Current Medical Imaging Reviews (v.12, #3)

Meet Our Editorial Board Member by David J. Sahn (157-158).

Investigation on the Detection of a Brain Disease by the Use of the Active Contour Methods by Mohamed H. Bendaoud, Noureddine Benabadji, Ahmed H. Belbachir (159-166).
The rapid development of medical imaging technology is revolutionizing medicine every day. Medical imaging allows scientists and doctors to disclose potentially vital information by scanning the human body non-invasively. The objective of this study is to isolate the pathology through the segmentation; it is considered the heart of medical imaging. Several methods have been proposed to visualize the evolution of the initial shape. In the present work, we propose two methods to define the outline of brain pathology using: Level Set without re-initialization and local region based method. The processed images are given by a magnetic resonance scanner (MRI) 1.5 tesla, of three patients. Knowing that the images are T2, T1 weighted. To give credibility to this study, a comparative study is implemented between the two methods studied. In the final analysis, we will reap the benefits of each method and the Downside.
We have shown that the evolution of the level set algorithm without re-initializing is faster than the algorithm of Local Region based method, but is still less accurate in the localization of the pathology. The evolution of the algorithm local region based methoda is very slow but much more accurate than the level set method without re-initialization. The only inconvenience is the requirement to initialize the curve C adjacent of pathology instead of taking the whole image.

Dual-Energy CT (DECT): A New Technique for Artifact Reduction from Metallic Orthopedic Implants by Laura Filograna, Nicola Magarelli, Michele La Torre, Luigi Pedone, Antonio Leone, Michael J. Thali, Lorenzo Bonomo (167-172).
Computed tomography (CT) plays an important role in the postoperative evaluation after metallic orthopedic implant placements for investigating complications (for example, implant fracture, faults, loosening and infections). Despite advances in CT technology, artefacts from metal implants can seriously impair CT image quality and diagnostic capabilities. In this article, basic principles of dual-energy CT (DECT) and its ability to reduce metal artefact are reviewed. Advantages and drawbacks of DECT in reducing metal artefacts in comparison with other single-energy CT approaches are also discussed. Although further systematic studies are needed to create optimized and standardized protocols, DECT with its monoenergetic reconstructions has to be considered a technique that opens new and effective possibilities for metallic artefact reduction in CT examinations.

An Intelligent Three-dimensional Ultrasound Program for Rapidly Imaging of the Fetal Cranial Mid-sagittal Plane by Yan Yi, Yi Xiong, Yao X. Zou, Mu Q. Lin, Qi Lin, Yang Jiao, Jin F. Xu (173-177).
This paper described a novel intelligent technique for rapid visualization of the fetal cranial mid-sagittal view to allow for the differentiation of fetal midline anomalies. Two hundred consecutive normal singleton pregnancies and twenty abnormal fetuses with cranial midline anomalies were imaged to display the mid-sagittal view using this new intelligent three-dimensional imaging program developed by our team. The cranial transverse plane was used as starting plane to acquire the threedimensional volumes and then scanned with this new program. The three-dimensional median planes were also evaluated by other two doctors. The reference landmarks of the mid-sagittal plane were that the falxs of fetal head on Plane A and Plane B were parallel to the X-axis and the reference dot was put on the falx. If one doctor thought it was not the median plane or the structure of corpus callosum or cerebellar vermis was not clearly visualized, the case was labelled as failed case. The cranial mid-sagittal view was successfully visualized in 190 normal cases (95%) and 18 abnormal cases (90%) by Smart MSP program. The failed 12 cases becuase the cerebral falxs of these fetuses were unable to be recognized by the program. In conclusion, this new intelligent three-dimensional program is a feasible method for quick visualization of fetal cranial mid-sagittal plane and may become a potential tool for routinely screening the fetal midline anomalies.

An Enhanced Hyper Spectral Image (HSI) Compression Based On Residual Dependent Arithmetic Coder (RDAC) by Senthivel Thiyagarajan, Gnanadurai Dhavamani, Somasundaram Malathi (178-188).
Background: The Hyper Spectral Image (HSI) compression is a challenging and demanding task in many remote sensing applications, because it has the large hyperspectral data. Optical remote sensing is much increased due to newly imported sensor technologies and advancements. Lossy HSI compression is an essential part for long-terms spectral storage data. In this paper, we provide a new lossy HSI compression algorithm with the help of Residual Dependent Arithmetic Coder (RDAC).
Methods: The main intention of this work is to reduce the complexity while compressing the large volume of data by compressing the spectral bands. Here, the Gray Level Co-occurrence Matrix (GLCM) technique is employed to extract the texture features of the given HSI band image. Then, the k-means clustering algorithm is employed to select the reference band in each cluster based on the cluster prominence value. Moreover, the RDAC is used to compress the reference band and the residual band information of each cluster. Finally, the HSI is decompressed with the help of compressed HSI band images.
Results: In experiments, the performance of the proposed method is analyzed and evaluated in terms of Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR) and Compression Ratio (CR). Moreover, it is compared with some of the existing HSI compression techniques such as, Set Partitioning in Hierarchical Trees (SPIHT), Joint Photographic Expert Group (JPEG), Set Partitioning Embedded bloCK (3D-SPECK), Inverse Wavelet Transform (IWT) and Reverse Karhunen-Loeve Transform (RKLT).
Conclusion: This paper proposes a new RDAC technique for lossy HSI compression. For this purpose, different image processing techniques are used. In this analysis, it is proved that the proposed HSI compression technique provides the best results, when compared to the other techniques.

