Article

Impact of advanced parallel or cloud computing technologies for image guided diagnosis and therapy

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... attention of many researchers in the field [22]. Some of the potential reasons to use cloud computing for medical image analysis are huge data storage, remote availability, faster processing demands, and scalability. ...
... It allowed users to upload MRI or CT scan images for detecting the tumor in real-time. Parallel computing approach to accelerate microscopy data High throughput Does not considered big data frameworks [33] A special issue on the use of cloud and parallel computing have been published in the Journal of X-Ray Science and Technology to highlight the state-of-the-art advances in image-guided diagnosis and therapy [22]. For instance, the use of parallel computing is investigated for modeling blood flow using computational fluid dynamics and smooth particle hydrodynamics in Reference [34]. ...
Article
Full-text available
For processing large-scale medical imaging data, adopting high-performance computing and cloud-based resources are getting attention rapidly. Due to its low–cost and non-invasive nature, microwave technology is being investigated for breast and brain imaging. Microwave imaging via space-time algorithm and its extended versions are commonly used, as it provides high-quality images. However, due to intensive computation and sequential execution, these algorithms are not capable of producing images in an acceptable time. In this paper, a parallel microwave image reconstruction algorithm based on Apache Spark on high-performance computing and Google Cloud Platform is proposed. The input data is first converted to a resilient distributed data set and then distributed to multiple nodes on a cluster. The subset of pixel data is calculated in parallel on these nodes, and the results are retrieved to a master node for image reconstruction. Using Apache Spark, the performance of the parallel microwave image reconstruction algorithm is evaluated on high-performance computing and Google Cloud Platform, which shows an average speed increase of 28.56 times on four homogeneous computing nodes. Experimental results revealed that the proposed parallel microwave image reconstruction algorithm fully inherits the parallelism, resulting in fast reconstruction of images from radio frequency sensor’s data. This paper also illustrates that the proposed algorithm is generalized and can be deployed on any master-slave architecture.
Conference Paper
Full-text available
Cloud computing and machine learning (ML) has recently gained prominence in the healthcare industry.Machine learning and cloud computing play a dynamic role in illness detection, but mainly among those living in rural locations with few medical resources. Machine-learning-based diagnosis systems function as secondary readers, assisting radiologists in accurately identifying illnesses, while cloud-based systems can enable telehealth and remote diagnostics.Artificial neural networks techniques have piqued many academics who want to investigate their potential for illness detection. An extreme learning machine (ELM) is a type of artificial neural network (ANN) with a lot of potential for handling classification issues. Cloud computing platforms provide a more precise and real-time forecast of the disease's growing tendency. Patients with life-threatening illnesses such as cancer, diabetes, neurological disorders, coronary heart disease, and HIV/AIDS are more susceptible to severe complications. This necessitates creating a solid mathematical foundation for tracking its spread and the automation of monitoring tools for dynamic online decision-making.Innovative solutions are needed to create, manage, and analyse massive data on the growing network of infected patients, patient information, and community movements and combine clinical trials, pharmaceutical, genetic, and public health data. Hence, ML can be used to manage enormous amounts of data and intelligently forecast disease spread. Cloud computing is utilised to improve the prediction process quickly by using high-speed computations.Cloud computing is gaining traction to solve the problem of delivering sophisticated services and data via the Internet.In addition, the current paper is conducted to understand the impact of cloud computing and machine learning Techniques in clinical diagnosis. The research design has been used in the present study is descriptive. Secondary data is analysed for literature review and purpose of analysis to get a precise conclusion.
Article
Medical images may be corrupted by noise. This noise affects the image quality and can obscure important information required for accurate diagnosis. Effectively apply filtering techniques can facilitate diagnosis or reduce radiation exposure. In this paper, we introduce a parallel method designed to reduce mixed Gaussian-impulse noise from digital images. The method uses fuzzy logic and the fuzzy peer group concept. Implementations of the method on multi-core interface using the open multi-processing (OpenMP) and on graphics processing units (GPUs) using CUDA are presented. Efficiency is measured in terms of execution time and in terms of MAE, PSNR and SSIM over medical images from the mini-MIAS database and over computed radiography (CR) images generated at different exposure levels. These images have been contaminated with impulsive and/or Gaussian noise. Experiments show that the proposed method obtains good performance in terms of the above mentioned objective quality measures. After applying multi-core and GPUs optimization strategies, the observed time shows that the new filter allows to remove mixed Gaussian-impulse noise in real-time.
