LV wall segmentation using the variational level set method (LSM) with additional shape constraint for oedema quantification

In this paper an automatic algorithm for the left ventricle (LV) wall segmentation and oedema quantification from T2-weighted cardiac magnetic resonance (CMR) images is presented. The extent of myocardial oedema delineates the ischaemic area-at-risk (AAR) after myocardial infarction (MI). Since AAR can be used to estimate the amount of salvageable myocardial post-MI, oedema imaging has potential clinical utility in the management of acute MI patients. This paper presents a new scheme based on the variational level set method (LSM) with additional shape constraint for the segmentation of T2-weighted CMR image. In our approach, shape information of the myocardial wall is utilized to introduce a shape feature of the myocardial wall into the variational level set formulation. The performance of the method is tested using real CMR images (12 patients) and the results of the automatic system are compared to manual segmentation. The mean perpendicular distances between the automatic and manual LV wall boundaries are in the range of 1-2?mm. Bland-Altman analysis on LV wall area indicates there is no consistent bias as a function of LV wall area, with a mean bias of?-121?mm(2)?between individual investigator one (IV1) and LSM, and?-122?mm(2)?between individual investigator two (IV2) and LSM when compared to two investigators. Furthermore, the oedema quantification demonstrates good correlation when compared to an expert with an average error of 9.3% for 69 slices of short axis CMR image from 12 patients.     

Ventricular structure,function, and focal fibrosis in anabolic steroid users: a CMR study

Purpose

Anabolic steroid (AS) misuse is widespread amongst recreational bodybuilders; however, their effects on the cardiovascular system are uncertain. Our aim was to document the impact of AS use on cardiac structure, function and the presence of focal fibrosis using the gold standard cardiovascular magnetic resonance imaging (CMR).

Methods

A cross-sectional cohort design was utilised with 21 strength-trained participants who underwent CMR imaging of the heart and speckle-tracking echocardiography. Thirteen participants (30 ± 5 years) taking AS for at least 2 years and currently on a “using”-cycle were compared with age and training-matched controls (n = 8; 29 ± 6 years) who self-reported never having taken AS (NAS).

Results

AS users had higher absolute left ventricular (LV) mass (220 ± 45 g) compared to NAS (163 ± 27 g; p < 0.05) but this difference was removed when indexed to fat-free mass. AS had a reduced right ventricular (RV) ejection fraction (AS 51 ± 4 % vs. NAS 59 ± 5 %; p < 0.05) and a significantly lower left ventricular E’:A’ myocardial tissue velocity ratio [AS 0.99(0.54) vs. NAS 1.78(0.46) p < 0.05] predominantly due to greater tissue velocities with atrial contraction. Peak LV longitudinal strain was lower in AS users (AS ?14.2 ± 2.7 % vs. NAS ?16.6 ± 1.9 %; p < 0.05). There was no evidence of focal fibrosis in any participant.

Conclusions

AS use was associated with significant LV hypertrophy, albeit in-line with greater fat-free mass, reduced LV strain, diastolic function, and reduced RV ejection fraction in male bodybuilders. There was, however, no evidence of focal fibrosis in any AS user.     

点对称距离模糊C均值聚类算法在脑部MRI图像分割中的应用

背景：在传统的图像分割方法中,模糊C均值聚类算法应用十分广泛。 目的：将改进的模糊C均值聚类算法应用到MRI图像的分割中,提高MRI图像分割的准确度。 方法：针对传统的基于Minkowski距离的模糊C均值聚类算法,提出了基于点对称距离的模糊C均值聚类算法,并将其运用到了脑部MRI图像分割中。 结果与结论：实验结果表明,与模糊C均值聚类算法相比,点对称距离的模糊C均值聚类算法有明显的优势。     

基于塔分割和多中心模糊聚类的医学图像分割

本文提出一种基于塔分割和多中心模糊C均值算法结合的无监督MR图像分割方法。文中采用根标记方法对塔图像进行过分割；在塔的最底层模糊刖像上应用HSC（hierarchical subtractive clustering）计算初始的聚类中心及聚类数。进而应用FCM算法合并过分割的结果。由于塔分割有效地降低了聚类样本数和HSC自动获得有效的初始聚类中心和聚类数，实验结果表明，在聚类性能不变情况下显著地减少FCM算法的运算时间，从而实现医学图像的快速分割。     