Assessing Degree of Hepatic Fibrosis in Patients with Chronic Liver Disease Using Gadoxetic Acid-Enhanced Liver MRI: Preliminary Results by Stefano Palmucci, Giancarlo Attinà, Giuseppina Cappello, Maria L. Lanza, Giovanni Failla, Piero Costa, Rita O. A. Siverino, Claudia Vecchio, Sergio Neri, Pietro V. Foti, Pietro Milone, Gaetano Bertino, Giovanni C. Ettorre (189-195).
Purpose: To assess different degree of hepatic fibrosis using gadoxetic acid-enhanced MRI in patients with chronic liver disease.
Materials and Methods: 31 patients with chronic liver disease and 7 healthy subjects were studied using a 1.5 Tesla MRI. Unenhanced and dynamically enhanced images were acquired using gadoxetic-acid contrast (0.025 mmol/kg). Hepatobiliary phase was obtained with a 20 minute time delay. Liver relative enhancement (LRE) values were calculated sampling 1 cm Regions Of Interest (ROI) per segment. According to Metavir score, patients were divided into different groups: Group A = F0-F1 (absent/mild fibrosis), Group B = F2 (moderate fibrosis) and Group C = F3-F4 (advanced fibrosis). Mann-Whitney test with Bonferroni correction ad hoc was adopted to compare LRE values between classes. Receiver Operating Characteristics (ROC) curves were created for prediction of different degree of fibrosis.
Results: When comparing LRE mean values a statistical difference between F0-F1 and F3-F4 groups (p=0.008) was observed. A significant report was also observed between moderate and advanced fibrosis (p=0.02). In the comparison between group A and group B no significant difference for LRE mean values (p=0.93) was observed. For prediction of advanced fibrosis, ROC curve showed an Area Under Curve (AUC) of 0.793, with sensitivity of 94.12% and specificity of 57.89% using a threshold LRE value 1.0031. For prediction of 1.0031.
Conclusion: LRE values showed significant difference only between moderate and advanced fibrosis. In patients with chronic liver disease, the assessment of early stages of hepatic fibrosis (mild or moderate fibrosis) is not statistically reliable using gadoxetic acid-enhanced MRI.

With the rapid advancement of imaging technology, an inordinate amount of digital medical images have been generated in hospitals and medical institutions. To exploit those medical images in an effort to aid the diagnoses and research, content-based image retrieval systems are required to effectively access the medical image databases. This study presents a content-based image retrieval system which enables medical professionals to locate calcification lesions that are pathologically similar to a given example. More importantly, the type and distribution features of calcification lesions are extracted to represent the characteristics of mammographic lesions according to the Breast Imaging Reporting and Data System (BI-RADS), which is widely utilized by radiologists to describe mammographic lesions. In performance evaluation, a mammogram dataset was used to assess the effectiveness of the extracted type and calcification features. Our experimental results demonstrated that when the retrieval system only compares the calcification type and/or the calcification distribution characteristic, the pleomorphic type presents a higher precision-recall curve than the other three varieties of calcification types, and the cluster distribution performs best among the three lesion distributions. When the retrieval system takes both type and distribution characteristics into consideration, the pleomorphic and clustered class shows the best performance amongst all the calcification lesions.

