01-09-2016, 11:43 AM
1452058798-PaperCopy.docx (Size: 20.13 KB / Downloads: 7)
This paper [DespinaKontos, VasileiosMegalooikonomou, James Gee; 2005], illustrates the application of intelligent medical image analysis techniques in order toreduce the computational cost of statistical voxel-wise analysis for detecting discriminativeregions of morphological variability among different populations. It also demonstrates that novelstatistical image processing techniques that operate selectively on groups of pixels aresuitable for morphological analysis of anatomical structures visualized by modern medicalimaging modalities. This proposed methodology effectively decreases thenumber of statistical tests performed, alleviating the effect of the multiple comparison problem and proves that this approach detects regions of statistically significantmorphological variability.
This paper [Silvia D.Olabarriaga, Jeroen G. Snel, Charl P. Botha, Robert G. Belleman; 2007], mainly describes the properties of distributed systems to support and facilitate the development, evaluation and clinical application of Medical Image Analysis (MIA) methods. The phases in the method’s lifecycle (development, parameter optimisation, evaluation and clinical routine) are analysed. The requirements aredescribed, proposing a grid-oriented paradigm that emphasizesvirtual collaboration among users, pieces of software, and devicesdistributed among geographically dispersed healthcare, research and development enterprises. Finally, the characteristics of the existingsystems are analysed according to these requirements. Theproposed requirements offer a useful framework to evaluate, compare and improve the existing systems that support MIA development.
In this paper [M.A. Ansari, R.S. Anand; 2007], a region based segmentation and image analysis with application to medical images have done. Image segmentationwhich is one of the most important steps- includesclustering, object detection and boundarydetections. This paper also describes various edge detector techniques like- Robert’s detector, Laplace detector, Perwitt detector, Sobel detector. It also speaks about algorithm for thresholding- (global, local and adaptive) and algorithm for region merging techniques.
This paper [CorneliuFlorea, ConstantinVertan; 2007], propose a method that boosts the dynamic range, by combining a set of digital images of the same X-Ray, acquired under different exposure values. The fusion of the acquired images is performed according to weighting scheme derived from the confidence information extracted from an experimentally derived digital still camera response function. The resulting image can be obtained at various high dynamic range values and shows improved feature visibility. This method have been successfully applied to enhance the dynamic range of the hip prosthesis radiography images.
The 3 main fields on which medical image processing and analysing focuses are Stuctural Imaging, Molecular Imaging and Functional Imaging. This paper [JieTian, JianXue, Yakang Dai, Jian Chen and JianZheng;2008], introduces a novel software platform that is investigated and developed specially for medical image processing and analyses. This paper proposes a full platform solution for medical image processing and analysing, including the Medical Imaging Toolkit (MITK) and the 3-Dimensional Medical Image Processing and Analysing System (3DMed). This platform proves to provide an effective method for developing robust medical imaging software.
In this paper [Ili AyuniMohdIkhsan, AiniHussain, MohdAsyrafZulkifley, Nooritawati Md. Tahir, Aouache Mustapha;2014], various methods of pre-processing techniques for vertebral bone segmentation have been analysed. Three methods are considered which are Histogram Equalisation (HE), Gamma Correction (GC), and Contrast Limited Adaptive Histogram Equaliser (CLAHE). This paper aims to compare and quantify the precision and accuracy of the techniques that are used to enhance the image quality. Experimental results of the system yield favourable results where the most accurate technique is CLAHE, followed by GC and HE.
The main methods oral surgeons use to assess the progressin treatment of bone defects are classicalrentgenodiagnostic and histomorphometric analyses.Histomorphometry is of limited use in routine procedures due to its invasiveness and classical X-Ray may be insufficient because the human eye cannot see and quantify all the subtle brightness differences on radiographs. This paper [Lukasz Karolczak, AndrzejMaterka, MarcinKozakiewicz; 2014], describes a newly developed algorithm for quantitative computer analysis of intra-oral radiographs. The newmethod of X-ray numerical analysis for bone regenerationmonitoring, proposed in this paper, is developed to reduce theinfluence of those three effects (Image Background Inhomogeneity, Poor Contrast and Substantial Variation of Radiographs Acquisition Parameters) on the key quantitativefeatures that are the basis for the monitoring and results show that this algorithm outperforms the technique used in earlier studies.
This paper [Ching-Chun Huang, Hung-Nguyen Manh, Chen-Yu Tseng; 2014], had proposed a tissue attenuation method to enhance the bright regions. Ingeneral, the bright regions of an X-Rayimage are of interest, since most important matters compactly will be located in thoseregions. By estimating the amount of the tissuecomponent and locally adjusting the ratios of remaining tissues, the system can achieve contrast enhancement in anefficient manner especially for bright regions.To adjust the ratios, a two-step procedure was proposed.First, the tissue component was separated from a givenimage based on local contrast maximization. Second, anattenuation adjustment was performed to control the ratio ofremovable tissues in order to correctly enhance contrast. Experimental results have demonstrated that the algorithm could effectivelypresent X-ray details for inspection.