基于两阶段改进FCM算法的彩色血液图像分割研究

介绍了一种基于FCM算法（Fuzzy c-means algorithm）的彩色血液细胞图像分割新方法。通过将原始血液显微图像转换为索引图像再对颜色映射表矩阵做模糊聚类来回避直接对像素值聚类，大大压缩了FCM算法的计算数据量。针对彩色血液细胞图像的固有特性将分割过程分为两阶段，包括确定聚类数目和初始聚类中心，并在第二阶段引入一个距离加权矩阵，改变距离度量方式以提高聚类的准确度。解决了FCM算法难以准确收敛到最佳极值的问题，减少了迭代收敛的次数，并且缩短了算法执行时间，实现了对彩色血液细胞图像各组分的正确分割。     

A computerized volumetric segmentation method applicable to multi-centre MRI data to support computer-aided breast tissue analysis,density assessment and lesion localization

Density assessment and lesion localization in breast MRI require accurate segmentation of breast tissues. A fast, computerized algorithm for volumetric breast segmentation, suitable for multi-centre data, has been developed, employing 3D bias-corrected fuzzy c-means clustering and morphological operations. The full breast extent is determined on T1-weighted images without prior information concerning breast anatomy. Left and right breasts are identified separately using automatic detection of the midsternum. Statistical analysis of breast volumes from eighty-two women scanned in a UK multi-centre study of MRI screening shows that the segmentation algorithm performs well when compared with manually corrected segmentation, with high relative overlap (RO), high true-positive volume fraction (TPVF) and low false-positive volume fraction (FPVF), and has an overall performance of RO 0.94 ± 0.05, TPVF 0.97 ± 0.03 and FPVF 0.04 ± 0.06, respectively (training: 0.93 ± 0.05, 0.97 ± 0.03 and 0.04 ± 0.06; test: 0.94 ± 0.05, 0.98 ± 0.02 and 0.05 ± 0.07).     

基于FCM聚类算法的MRI脑组织图像分割方法比较研究

目的磁共振成像(magnetic resonance imaging,MRI)对脑组织有较好的成像效果,但噪声、偏移场和部分容积效应(partial volume effect,PVE)的存在,使得全自动分割MRI图像面临一定的困难。模糊C均值(fuzzy C-means,FCM)聚类算法在脑组织分割中得到较广泛研究。本文以存在噪声和偏移场影响的脑MRI图像分割为应用背景,研究了大量相关方法,探讨FCM算法分割脑部图像的改进思想。方法本文主要研究了9种FCM算法的理论基础,并通过脑组织分割实验对各种算法进行了分析。结果比较了不同算法的优劣,给出各类算法直观及定量评价结果。结论偏移场和噪声对脑磁共振图像组织分类质量有明显影响。其中几种方法可以减弱这些不利影响,但由于难以选择合适的参数,其分类效果并不理想。如何合理利用空间信息在未来仍有较大研究价值。     

基于数据融合的脑组织图像分割

随着医学影像技术的发展,我们可以用不同的成像方法对同一个脑断层得到多模态的核磁共振图像,针对脑组织分割的需要,文中介绍了一种基于数据融合的多模分割方法.算法首先用基于模糊C均值聚类(Fuzzy C-Means,FCM)的方法分别对单一模态的图像聚类进行分割,然后采用数据融合的方法得出最终的分割结果.实验结果表明,此方法能有效地分割出白质、灰质和脑脊液,并且分割精度要明显高于对单一模态图像的分割结果.     

Telomere length and adrenergic-induced left ventricular dilatation and systolic chamber dysfunction in rats

Purpose

The mechanisms responsible for telomere shortening in heart failure are uncertain. We evaluated whether left ventricular (LV) dilatation and systolic chamber dysfunction produced by chronic β-adrenergic receptor (β-AR) activation is associated with leukocyte or cardiac telomere shortening.

Methods

Following 6 months of daily injections of the β-AR agonist, isoproterenol (0.02 mg/kg/day) or the saline vehicle to rats, the extent of LV dilatation and LV systolic chamber dysfunction were determined using echocardiography and isolated perfused heart procedures, and relative telomere length of leukocyte (LTL) and cardiac (CTL) deoxyribonucleic acid were determined using a quantitative real-time polymerase chain reaction assay.