Preoperative and Follow up Multi-Detector Row CT Angiography (MDCTA) in the Evaluation of Interrupted Aortic Arch (IAA) by Xiangmin Li, Zhenpeng Peng, Xuhui Zhou, Zhi Dong, Chaogui Yan, Bingsheng Huang, Min-Yi Cui, Shi-Ting Feng (205-212).
Purpose: The aim of our study is to investigate the ability of preoperative and follow-up multi-detector row CT angiography (MDCTA) in evaluating interrupted aortic arch (IAA).
Materials and Methods: MDCTA and echocardiography (ECHO) were performed preoperatively and postoperatively in nine patients (6 males and 3 females, average age: 4.7 years) with surgically confirmed IAA. The preoperative and follow-up images were analyzed retrospectively and compared with operative findings.
Results: Of the nine IAA cases, eight were type A, and one was type B. All cases were diagnosed correctly by MDCTA with 100% sensitivity. The MDCTA findings were highly concordant with the operative findings as well as with the various combined cardiac anomalies, including patent ductus arteriosus (PDA) (N=7); ventricular septal defect (VSD) (N=6); double outlet right ventricle and atrial septal defect (ASD) (N=1); aortopulmonary window and right pulmonary artery arising from the ascending aorta (N=1); and abnormal enlarged collateral vessels connecting with the descending aorta and enlarged internal mammary artery without PDA, ASD or VSD (N=2; ages 15 and 13 years old). Only five patients correctly diagnosed by ECHO with 56% sensitivity, two were misdiagnosed with coarctation of the aorta (CoA), and the remaining two were missed diagnoses. In the follow-up period, descending aorta saccular aneurysm combining thrombus at the distal end of the stoma was found in one case by MDCTA but missed by ECHO, and this MDCTA finding was further confirmed by operative findings. No postoperative complications were found in the other cases.
Conclusions: MDCTA can display the pathological anatomy of IAA and its combined malformations in a reliable manner. With an improved detection rate of malformations compared with that of ECHO, MDCTA may have high diagnostic value in detecting IAA and its postoperative complications.

Feature Based Fusion of Multimodal Medical Image Slices with Combined Transforms by Kavitha C. Thankam, Chellamuthu Chinnagounder (213-219).
Medical image fusion plays an important role in radiological practice, helping the doctors in exact diagnosis and medical treatment planning. The proposed feature based fusion of medical image slices in combined transform improves the quality of the fused image. The image is initially analysed using integer wavelet transform and then by discrete ripplet transform. Significant features of the low pass and the high pass regions are extracted. Decision rules are framed based on the features extracted, and the fused coefficients are obtained. Then, the fused coefficients are synthesized by applying the inverse transforms. In the proposed work, a generalised algorithm is applied to different pairs of medical images like i) CT and MRI image slices, ii) PET and MRI image slices and iii) SPECT and MRI image slices. The fused image is validated, and the quantitative results show that the proposed method is better than the existing methods and it has better visual quality with clear edges.

Assessment of Gastroesophageal Reflux Disease in Chronic Obstructive Pulmonary Disease: Is Measuring Cardioesophageal Angle on Multidetector Computed Tomography Images Useful? by Ebru Unlu, Sevinc S. Ulasli, Emre Kacar, Mehtap B. Acay, Serife Ozdinc, Nesrin Atci, Cinar Balcik, Muzaffer Sariaydin, Ersin Gunay (220-224).
We aimed to investigate whether there were any differences between patients with chronic obstructive pulmonary disease (COPD) and controls according to their cardioesophageal angles (COA) by Computed tomography (CT) and also the relationship between the degrees of COA and pulmonary function values in COPD patients. Degrees of COA were measured on curved oblique coronal CT images of 198 COPD patients with gastroesophageal reflux (GER) and 298 control subjects. A comparative analysis of COA values in two groups was performed and possible correlations between degrees of COA and pulmonary function parameters were evaluated. COA values were significantly higher in COPD patients compared to controls (p< 0.001). The degrees of COA was negatively correlated with FVC and FEV1 values (% and L) in COPD patients (p<0.001, r=-0.579; p<0.001, r=-0.64 and p<0.001, r=-0.65; p<0.001, r=-0.615, respectively). A positive correlation between age and COA was found in all the study population (p=0.002; r=0.134). A negative correlation between COA and BMI was found in COPD patients (p= 0.037, r=-0.149). In conclusion, COPD patients have significantly increased COA value which is a predisposing factor for GER. Also, the association between obtuse COA and decreased levels of pulmonary function parameters in COPD patients may indicate reflux-induced bronchoconstriction and should be considered in COPD management as a non invasive indicator of reflux.

Lung cancer is one of the most common lethal type of diseases. One of the most important and difficult tasks a doctor has to carry out is the detection and diagnosis of cancerous lung cells from the Computed Tomography (CT) images result. Segmentation and classification of lung CT image, based on soft computing, is still a challenging task in the medical field, due to more computational time and accuracy. This paper deals with an improvement in lung cancer detection using Possibilistic Fuzzy C-Means (PFCM) based segmentation. This work also focuses on the normal and abnormal cancer cells that is classified by using the algorithms of SVM (Support Vector Machine), Gaussian Interval Type II Fuzzy Logic System and Genetic Algorithm (SVMFLGA). The results demonstrate that the SVMFLGA outperforms the Gaussian Interval type II fuzzy logic system (GAIT2FLS) in terms of classification accuracy.