces hormones. Hyperthyroidism in caused due to increase in the production of thyroid hormones. In this paper a methodology using an online ensemble of decision trees to detect thyroid-related diseases is proposed. The aim of this work is to improve the diagnostic accuracy of thyroid disease. Initially, feature rejection method is applied to discard 10 irrelevant and redundant features from 29 features. Then, it's shown that the offline ensemble of decision trees provides higher performance than state-of-the-art methodologies. Afterwards, the exponential weights based online ensemble method is implemented which reaches comparable classification performance with offline methodology. The proposed system consists of three stages: feature rejection, training decision trees with different cost schemes and the online classification stage where each classifier is weighted using an exponential weight based algorithm. The performance of online algorithm increases as the number of samples increases, because it continuously updates the weights to improve accuracy. The achieved classification accuracy proves the robustness and effectiveness of online version of proposed system in thyroid disease diagnosis.
Open Access
From patch-level to ROI-level deep feature representations for breast histopathology classification
(SPIE, 2019) Mercan, Caner; Aksoy, Selim; Mercan, E.; Shapiro, L. G.; Weaver, D. L.; Elmore, J. G.; Tomaszewski, J. E.; Ward, A. D.
We propose a framework for learning feature representations for variable-sized regions of interest (ROIs) in breast histopathology images from the convolutional network properties at patch-level. The proposed method involves fine-tuning a pre-trained convolutional neural network (CNN) by using small fixed-sized patches sampled from the ROIs. The CNN is then used to extract a convolutional feature vector for each patch. The softmax probabilities of a patch, also obtained from the CNN, are used as weights that are separately applied to the feature vector of the patch. The final feature representation of a patch is the concatenation of the class-probability weighted convolutional feature vectors. Finally, the feature representation of the ROI is computed by average pooling of the feature representations of its associated patches. The feature representation of the ROI contains local information from the feature representations of its patches while encoding cues from the class distribution of the patch classification outputs. The experiments show the discriminative power of this representation in a 4-class ROI-level classification task on breast histopathology slides where our method achieved an accuracy of 66.8% on a data set containing 437 ROIs with different sizes.
Open Access
Influence function based Gaussianity tests for detection of microcalcifications in mammogram images
(IEEE, 1999-10) Gürcan, M. Nafi; Yardımcı, Y.; Çetin, A. Enis
In this paper, computer-aided diagnosis of microcalcifications in mammogram images is considered. Microcalcification clusters are an early sign of breast cancer. Microcalcifications appear as single bright spots in mammogram images. We propose an effective method for the detection of these abnormalities. The first step of this method is two-dimensional adaptive filtering. The filtering produces an error image which is divided into overlapping square regions. In each square region, a Gaussianity test is applied. Since microcalcifications have an impulsive appearance, they are treated as outliers. In regions with no microcalcifications, the distribution of the error image is almost Gaussian, on the other hand, in regions containing microcalcification clusters, the distribution deviates from Gaussianity. Using the theory of the influence function and sensitivity curves, we develop a Gaussianity test. Microcalcification clusters are detected using the Gaussianity test. Computer simulation studies are presented.
Open Access
Multi-instance multi-label learning for whole slide breast histopathology
(International Society for Optical Engineering SPIE, 2016-02-03) Mercan, Caner; Mercan, E.; Aksoy, Selim; Shapiro, L. G.; Weaver, D. L.; Elmore, J. G.
Digitization of full biopsy slides using the whole slide imaging technology has provided new opportunities for understanding the diagnostic process of pathologists and developing more accurate computer aided diagnosis systems. However, the whole slide images also provide two new challenges to image analysis algorithms. The first one is the need for simultaneous localization and classification of malignant areas in these large images, as different parts of the image may have different levels of diagnostic relevance. The second challenge is the uncertainty regarding the correspondence between the particular image areas and the diagnostic labels typically provided by the pathologists at the slide level. In this paper, we exploit a data set that consists of recorded actions of pathologists while they were interpreting whole slide images of breast biopsies to find solutions to these challenges. First, we extract candidate regions of interest (ROI) from the logs of pathologists' image screenings based on different actions corresponding to zoom events, panning motions, and fixations. Then, we model these ROIs using color and texture features. Next, we represent each slide as a bag of instances corresponding to the collection of candidate ROIs and a set of slide-level labels extracted from the forms that the pathologists filled out according to what they saw during their screenings. Finally, we build classifiers using five different multi-instance multi-label learning algorithms, and evaluate their performances under different learning and validation scenarios involving various combinations of data from three expert pathologists. Experiments that compared the slide-level predictions of the classifiers with the reference data showed average precision values up to 62% when the training and validation data came from the same individual pathologist's viewing logs, and an average precision of 64% was obtained when the candidate ROIs and the labels from all pathologists were combined for each slide. © 2016 SPIE.