Medical Ultrasound Image Enhancement and De-noising has been an important subject of medical image processing.Medical Ultrasound images have unique speckle noise, but may also produce other noise- Gaussian noise and impulse noise in the process of storage and transmission. This paper [He Wen, Wu Qi; 2015], proposes a kind of ultrasonic image enhancement and de-noising algorithm based on wavelet analysis theory and fuzzy theory. Experiments show that the algorithm can make ultrasound images with good visual effects, and lay the foundation for accurate processing.
Quantitatively analyse the characteristics ofintraventricular flow is critical for the diagnosis of congenitalheart disease (CHD). Although several techniques are applied inclinic currently, the information which is provided is notsatisfied with the requirement of the therapies of CHD.This paper [Lijun Chen, Jinlong Liu, Haifa Hong, Aimin Sun, Jinfen Liu, Yuqi Zhang, MitsuoUmezu; 2015], introduces the method of computational fluid dynamics (CFD) to investigate the intraventricular hemodynamics using a reconstructed model of left ventricle (LV) based on cardiac magneticresonance (CMR) images.
In this paper [KanyanatMeejaroen, Charoen Chaweechan, WanusKhodsiri, VorapraneeKhu-smith, UkritWatchareeruetai, PattanaSornmagura, Taya Kittiyakara;2015], an image-processing-based method designed to detect fibrosis in liver biopsy images is proposed. The proposed method first enhances the colour difference between liver tissue and fibrosis areas. Then, a low-pass filtering is applied to each colour band to reduce noise. In order to calculate the percentage of fibrosis against total liver tissue, the background area, i.e. empty slide area, is detected. Next, Bayesian classifier is used to separate fibrosis from liver tissue based on the colour information. Finally, the proportion of the fibrosis area to the tissue area is computed.
In medical image analysis, segmentation of medical images such as Computed Tomography (CT) volumetric images is necessary for further medical image analysis and computer aided intervention. This paper [AsukaOkagawa, Yuji Oyamada, Yoshihiko Mochizuki, Hiroshi Ishikawa; 2015], proposes a method for medical image segmentation by higherorder energy minimization. Specifically, they introduce a higher-order term that describes the continuity around the edge points of a CT image. The parameters of the energy terms are determined according to various conditional probabilities learned from sample data with the ground truth. Then they minimize the energy using graph cuts and evaluate the effectiveness of the introduction of the term into the traditional energy.In the experimental results, the accuracy is improved in some cases and the problem of robustness should be solved in the future research.
Diabetic Retinopathy (DR) disease is the major cause of adult blindness in India, USA and China. According to medical experts, early detection of DR is essential to prevent the blindness. Furthermore, in case of low contrast noisy retinal images, the detection of MAs become complicated. Therefore prior to the medical analysis, Contrast Enhancement of retinal image is an essential step. In this paper [N. S. Datta, P. Saha, H. S. Dutta, D. Sarkar, S. Biswas, P. Sarkar;2015], a new Contrast Enhancement method is introduced for retinal images which produces better quality of image and also able to preserve the mean brightness satisfactorily and as a result it obviously improves the overall MAs (small round shape red spot on retina called Micro-aneurysms are the first sign of DR)detection method.
This paper [S. Makrogiannis, K. W. Fishbein, A. Z. Moore, R. G. Spencer, L. Ferrucci; 2015], presents a method for identification and characterization of muscle and adipose tissue in the mid-thigh region using MRI. It proposes an image-based muscle quality prediction technique that estimates tissue-specific probability density models and their Eigen structures in the joint domain of water- and fat-suppressed voxel signal intensities along with volumetric and intensity-based tissue characteristics computed during the quantification stage. The central hypothesis is that we can use the MRI-based muscle quality signatures to predict biomechanical properties of the mid-thigh, namely the muscle quality index MQ.
The analysis presented in this paper [Dick W. Harberts, Mark van Helvoort; 2015], shows that the requirements on diagnostic image quality pose higher limitations on the use of magnetic materials in medical implants than the requirements for magneto-mechanical forces. From experiments with ferrites in a 1.5-T MRI system, it is concluded that if diagnostic image quality is required for distances of at least 2 cm from the medical implant, then only 0.1 millimetre cube ferrite can be accepted. This amount of ferrite is too low for any practical application. Therefore, medical implants can only be coexistent with MRI if they do not contain ferrites.