Fully automatic and segmentation-robust classification of breast tumors based on local texture analysis of ultrasound images

Region of interest (ROI) is a region used to extract features. In breast ultrasound (BUS) image, the ROI is a breast tumor region. Because of poor image quality (low SNR (signal/noise ratio), low contrast, blurry boundaries, etc.), it is difficult to segment the BUS image accurately and produce a ROI which precisely covers the tumor region. Due to the requirement of accurate ROI for feature extraction, fully automatic classification of BUS images becomes a difficult task. In this paper, a novel fully automatic classification method for BUS images is proposed which can be divided into two steps: “ROI generation step” and “ROI classification step”. The ROI generation step focuses on finding a credible ROI instead of finding the precise tumor location. The ROI classification step employs a novel feature extraction and classification strategy. First, some points in the ROI are selected as the “classification checkpoints” which are evenly distributed in the ROI, and the local texture features around each classification checkpoint are extracted. For each ROI, all the classification checkpoints are classified. Finally, the class of the BUS image is determined by analyzing every classification checkpoint in the corresponding ROI. Both steps were implemented by utilizing a supervised texture classification approach. The experiments demonstrate that the proposed method is very robust to the segmentation of BUS images, and very effective and useful for classifying breast tumors.     

Computer-aided detection and classification of microcalcifications in mammograms: a survey

Breast cancer continues to be a significant public health problem in the world. Approximately, 182,000 new cases of breast cancer are diagnosed and 46,000 women die of breast cancer each year in the United States. Even more disturbing is the fact that one out of eight women in US will develop breast cancer at some point during her lifetime. Primary prevention seems impossible since the causes of this disease still remain unknown. Early detection is the key to improving breast cancer prognosis. Mammography is one of the reliable methods for early detection of breast carcinomas. There are some limitations of human observers, and it is difficult for radiologists to provide both accurate and uniform evaluation for the enormous number of mammograms generated in widespread screening. The presence of microcalcification clusters (MCCs) is an important sign for the detection of early breast carcinoma. An early sign of 30–50% of breast cancer detected mammographically is the appearance of clusters of fine, granular microcalcification, and 60–80% of breast carcinomas reveal MCCs upon histological examinations. The high correlation between the appearance of the microcalcification clusters and the diseases show that the CAD (computer aided diagnosis) systems for automated detection/classification of MCCs will be very useful and helpful for breast cancer control. In this survey paper, we summarize and compare the methods used in various stages of the computer-aided detection systems (CAD). In particular, the enhancement and segmentation algorithms, mammographic features, classifiers and their performances are studied and compared. Remaining challenges and future research directions are also discussed.     

Fully automatic segmentation of breast ultrasound images based on breast characteristics in space and frequency domains

Due to the complicated structure of breast and poor quality of ultrasound images, accurately and automatically locating regions of interest (ROIs) and segmenting tumors are challenging problems for breast ultrasound (BUS) computer-aided diagnosis systems. In this paper, we propose a fully automatic BUS image segmentation approach for performing accurate and robust ROI generation, and tumor segmentation. In the ROI generation step, the proposed adaptive reference point (RP) generation algorithm can produce the RPs automatically based on the breast anatomy; and the multipath search algorithm generates the seeds accurately and fast. In the tumor segmentation step, we propose a segmentation framework in which the cost function is defined in terms of tumor?s boundary and region information in both frequency and space domains. First, the frequency constraint is built based on the newly proposed edge detector which is invariant to contrast and brightness; and then the tumor pose, position and intensity distribution are modeled to constrain the segmentation in the spatial domain. The well-designed cost function is graph-representable and its global optimum can be found. The proposed fully automatic segmentation method is applied to a BUS database with 184 cases (93 benign and 91 malignant), and the performance is evaluated by the area and boundary error metrics. Compared with the newly published fully automatic method, the proposed method is more accurate and robust in segmenting BUS images.     