Results

β-AR activation resulted in LV dilatation as indexed by increased LV diastolic diameters (9.2 ± 0.6 vs. 8.4 ± 0.9 mm, P = 0.01) and increased diastolic volume intercepts at zero pressure of the LV diastolic pressure–volume relationship (isolated, perfused heart preparation) (0.40 ± 0.06 vs. 0.37 ± 0.08 ml, P = 0.03). Moreover, β-AR activation resulted in LV systolic chamber dysfunction as indexed by reductions in LV endocardial fractional shortening (0.40 ± 0.05 vs. 0.45 ± 0.06, P = 0.01) and the slope of the LV systolic pressure–volume relation (609 ± 176 vs. 901 ± 230, P = 0.01). Although LTL decreased with age in rats receiving either the β-AR agonist or the saline vehicle (P < 0.05), neither CTL (?0.10 ± 0.14 vs. ?0.15 ± 0.12, P = 0.3) nor LTL (?0.11 ± 0.19 vs. ?0.15 ± 0.18, P = 0.5) were modified by β-AR activation.

Conclusion

In conclusion, chronic β-AR activation sufficient to produce LV dilatation and systolic chamber dysfunction is not associated with alterations in leukocyte or cardiac telomere length. Telomere shortening in chronic heart failure is unlikely to be attributed to chronic β-AR activation.     

磁共振颅脑图像快速模糊聚类分割算法的研究

尽管模糊C-均值（简称FCM）聚类算法已广泛应用于图像分割研究,但是,由于模糊C-均值聚类算法所固有的一些缺点,特别是运算开销太大造成了该算法在实际应用中难以推广使用。根据模糊C-均值聚类算法和磁共振颅脑图象的特点,我们提出了一种分割磁共振颅脑图象的快速模糊C-均值（简称FFCM）聚类算法,该算法利用K-均值聚类结果指导模糊聚类的初始化,使模糊聚类的迭代次数明显减少。从而极大地提高模糊聚类的速度,实际应用表明,FFCM的分割速度比FCM快6.5倍以上,而分割精度与FCM相比无显著性差异。     

数据驱动的三叉神经纤维束自动分割算法

目前三叉神经的纤维跟踪成像过程中普遍存在人工依赖性问题,主要包括人工绘制感兴趣区域(ROI)及手动筛选目标纤维束,导致结果的不确定性和数据误差。针对此类问题,提出一种数据驱动的三叉神经纤维自动分割算法。利用多组大脑样本的纤维数据,建立数据驱动的纤维聚类图谱,实现新样本纤维数据的自动分割,直接得到三叉神经纤维束。在实验中,选择25组青年健康人的数据作为样本数据。首先,利用FSL软件分割工具提取脑干作为ROI,进行确定性纤维跟踪。其次,通过对20组纤维数据进行多样本配准和谱聚类,创建数据驱动的纤维聚类图谱。根据三叉神经细小的特点,在建立纤维图谱过程中,通过对脑干纤维束进行二次分类来标注三叉神经纤维束。最后,选择5组青年健康人的新样本数据,将其脑干纤维数据应用纤维图谱自动分割得到三叉神经纤维束,并计算同一样本数据的自动分割结果与手动分割结果之间的加权Dice系数。结果显示,所提出的方法成功分割5组数据的三叉神经纤维束,而传统人工方法成功识别4组三叉神经纤维束,两者结果之间的加权Dice系数分别为0.865,0.939,0.824,0.942。该方法可以有效避免人为因素的影响,提高神经外科医生...     

基于改进空间模糊聚类的DTI图像分割算法

针对模糊C均值(FCM)聚类算法初始聚类中心选择的随机性和噪声的敏感性等问题,提出一种基于改进空间模糊聚类的图像分割算法来分割人脑DTI图像。使用局部密度核函数和中心距离函数精确选取初始聚类中心,不仅可以解决因聚类中心随机选取造成的聚类效果不稳定的问题,而且还可以使目标函数迅速收敛,提高分割效率;通过将正态分布空间信息融入模糊隶属度函数,能减小图像噪声以及人为因素对分割结果的影响。用该方法与FCM、SFCM方法对人脑DTI数据进行分割,以评价算法的聚类效果。实验对美国明尼苏达大学生物医学功能成像与神经工程实验室提供的58例DTI数据、3例FA参数图像以及6例迭加过噪声的人脑DTI图像进行分割,结果表明:该算法分割系数最高,可达到0.984 1;在同一图像中,该算法在划分系数上比FCM最高提升20.2%,并且在划分熵上比SFCM最高下降19.8%;该算法目标函数平均迭代次数为32,较FCM的52次与空间FCM的76次有明显降低。实验证明,该算法能够准确、快速地分割出重要目标,且对图像噪声不敏感。     