Article
Full-text available
Real-time monitoring and amendment of patient position is important for the radiotherapy. However, using electronic portal imaging device (EPID) and cone beam computer tomography (CBCT) in the clinical practice generate different degrees of delay, so that they cannot achieve the purpose of real-time application. Meanwhile, a few products come with the function of the real-time monitoring and amendment, such as CyberKnife, which is too expensive for the common people. The objective of this study is to develop and test a novel independent system to monitor treatment center and amend the position of patient, which is applicable to most accelerators, based on binocular location. The system monitors the treatment center by tracking the markers attached to the patient. Once the treatment center shifts, the system uses the magic finger, which is developed to control the treatment bet automatically to adjust the treatment bed position. To improve the monitoring accuracy, we trained the data collected from the clinic based on SVM (Support Vector Machine). Thus, the training results assist users to adjust the feasible degree of the monitoring. The experiment results showed that using this new monitoring system, the monitoring resolution reached 0.5 mm, and the error ratio of the judgment was less than 1.5%.
Article
Full-text available
Brain tumor segmentation aims to separate the different tumor tissues such as active cells, necrotic core, and edema from normal brain tissues of White Matter (WM), Gray Matter (GM), and Cerebrospinal Fluid (CSF). MRI-based brain tumor segmentation studies are attracting more and more attention in recent years due to non-invasive imaging and good soft tissue contrast of Magnetic Resonance Imaging (MRI) images. With the development of almost two decades, the innovative approaches applying computer-aided techniques for segmenting brain tumor are becoming more and more mature and coming closer to routine clinical applications. The purpose of this paper is to provide a comprehensive overview for MRI-based brain tumor segmentation methods. Firstly, a brief introduction to brain tumors and imaging modalities of brain tumors is given. Then, the preprocessing operations and the state of the art methods of MRI-based brain tumor segmentation are introduced. Moreover, the evaluation and validation of the results of MRI-based brain tumor segmentation are discussed. Finally, an objective assessment is presented and future developments and trends are addressed for MRI-based brain tumor segmentation methods.
Article
Full-text available
To evaluate the value of apparent diffusion coefficient (ADC) histogram analysis for predicting tumor recurrence in patients with uterine cervical cancer treated with chemoradiation therapy (CRT). Our institutional review board approved this retrospective study and waived informed consent from each patient. Forty-two patients (mean age, 56 ± 14 years) with biopsy-proven uterine cervical squamous cell carcinoma who underwent both pre-treatment pelvic magnetic resonance imaging with a 3.0 T magnetic resonance scanner and concurrent CRT were included. All patients were followed-up for more than 6 months (mean, 36.4 ± 11.9 months; range 9.0-52.8 months) after completion of CRT. Baseline ADC parameters (mean ADC, 25th percentile, 50th percentile, and 75th percentile ADC values) of tumors were calculated and compared between the recurrence and no recurrence groups. In the recurrence group, the mean ADC and 75th percentile ADC values of tumors were significantly higher than those of the no recurrence group (p = 0.043 and p = 0.008, respectively). In multivariate analysis, the 75th percentile ADC value of tumors was a significant predictor for tumor recurrence (p = 0.009; hazard ratio, 1.319). When the cut-off value of the 75th percentile ADC (0.936 × 10(-3) mm(2)/sec) was used, the overall recurrence free survival rate above the cut-off value was significantly lower than that below the cut-off value (51.9% vs. 91.7%, p = 0.003, log-rank test). Pre-CRT ADC histogram analysis may serve as a biomarker for predicting tumor recurrence in patients with uterine cervical cancer treated with CRT.
Article
Full-text available
This paper presents a novel model-based patient-specific method for automatic detection of seizures in the intracranial EEG recordings. The proposed method overcomes the complexities in the practical implementation of the patient-specific approach of seizure detection. The method builds a seizure model (set of basis functions) for a priori known seizure (the template seizure pattern), and uses the statistically optimal null filters as a building block for the detection of similar seizures. The process of modeling the template seizure is fully automatic. Overall, the detection method involves the segmentation of the template seizure pattern, rejection of the redundant and noisy segments, extraction of features from the segments to generate a set of models, selection of the best seizure model, and training of the classifier. The trained classifier is used to detect similar seizures in the remaining data. The resulting seizure detection method was evaluated on a total of 304 h of single-channel depth EEG recordings from 14 patients. The system performance is further compared to the Qu-Gotman patient-specific system using the same data. A significant improvement in the proposed system, in terms of specificity, is observed over the compared method.
