Current Medical Imaging - Volume 12, Issue 4, 2016

Diabetic retinopathy (DR), glaucoma and hypertension are the most common forms of retinal diseases and diagnosis of these diseases at their early stages can prevent people from blindness. In practice, computer aided detection (CAD) systems are being developed for the detection of these eye diseases. These CAD systems provide second opinion to less experienced opthalmologists for the detection and analysis of these diseases using digital fundus images. In this paper, the state-of-the-art CAD systems are studied and compared systematically in terms of their classification accuracy, methodological approach and efficiency. The comparison results indicate that the accuracy of these CAD systems is not up to the mark.

Papillary breast tumors are rare breast tumors. Presentation of papillary breast lesions varies clinically and radiologically. Standard diagnostic work-up for papillary breast lesions includes various radiological modalities such as mammography, galactography, ultrasound and MRI. Papillary lesions often have a wide spectrum of appearance on different radiological modalities, so that optimal differentiation of papillary lesions is not easy with various imaging methods. The purpose of this review article is to describe the different imaging appearances of benign and malignant papillary lesions of the breast with special emphasis on atypical appearances.

As the increasing incidence of thyroid nodules has been recently reported, differential diagnosis between malignancy and benign nodules plays a critical role for determination of therapy plans, but is still a challenge for clinicians. Because of uncertain results of fine needle aspiration (FNA) cytology in some cases, medical images give great noninvasive contribution to differential diagnosis of thyroid nodules. This review of current literature focused on the development of various image modalities on thyroid nodules, in order to supply essentials to clinicians for making pretreatment determination. When ultrasonography is recognized as a routine tool for detecting and differentiating thyroid nodules, improved accuracy has been found when it is combined with FNA. The newly-developed modalities of elastography and contrast enhanced ultrasonography could be used for evaluating the nodule stiffness and vascularity, respectively. Although not recommended as the first-line tool in the managment guideline, scintigraphy supplies supplementary information of the indeterminate nodules after FNA, especially in follicular thyroid nodules. Computed tomography( CT)-detected calcification and its patterns have been regarded to be important sign related to malignancy. Apparent diffusion coefficient ratios on diffusion-weighted magnetic resonance image( MRI) are detected significantly lower in malignant nodules than those in benign ones. CT and MRI are also preferred to demonstrating the surrounding anatomic structures and lymph nodes. The role of PET and PET/CT in the assessment of nodules has been increased, while SUVmax is attractive with a cutoff value about 6.0. In conclusion, with their respective functions, the medical image modalities could supply much helpful information for differential diagnosis of thyroid nodules.

The increasing incidence of neurodegenerative disorders with dementia (NDD) requires structural and functional imaging methods which are easily available in the clinical setting. The diagnostic work-up for patients who have NDD includes magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT), and positron emission tomography (PET). Both, MRI and SPECT can show characteristic findings of NDD. However, the interpretation of these findings has to be done in the context of clinical data, which includes biochemical and histological examinations. In this article, we review recent advances in pathology and the impact of these advances on MRI and SPECT studies interpretations. We present a literature review describing typical MRI and SPECT findings and the diagnostic accuracies reported for NDD. Combining MRI and SPECT examinations with quantitative evaluation of clinical data further improves the diagnostic potential of these imaging procedures for NDD.

Elastography is relatively a new diagnostic modality that is being used in the field of medicine. There are 2 basic types of elastography techniques; strain elastography and shear-wave elastography. This review details the principle, its applications and draws differences between the two imaging modalities. Literature from PubMed was searched within the year 2015 only, that contain the search terms ‘strain elastography’ and ‘shear wave elastography’ individually. Articles were carefully selected that must cover at least one application of Liver, Breast, Thyroid, Gastrointestinal tract, Prostate and Musculoskeletal. This review opens a new insight into comparative studies for strain elastography and shear wave elastography techniques, as limited data is available as how these two imaging diagnostic modalities behave under same circumstances.