基于空间FCM与MRF方法的乳腺MRI序列三维病灶分割研究

针对乳腺DCE-MRI病灶分割,提出一种空间FCM聚类与MRF随机场相结合的三维分割方法。首先,对MRI图像进行空间FCM粗分割,提取病灶粗轮廓。然后,在其基础上进行MRF精分割,并结合病灶三维信息：用相邻切片分割结果对应标号矩阵初始化MRF精分割标号场,同时用该张切片粗分割所得隶属度矩阵对MRF精分割进行参数自适应调整。用该方法与空间FCM、水平集、模糊MRF方法对50例MRI数据进行分割对比实验,得到良、恶性病灶分割重叠率分别为76.4、75.5;相比于空间FCM的68.%、69.5水平集的70.8、72.6以及模糊MRF的72.9、73.6有明显提升。对所有175例MRI数据分割结果进行非监督评价,得到良、恶性病灶区域均匀性均大于0.92;区域内差异性良性病灶92%小于150、恶性病灶98%小于150;区域间差异性良性病灶87%大于0.25、恶性病灶90%大于0.3综上表明,该方法具有较高的分割精度。     

Automatic segmentation of the left ventricle cavity and myocardium in MRI data

A novel approach for the automatic segmentation has been developed to extract the epi-cardium and endo-cardium boundaries of the left ventricle (lv) of the heart. The developed segmentation scheme takes multi-slice and multi-phase magnetic resonance (MR) images of the heart, transversing the short-axis length from the base to the apex. Each image is taken at one instance in the heart's phase. The images are segmented using a diffusion-based filter followed by an unsupervised clustering technique and the resulting labels are checked to locate the (lv) cavity. From cardiac anatomy, the closest pool of blood to the lv cavity is the right ventricle cavity. The wall between these two blood-pools (interventricular septum) is measured to give an approximate thickness for the myocardium. This value is used when a radial search is performed on a gradient image to find appropriate robust segments of the epi-cardium boundary. The robust edge segments are then joined using a normal spline curve. Experimental results are presented with very encouraging qualitative and quantitative results and a comparison is made against the state-of-the art level-sets method.     

Microscopic Cell Nuclei Segmentation Based on Adaptive Attention Window

This paper presents an adaptive attention window (AAW)-based microscopic cell nuclei segmentation method. For semantic AAW detection, a luminance map is used to create an initial attention window, which is then reduced close to the size of the real region of interest (ROI) using a quad-tree. The purpose of the AAW is to facilitate background removal and reduce the ROI segmentation processing time. Region segmentation is performed within the AAW, followed by region clustering and removal to produce segmentation of only ROIs. Experimental results demonstrate that the proposed method can efficiently segment one or more ROIs and produce similar segmentation results to human perception. In future work, the proposed method will be used for supporting a region-based medical image retrieval system that can generate a combined feature vector of segmented ROIs based on extraction and patient data.     

Tamper Localization and Lossless Recovery Watermarking Scheme with ROI Segmentation and Multilevel Authentication

Tamper localization and recovery watermarking scheme can be used to detect manipulation and recover tampered images. In this paper, a tamper localization and lossless recovery scheme that used region of interest (ROI) segmentation and multilevel authentication was proposed. The watermarked images had a high average peak signal-to-noise ratio of 48.7 dB and the results showed that tampering was successfully localized and tampered area was exactly recovered. The usage of ROI segmentation and multilevel authentication had significantly reduced the time taken by approximately 50 % for the tamper localization and recovery processing.     

A Fully Automatic Method for Lung Parenchyma Segmentation and Repairing

Considering that the traditional lung segmentation algorithms are not adaptive for the situations that most of the juxtapleural nodules, which are excluded as fat, and lung are not segmented perfectly. In this paper, several methods are comprehensively utilized including optimal iterative threshold, three-dimensional connectivity labeling, three-dimensional region growing for the initial segmentation of the lung parenchyma, based on improved chain code, and Bresenham algorithms to repair the lung parenchyma. The paper thus proposes a fully automatic method for lung parenchyma segmentation and repairing. Ninety-seven lung nodule thoracic computed tomography scans and 25 juxtapleural nodule scans are used to test the proposed method and compare with the most-cited rolling-ball method. Experimental results show that the algorithm can segment lung parenchyma region automatically and accurately. The sensitivity of juxtapleural nodule inclusion is 100 %, the segmentation accuracy of juxtapleural nodule regions is 98.6 %, segmentation accuracy of lung parenchyma is more than 95.2 %, and the average segmentation time is 0.67 s/frame. The algorithm can achieve good results for lung parenchyma segmentation and repairing in various cases that nodules/tumors adhere to lung wall.     