Cancer classification by gradient LDA technique using microarray gene expression data

Cancer classification is one of the major applications of the microarray technology. When standard machine learning techniques are applied for cancer classification, they face the small sample size (SSS) problem of gene expression data. The SSS problem is inherited from large dimensionality of the feature space (due to large number of genes) compared to the small number of samples available. In order to overcome the SSS problem, the dimensionality of the feature space is reduced either through feature selection or through feature extraction. Linear discriminant analysis (LDA) is a well-known technique for feature extraction-based dimensionality reduction. However, this technique cannot be applied for cancer classification because of the singularity of the within-class scatter matrix due to the SSS problem. In this paper, we use Gradient LDA technique which avoids the singularity problem associated with the within-class scatter matrix and shown its usefulness for cancer classification. The technique is applied on three gene expression datasets; namely, acute leukemia, small round blue-cell tumour (SRBCT) and lung adenocarcinoma. This technique achieves lower misclassification error as compared to several other previous techniques.     

Thermography based breast cancer analysis using statistical features and fuzzy classification

Medical thermography has proved to be useful in various medical applications including the detection of breast cancer where it is able to identify the local temperature increase caused by the high metabolic activity of cancer cells. It has been shown to be particularly well suited for picking up tumours in their early stages or tumours in dense tissue and outperforms other modalities such as mammography for these cases. In this paper we perform breast cancer analysis based on thermography, using a series of statistical features extracted from the thermograms quantifying the bilateral differences between left and right breast areas, coupled with a fuzzy rule-based classification system for diagnosis. Experimental results on a large dataset of nearly 150 cases confirm the efficacy of our approach that provides a classification accuracy of about 80%.     

深度学习在口腔医学影像中的应用与挑战

口腔医学影像是进行临床口腔疾病检测、筛查、诊断和治疗评估的重要工具,对口腔影像进行准确分析对于后续治疗计划的制定至关重要。常规的口腔医学影像分析依赖于医师的水平和经验,存在阅片效率低、可重复性低以及定量分析欠缺的问题。深度学习可以从大样本数据中自动学习并获取优良的特征表达,提升各类机器学习任务的效率和性能,目前已广泛应用于医学影像分析处理的各类任务之中。基于深度学习的口腔医学影像处理是目前的研究热点,但由于口腔医学领域内在的特殊性和复杂性,以及口腔医学影像数据样本量通常较小的问题,给深度学习方法在相关学习任务和场景的应用带来了新的挑战。本文从口腔医学影像领域常用的二维X射线影像、三维点云/网格影像和锥形束计算机断层扫描影像3种影像出发,介绍深度学习技术在口腔医学影像处理及分析领域应用的思路和现状,分析了各算法的优缺点及该领域所面临的问题和挑战,并对未来的研究方向和可能开展的临床应用进行展望,以助力智慧口腔建设。     

A novel method for pulmonary embolism detection in CTA images

In this paper, we propose a new computer-aided detection (CAD) – based method to detect pulmonary embolism (PE) in computed tomography angiography images (CTAI). Since lung vessel segmentation is the main objective to provide high sensitivity in PE detection, this method performs accurate lung vessel segmentation. To concatenate clogged vessels due to PEs, the starting region of PEs and some reference points (RPs) are determined. These RPs are detected according to the fixed anatomical structures. After lung vessel tree is segmented, the region, intensity, and size of PEs are used to distinguish them. We used the data sets that have heart disease or abnormal tissues because of lung disease except PE in this work. According to the results, 428 of 450 PEs, labeled by the radiologists from 33 patients, have been detected. The sensitivity of the developed system is 95.1% at 14.4 false positive per data set (FP/ds). With this performance, the proposed CAD system is found quite useful to use as a second reader by the radiologists.     