In recent years, fluorine 18 fluorodeoxyglucose (18F-FDG) positron emission tomography (PET)/computed tomography (CT) has emerged as a new modality for detecting the primary site of cervical metastases from unknown primary carcinoma (UPC). The aim of this meta-analysis was to assess the diagnostic value of 18F-FDG PET/CT in detecting primary tumors in UPC patients with cervical metastases. After systematic search for eligible studies and data extraction, we determined pooled sensitivities and specificities across studies, calculated positive and negative likelihood ratios, and constructed summary receiver operating characteristic curves with area under the curve (AUC) and Q* obtained. We also analyzed the heterogeneity between studies based on subgroup-analysis and publication bias. Totally, 39 studies involving 1468 patients met the inclusion criteria. The pooled sensitivity and specificity were 0.90 (95% confidence interval [CI], 0.87-0.92) and 0.79 (95% CI, 0.75- 0.81), respectively. Likelihood ratio syntheses yielded overall PLR of 3.50 (95% CI, 3.00-4.08) and NLR of 0.20 (95% CI, 0.15-0.26). The AUC and Q* index were 0.8930 and 0.8238, respectively. The heterogeneity was only significantly observed in sensitivity. PET/CT is beneficial in the overall assessment of primary tumors in UPC patients with cervical metastases. Large, multicenter, and prospective studies with standard protocols are now needed to investigate the true value of PET/CT for detecting primary tumors of cervical metastases from UPC and the broad application of this method in clinical practice.

Cancer remains one of the major concerns of deaths worldwide. Early detection is the key point in reducing the cancer mortality. Automatic systems are needed to assist radiologists in the cancer detection and diagnosis. Hence, there are strong needs for the development of computer aided diagnosis (CAD) systems which have the capability to help radiologists in decision making. The aim of this work is to develop a computer aided system for breast cancer diagnosis in ultrasound images. The developed system consists of segmentation, feature extraction, feature selection and classification. The marker controlled watershed technique is used to segment the region of interest (ROI). In the feature extraction step, the wavelet transform is applied then the texture and statistical features of ROI are extracted. In addition, a set of morphological features are extracted directly from ROI in spatial domain. The obtained features are combined together to produce the feature vector. In order to select the most discriminative feature, a feature ranking technique is used to determine the capability of each feature. In the classification step, support vector machine (SVM), classification and regression trees (CART) and classification rule classifiers are used to classify the ROI as benign or malignant. The proposed method is validated using 10 fold cross-validation. The results show that classification rule classifier outperforms SVM and CART classifiers.

Quantum noise are more likely to occur in mammographic images and it effects the accuracy of classification. In this paper, Genetic Programming (GP) based novel hybrid ensemble method designed for noise amputation from breast mammogram images has been proposed. Three different filters Frost filter, Weiner filter and Non Local Means has been fused by using Genetic Programming for noise removal. The fused GP based filter calculate the value by combining three filters and replacing them with the noisy pixels. Peak signal to noise ratio (PSNR) and Structure Similarity Index measure (SSIM) as a quantitative measures has been employed for evaluation of proposed method with different existing methods. Experimental results show that proposed method bring forward better results as compared to existing methods.

In medical image processing, image fusion technique is used to enhance the brain tumors or inertial component of the brain for better medical diagnosis and further clinical treatment. In this paper, the brain tumor is detected and diagnosed by the following stages; preprocessing, fuzzy logic based fusion, feature extraction, Genetic algorithm and classification. Mamdani Fuzzy rules are constructed and used for brain tumor enhancement. Local binary and ternary pattern are extracted from the fused image and best features are selected by genetic algorithm. The extracted features are trained and classified into normal or abnormal brain image by feed forward back propagation neural networks. Morphological operations are used to segment the brain tumor from the classified brain image. The methodology presented in this paper is tested over the images available from the public datasets. The proposed system achieved the sensitivity rate of 99.67%, specificity rate of 99.56% and accuracy of 98.75%.

Objectives: Soft tissue attenuation artifacts are the most common cause of false-positives in myocardial perfusion SPECT (MPS). Few studies assessing the value of prone imaging in women have been published. Our study evaluated the impact of prone position imaging on the interpretation of MPS scans of women with suspect coronary artery disease (CAD) and also defined the relationship between breast attenuation, age, bra cup size and body mass index (BMI). Methods: MPS scans of women with suspected CAD (n=431) were retrospectively analyzed by two blinded experts not aware if post-stress images were acquired in supine or prone position. After semi-quantitative analysis summed stress, rest and difference scores (SSS, SRS and SDS, respectively) were calculated and scans were classified as normal, abnormal or equivocal. Results: The SSS and SDS values were distinct and lower for images in prone position (p < 0.01). The analysis of the 17 segments of the left ventricle showed similar findings for most of the anterior (p < 0.01) and inferior (p<0.01) wall segments .One hundred forty-five studies were considered equivocal by the observers, but after the combined analysis with prone images, 70 (48.3%) were reclassified to normal (p < 0.01). Conclusions: Prone position imaging had impact on the interpretation of MPS images. The influence of soft tissue attenuation was reduced and studies initially classified as equivocal were reclassified to normal.