Current Issue : July - September Volume : 2017 Issue Number : 3 Articles : 5 Articles
Background: Positron Emission Tomography ââ?¬â?? Computed Tomography (PET/CT) imaging is the basis for the\nevaluation of response-to-treatment of several oncological diseases. In practice, such evaluation is manually\nperformed by specialists, which is rather complex and time-consuming. Evaluation measures have been proposed,\nbut with questionable reliability. The usage of before and after-treatment image descriptors of the lesions for\ntreatment response evaluation is still a territory to be explored.\nMethods: In this project, Artificial Neural Network approaches were implemented to automatically assess treatment\nresponse of patients suffering from neuroendocrine tumors and Hodgkyn lymphoma, based on image features\nextracted from PET/CT.\nResults: The results show that the considered set of features allows for the achievement of very high classification\nperformances, especially when data is properly balanced.\nConclusions: After synthetic data generation and PCA-based dimensionality reduction to only two components,\nLVQNN assured classification accuracies of 100%, 100%, 96.3% and 100% regarding the 4 response-to-treatment\nclasses....
Background: Obstructive sleep apnea (OSA) is a public health problem. Detailed analysis of the para-pharyngeal fat\npads can help us to understand the pathogenesis of OSA and may mediate the intervention of this sleeping disorder.\nA reliable and automatic para-pharyngeal fat pads segmentation technique plays a vital role in investigating larger\ndata bases to identify the anatomic risk factors for the OSA.\nMethods: Our research aims to develop a context-based automatic segmentation algorithm to delineate the fat\npads from magnetic resonance images in a population-based study. Our segmentation pipeline involves texture\nanalysis, connected component analysis, object-based image analysis, and supervised classification using an\ninteractive visual analysis tool to segregate fat pads from other structures automatically.\nResults: We developed a fully automatic segmentation technique that does not need any user interaction to extract\nfat pads. Our algorithm is fast enough that we can apply it to population-based epidemiological studies that provide a\nlarge amount of data. We evaluated our approach qualitatively on thirty datasets and quantitatively against the\nground truths of ten datasets resulting in an average of approximately 78% detected volume fraction and a 79% Dice\ncoefficient, which is within the range of the inter-observer variation of manual segmentation results.\nConclusion: The suggested method produces sufficiently accurate results and has potential to be applied for the\nstudy of large data to understand the pathogenesis of the OSA syndrome....
CT-scan is the most irradiating tool in diagnostic radiology. For 5% - 10% of\ndiagnostic X-ray procedures, it is responsible for 34% of irradiation according\nto UNSCEAR. Patients radiation protection must therefore be increased during\nCT-scan procedures. This requires the rigorous application of optimization\nprinciple which imposes to have ââ?¬Å?diagnostic reference levelsââ?¬Â. Objective: The\naim of this study was to determine the diagnostic reference levels (DRLs) of the\nfour most frequent CT-scans examinations of adults in Cameroon. Material\nand Method: It was a cross-sectional pilot study carried out from April to\nSeptember 2015 in five health facilities using CT-scan in Cameroon. The\nstudied variables were: patients age and sex, type of CT-scan examination\n(cerebral, chest, abdomino-pelvic, lumbar spine), Used of IV contrast (IVâË?â??/\nIV+), acquisition length, time of tube rotation, voltage (kV), mAs, pitch,\nthickness of slices, CTDIvol and DLP. For each type of examination, at least 30\npatients were included per center, consecutively on the randomly predetermined\ndays. The DRL for each type of examination was defined as the 75th percentile\nof its PDL and CTDIvol. Results: Of the 696 examinations, 41.2% were\ncerebral, 26.9% abdomino-pelvic, 17.7% lumbar spine and 14.2% chest. The\nmean age of patients was 52 Ã?± 15 years [20 - 90 years], 58.9% were 50 years and\nolder. The sex-ratio was 1.26 (55.9% males). The CT machines were 4, 8 and 16\nmultidetectors. The 75th percentile of DLP or DRLs [standard deviation] was:2\n[1150 Ã?± 278 mGyâË?â?¢cm], [770 Ã?± 477 mGyâË?â?¢cm], [720 Ã?± 170 mGyâË?â?¢cm] and [715 Ã?±\n187 mGyâË?â?¢cm] respectively for cerebral, lumbar spine, abdominopelvic and\nchest CT-scans. Taking in consideration the number of detectors, the 75th\npercentile of the Dose-Length product decreased with the increase number of\ndetectors for cerebral examinations but was the highest with 16 MDCT for the\nabdominopelvic, lumbar spine and chest CT-scans. For the chest and lumbar\nspine examinations, there was a significant increase in patient-dose with the\nincrease in the number of detectors. Conclusion: Our DRLs values lie between\nthe norms of some European countries and those of some African\ncountries. There is remarquable variation in dose for the commonest CTscans\nexaminations in Cameroon, requiring then an optimization process\nfrom these determined DRLs and establishment of national DRLs. Special\nattention to optimization should be paid when using 16 MDCT....
Optical colonoscopy is the most common approach to diagnosing bowel diseases through direct colon and rectum inspections.\nPeriodic optical colonoscopy examinations are particularly important for detecting cancers at early stages while still treatable.\nHowever, diagnostic accuracy is highly dependent on both the experience and knowledge of the medical doctor. Moreover, it\nis extremely difficult, even for specialist doctors, to detect the early stages of cancer when obscured by inflammations of the\ncolonic mucosa due to intractable inflammatory bowel diseases, such as ulcerative colitis. Thus, to assist the UC diagnosis, it is\nnecessary to develop a new technology that can retrieve similar cases of diagnostic target image from cases in the past that stored\nthe diagnosed images with various symptoms of colonic mucosa. In order to assist diagnoses with optical colonoscopy, this paper\nproposes a retrieval method for colonoscopy images that can cope with multiscale objects. The proposed method can retrieve\nsimilar colonoscopy images despite varying visible sizes of the target objects. Through three experiments conducted with real\nclinical colonoscopy images, we demonstrate that the method is able to retrieve objects of any visible size and any location at a\nhigh level of accuracy....
Ventricle segmentation is a challenging technique for the development of detection system of ischemic stroke in computed\ntomography (CT), as ischemic stroke regions are adjacent to the brain ventricle with similar intensity. To address this problem, we\ndeveloped an objective segmentation system of brain ventricle in CT.The intensity distribution of the ventricle was estimated based\non clustering technique, connectivity, and domain knowledge, and the initial ventricle segmentation results were then obtained. To\nexclude the stroke regions from initial segmentation, a combined segmentation strategy was proposed, which is composed of three\ndifferent schemes: (1) the largest three-dimensional (3D) connected component was considered as the ventricular region; (2) the\nbig stroke areas were removed by the image difference methods based on searching optimal threshold values; (3) the small stroke\nregions were excluded by the adaptive template algorithm.The proposed method was evaluated on 50 cases of patients with ischemic\nstroke.The mean Dice, sensitivity, specificity, and root mean squared error were 0.9447, 0.969, 0.998, and 0.219 mm, respectively.\nThis system can offer a desirable performance. Therefore, the proposed system is expected to bring insights into clinic research and\nthe development of detection system of ischemic stroke in CT....
Loading....