Road intersection detection and classification using hierarchical SVM classifier

In this paper, a hierarchical multi-classification approach using support vector machines (SVM) has been proposed for road intersection detection and classification. Our method has two main steps. The first involves the road detection. For this purpose, an edge-based approach has been developed using the bird’s eye view image which is mapped from the perspective view of the road scene. Then, the concept of vertical spoke has been introduced for road boundary form extraction. The second step deals with the problem of road intersection detection and classification. It consists on building a hierarchical SVM classifier of the extracted road forms using the unbalanced decision tree architecture. Many measures are incorporated for good evaluation of the proposed solution. The obtained results are compared to those of Choi et al. (2007).     

Plant-wide detection and diagnosis using correspondence analysis

This paper presents an approach based on the correspondence analysis (CA) for the task of fault detection and diagnosis. Unlike other data-based monitoring tools, such as principal components analysis/dynamic PCA (PCA/DPCA), the CA algorithm has been shown to use a different metric to represent the information content in the data matrix X. Decomposition of the information represented in the metric is shown here to yield superior performance from the viewpoints of data compression, discrimination and classification, as well as early detection and diagnosis of faults. Metrics similar to the contribution plots and threshold statistics that have been developed and used for PCA are also proposed in this paper for detection and diagnosis using the CA algorithm. Further, using the benchmark Tennessee Eastman problem as a case study, significant performance improvements are demonstrated in monitoring and diagnosis (in terms of shorter detection delays, smaller false alarm rates, reduced missed detection rates and clearer diagnosis) using the CA algorithm over those achievable using the PCA and DPCA algorithms.     

Microwave tomography for breast cancer detection on Cell broadband engine processors

Microwave tomography (MT) is a safe screening modality that can be used for breast cancer detection. The technique uses the dielectric property contrasts between different breast tissues at microwave frequencies to determine the existence of abnormalities. Our proposed MT approach is an iterative process that involves two algorithms: Finite-Difference Time-Domain (FDTD) and Genetic Algorithm (GA). It is a compute intensive problem: (i) the number of iterations can be quite large to detect small tumors; (ii) many fine-grained computations and discretizations of the object under screening are required for accuracy. In our earlier work, we developed a parallel algorithm for microwave tomography on CPU-based homogeneous, multi-core, distributed memory machines. The performance improvement was limited due to communication and synchronization latencies inherent in the algorithm. In this paper, we exploit the parallelism of microwave tomography on the Cell BE processor. Since FDTD is a numerical technique with regular memory accesses, intensive floating point operations and SIMD type operations, the algorithm can be efficiently mapped on the Cell processor achieving significant performance. The initial implementation of FDTD on Cell BE with 8 SPEs is 2.9 times faster than an eight node shared memory machine and 1.45 times faster than an eight node distributed memory machine. In this work, we modify the FDTD algorithm by overlapping computations with communications during asynchronous DMA transfers. The modified algorithm also orchestrates the computations to fully use data between DMA transfers to increase the computation-to-communication ratio. We see 54% improvement on 8 SPEs (27.9% on 1 SPE) for the modified FDTD in comparison to our original FDTD algorithm on Cell BE. We further reduce the synchronization latency between GA and FDTD by using mechanisms such as double buffering. We also propose a performance prediction model based on DMA transfers, number of instructions and operations, the processor frequency and DMA bandwidth. We show that the execution time from our prediction model is comparable (within 0.5 s difference) with the execution time of the experimental results on one SPE.     

Automated breast cancer detection by thermography : performance goal and diagnostic feature identification

This paper presents the methodology used for establishing a performance goal and identifying the diagnostic features in a program to develop an automated system for breast cancer detection based on thermographic principles. The receiver operating characteristic (ROC) curve approach is used to evaluate both observer classification and classification rules based on an observer's evaluation of diagnostic features. The multivariate logistic function is applied to two sets of observer evaluated feature sets using 623 normal and 122 breast cancer diagnosed subjects. It is shown that the observer outperforms the multivariate logistic function classifier based on the diagnostic features.     