Computerized Classification of Pneumoconiosis on Digital Chest Radiography Artificial Neural Network with Three Stages

It is difficult for radiologists to classify pneumoconiosis from category 0 to category 3 on chest radiographs. Therefore, we have developed a computer-aided diagnosis (CAD) system based on a three-stage artificial neural network (ANN) method for classification based on four texture features. The image database consists of 36 chest radiographs classified as category 0 to category 3. Regions of interest (ROIs) with a matrix size of 32 × 32 were selected from chest radiographs. We obtained a gray-level histogram, histogram of gray-level difference, gray-level run-length matrix (GLRLM) feature image, and gray-level co-occurrence matrix (GLCOM) feature image in each ROI. For ROI-based classification, the first ANN was trained with each texture feature. Next, the second ANN was trained with output patterns obtained from the first ANN. Finally, we obtained a case-based classification for distinguishing among four categories with the third ANN method. We determined the performance of the third ANN by receiver operating characteristic (ROC) analysis. The areas under the ROC curve (AUC) of the highest category (severe pneumoconiosis) case and the lowest category (early pneumoconiosis) case were 0.89 ± 0.09 and 0.84 ± 0.12, respectively. The three-stage ANN with four texture features showed the highest performance for classification among the four categories. Our CAD system would be useful for assisting radiologists in classification of pneumoconiosis from category 0 to category 3.     

A new methodological approach to assess cardiac work by pressure–volume and stress–length relations in patients with aortic valve stenosis and dilated cardiomyopathy

In experimental animals, cardiac work is derived from pressure–volume area and analyzed further using stress–length relations. Lack of methods for determining accurately myocardial mass has until now prevented the use of stress–length relations in patients. We hypothesized, therefore, that not only pressure–volume loops but also stress–length diagrams can be derived from cardiac volume and cardiac mass as assessed by cardiac magnetic resonance imaging (CMR) and invasively measured pressure. Left ventricular (LV) volume and myocardial mass were assessed in seven patients with aortic valve stenosis (AS), eight with dilated cardiomyopathy (DCM), and eight controls using electrocardiogram (ECG)-gated CMR. LV pressure was measured invasively. Pressure–volume curves were calculated based on ECG triggering. Stroke work was assessed as area within the pressure–volume loop. LV wall stress was calculated using a thick-wall sphere model. Similarly, stress–length loops were calculated to quantify stress–length-based work. Taking the LV geometry into account, the normalization with regard to ventricular circumference resulted in “myocardial work.” Patients with AS (valve area 0.73 ± 0.18 cm²) exhibited an increased LV myocardial mass when compared with controls (P < 0.05). LV wall stress was increased in DCM but not in AS. Stroke work of AS was unchanged when compared with controls (0.539 ± 0.272 vs 0.621 ± 0.138 Nm, not significant), whereas DCM exhibited a significant depression (0.367 ± 0.157 Nm, P < 0.05). Myocardial work was significantly reduced in both AS and DCM when compared with controls (129.8 ± 69.6, 200.6 ± 80.1, 332.2 ± 89.6 Nm/m², P < 0.05), also after normalization (7.40 ± 5.07, 6.27 ± 3.20, 14.6 ± 4.07 Nm/m², P < 0.001). It is feasible to obtain LV pressure–volume and stress–length diagrams in patients based on the present novel methodological approach of using CMR and invasive pressure measurement. Myocardial work was reduced in patients with DCM and noteworthy also in AS, while stroke work was reduced in DCM only. Most likely, deterioration of myocardial work is crucial for the prognosis. It is suggested to include these basic physiological procedures in the clinical assessment of the pump function of the heart.     

A New Medical Image Watermarking Technique with Finer Tamper Localization

Medical imaging and information management systems require transmission of medical images over the Internet. Many image watermarking techniques have been proposed in recent years to ensure the integrity and authenticity of medical images transferred over insecure networks. In this work, we propose a new medical image watermarking technique to detect tampered regions on medical images with finer accuracy by authenticating 4 × 4 blocks and without restricting region of interest (ROI) size. The proposed method can mark a 4 × 4 pixel block if it has even one tampered pixel, while similar methods (which have no ROI size restriction) mark 8 × 8, 16 × 16, and 40 × 40 pixel blocks. Modified difference expansion (MDE) and least significant bit (LSB) embedding techniques are used together first in the literature by the method to embed authentication bits into corresponding blocks. The method uses a small 4 × 4 window to mark the tampered region. Experimental results show that the proposed method detects tampered regions on medical images with high accuracy and can be used by all medical image modalities. The results also indicate that the method has finer accuracy and no ROI size restriction compared to similar works reported in the literature.