Selecting features in microarray classification using ROC curves

We present a new method based on the ROC (Receiver Operating Characteristic) curve to efficiently select a feature subset in classifying a high-dimensional microarray dataset with a limited number of observations. Our method has two steps: (1) selecting the most relevant features to the target label using the ROC curve and (2) iteratively eliminating a redundant feature using the ROC curves. The ROC curve is strongly related with a non-parametric hypothesis testing, which must be effective for a dataset with small numerical observations. Experiments with real datasets revealed the significant performance advantage of our method over two competing feature subset selection methods.     

Computer-aided diagnosis of human brain tumor through MRI: A survey and a new algorithm

Computer-aided detection/diagnosis (CAD) systems can enhance the diagnostic capabilities of physicians and reduce the time required for accurate diagnosis. The objective of this paper is to review the recent published segmentation and classification techniques and their state-of-the-art for the human brain magnetic resonance images (MRI). The review reveals the CAD systems of human brain MRI images are still an open problem. In the light of this review we proposed a hybrid intelligent machine learning technique for computer-aided detection system for automatic detection of brain tumor through magnetic resonance images. The proposed technique is based on the following computational methods; the feedback pulse-coupled neural network for image segmentation, the discrete wavelet transform for features extraction, the principal component analysis for reducing the dimensionality of the wavelet coefficients, and the feed forward back-propagation neural network to classify inputs into normal or abnormal. The experiments were carried out on 101 images consisting of 14 normal and 87 abnormal (malignant and benign tumors) from a real human brain MRI dataset. The classification accuracy on both training and test images is 99% which was significantly good. Moreover, the proposed technique demonstrates its effectiveness compared with the other machine learning recently published techniques. The results revealed that the proposed hybrid approach is accurate and fast and robust. Finally, possible future directions are suggested.     

面向肺癌CAD系统的感兴趣区域特征选择与分类算法

本文重点研究ROI 的特征提取与分类方法．首先,根据医学征象对ROI 进行特征提取;为了提高分类的准确性,采用概率分布可分性对原始提取的特征进行特征选择．然后,利用SVM 对选择的特征进行定量描述;采用特征量化参数对Mahalanobis 距离进行加权改进,加权的Mahalanobis 距离使类间差别明显增大．最后采用加权改进后的Mahalanobis 距离将ROI 分类为结节或非结节．利用所提ROI 特征选择和分类算法进行肺结节检测实验;肺结节检测灵敏度为94.6%,漏诊率为5.4%,可以为医生进行肺癌早期诊断提供帮助信息．     

面向乳腺超声计算机辅助诊断的两阶段深度迁移学习

目的 为了提升基于单模态B型超声（B超）的乳腺癌计算机辅助诊断（computer-aided diagnosis,CAD）模型性能,提出一种基于两阶段深度迁移学习（two-stage deep transfer learning,TSDTL）的乳腺超声CAD算法,将超声弹性图像中的有效信息迁移至基于B超的乳腺癌CAD模型之中,进一步提升该CAD模型的性能。方法 在第1阶段的深度迁移学习中,提出将双模态超声图像重建任务作为一种自监督学习任务,训练一个关联多模态深度卷积神经网络模型,实现B超图像和超声弹性图像之间的信息交互迁移;在第2阶段的深度迁移学习中,基于隐式的特权信息学习（learning using privilaged information,LUPI）范式,进行基于双模态超声图像的乳腺肿瘤分类任务,通过标签信息引导下的分类进一步加强两个模态之间的特征融合与信息交互;采用单模态B超数据对所对应通道的分类网络进行微调,实现最终的乳腺癌B超图像分类模型。结果 实验在一个乳腺肿瘤双模超声数据集上进行算法性能验证。实验结果表明,通过迁移超声弹性图像的信息,TSDTL在基于B超的乳腺癌诊断任务中取得的平均分类准确率为87.84±2.08%、平均敏感度为88.89±3.70%、平均特异度为86.71±2.21%、平均约登指数为75.60±4.07%,优于直接基于单模态B超训练的分类模型以及多种典型迁移学习算法。结论 提出的TSDTL算法通过两阶段的深度迁移学习,将超声弹性图像的信息有效迁移至基于B超的乳腺癌CAD模型,提升了模型的诊断性能,具备潜在的应用可行性。     

基于Ncut分割和SVM分类器的医学图像分类算法

为解决医疗诊断中由于疲劳和主观因素影响导致的诊断错误,本文提出了基于Ncut分割方法的医学CT图像的分割、特征提取和诊断的新方案.将Ncut分割方法应用于脑CT图像.先进行图像分割,提取感兴趣区域,再从边缘、灰度,纹理三方面提取特征,最后利用支持向量机(SVM)对图像进行分类,为医生的诊断提供参考.从表格化的分类结果看,所提方案有较大的应用价值.     

A unified framework for fault detection and isolation of sensor and actuator biases in linear time invariant systems using marginalized likelihood ratio test with uniform priors

This study aims to present a fault detection and isolation (FDI) framework based on the marginalized likelihood ratio (MLR) approach using uniform priors for fault magnitudes in sensors and actuators. The existing methods in the literature use either flat priors with infinite support or the Gamma distribution as priors for the fault magnitudes. In the current study, it is assumed that the fault magnitude is a realization of a uniform prior with known upper and lower limits. The method presented in this study performs detection of time of occurrence of the fault and isolation of the fault type simultaneously while the estimation of the fault magnitude is achieved using a least squares based approach. The newly proposed method is evaluated by application to a benchmark CSTR problem using Monte Carlo simulations and the results reveal that this method can estimate the time of occurrence of the fault and the fault magnitude more accurately compared to a generalized likelihood ratio (GLR) based approach applied to the same benchmark problem. Simulation results on a benchmark problem also show significantly lower misclassification rates.     

A single dynamic observer-based module for design of simultaneous fault detection,isolation and tracking control scheme

The problem of simultaneous fault detection, isolation and tracking (SFDIT) control design for linear systems subject to both bounded energy and bounded peak disturbances is considered in this work. A dynamic observer is proposed and implemented by using the H_∞/H_?/L₁ formulation of the SFDIT problem. A single dynamic observer module is designed that generates the residuals as well as the control signals. The objective of the SFDIT module is to ensure that simultaneously the effects of disturbances and control signals on the residual signals are minimised (in order to accomplish the fault detection goal) subject to the constraint that the transfer matrix from the faults to the residuals is equal to a pre-assigned diagonal transfer matrix (in order to accomplish the fault isolation goal), while the effects of disturbances, reference inputs and faults on the specified control outputs are minimised (in order to accomplish the fault-tolerant and tracking control goals). A set of linear matrix inequality (LMI) feasibility conditions are derived to ensure solvability of the problem. In order to illustrate and demonstrate the effectiveness of our proposed design methodology, the developed and proposed schemes are applied to an autonomous unmanned underwater vehicle (AUV).     

Multiscaled combination of MR and SPECT images in neuroimaging: A simplex method based variable-weight fusion

Single-photon emission computed tomography (SPECT) images alone are difficult to understand in diagnosis, since anatomical structures are absent from the data. Studies on combination attempt to locate functional changes of the SPECT image by the magnetic resonance (MR) image. Due to the low similarity between original images, fused results are always darkened, obscured or loss some crucial anatomical structures. This paper has solved these problems by the variable-weight matrix which is estimated by minimizing the cost function using the simplex method. Under the generalized intensity-hue-saturation (GIHS) framework, the multiscaled analysis is presented for a better detail preservation. Besides, interactive approaches are discussed for the gradual variation between original images and the control of detail performance. The similarity assessment evaluates several different methods on a normal brain atlas. Two clips show the interactive property of the proposed method, while two medical cases demonstrate its clinical values. We conclude that the proposed method is superior to traditional methods, when considering the definition and the information capacity of fused results.