Article
Full-text available
Study aim To explore various sources of variability in the measurement of LV volumes and dyssynchrony by 3D echocardiography (3DE). Methods We studied 100 patients (58 ± 18 years, 51 men) to assess the impact of: (1) manual editing; (2) 3D data set temporal resolution; (3) LV 16- or 17-segmentation model; (4) software sensitivity for automated endocardial surface detection; and (5) image quality, on the measurement of LV end-diastolic (EDV) and end-systolic (ESV) volumes, sphericity indices (EDSI, ESSI), ejection fraction (EF) and dyssynchrony (SDI). Two- and 4-beat LV full-volume data sets were analyzed and compared. Cardiac magnetic resonance (CMR) was used as reference in 26 patients. Results Manual editing of endocardial surface improved the agreement of LV volumes with CMR, but increased SDI (SDI17: 5.6 ± 0.5% vs. 4.3 ± 0.3%; P < 0.0001). Data set temporal resolution had no significant impact on LV parameters. Adding the 17th to 16-segment LV model did not significantly increase SDI. Reducing software sensitivity in endocardial surface detection increased EDV (101 ± 46 ml vs. 118 ± 50 ml) and sphericity, decreased SDI (SDI 17: 6.7 ± 3.3% vs. 2.9 ± 3.7%) (P < 0.05 for all), and improved agreement of EDV and ESV with CMR. Impact of software sensitivity in LV endocardium detection on LV parameters was related to image quality: higher on SDI in pts with suboptimal quality (SDI 17 bias 4.5% vs. 3.2%, P < 0.05); higher on LV volumes in patients with optimal quality (EDV bias 14 ml vs. 19 ml, ESV bias 5 ml vs. 9 ml; P = 0.01). Conclusions Manual editing, software settings and image quality significantly impact on 3D LV volumes and dyssynchrony assessment.
Article
Full-text available
Significant progress has been made towards the clinical application of human-machine interfaces (HMIs) based on electromyographic (EMG) pattern recognition for various rehabilitation purposes. Making this technology practical and available to patients with motor deficits requires overcoming real-world challenges, such as physical and physiological changes, that result in variations in EMG signals and systems that are unreliable for long-term use. In this study, we aimed to address these challenges by (1) investigating the stability of time-domain EMG features during changes in the EMG signals and (2) identifying the feature sets that would provide the most robust EMG pattern recognition. Variations in EMG signals were introduced during physical experiments. We identified three disturbances that commonly affect EMG signals: EMG electrode location shift, variation in muscle contraction effort, and muscle fatigue. The impact of these disturbances on individual features and combined feature sets was quantified by changes in classification performance. The robustness of feature sets was evaluated by a stability index developed in this study. Muscle fatigue had the smallest effect on the studied EMG features, while electrode location shift and varying effort level significantly reduced the classification accuracy for most of the features. Under these disturbances, the most stable EMG feature set with combination of four features produced at least 16.0% higher classification accuracy than the least stable set. EMG autoregression coefficients and cepstrum coefficients showed the most robust classification performance of all studied time-domain features. Selecting appropriate EMG feature combinations can overcome the impact of the studied disturbances on EMG pattern classification to a certain extent; however, this simple solution is still inadequate. Stabilizing electrode contact locations and developing effective classifier training strategies are suggested to further improve the robustness of HMIs based on EMG pattern recognition.
Article
Full-text available
A training algorithm that maximizes the margin between the training patterns and the decision boundary is presented. The technique is applicable to a wide variety of classifiaction functions, including Perceptrons, polynomials, and Radial Basis Functions. The effective number of parameters is adjusted automatically to match the complexity of the problem. The solution is expressed as a linear combination of supporting patterns. These are the subset of training patterns that are closest to the decision boundary. Bounds on the generalization performance based on the leave-one-out method and the VC-dimension are given. Experimental results on optical character recognition problems demonstrate the good generalization obtained when compared with other learning algorithms. 1
Article
Objective: To retrospectively reappraise characteristics of the electro-clinical seizure semiology of the bilateral asymmetric tonic seizure (BATS) in the patients with supplementary sensorimotor area (SSMA) epilepsy. Methods: From the retrospective analysis of the pre- and post-operative Magnetic Resonance Imaging (MRI) data involving 386 patients who received epilepsy surgery, 123 BATS were identified meeting the clinical criteria and included in the study. For comparison in four extremities involvement, limbs were paired and comparatively evaluated between the contralateral and ipsilateral sides, proximal and distal segments, and upper and lower limbs. For evaluation of sequential events, each tonic phase of the BATS was chronologically divided into 10 equal epochs. In each epoch, distribution of tonic events in 4 extremities and axes was visually evaluated and comparatively analyzed. Results: Asymmetric tonic posturing was the most constant findings in 6 patients, whose upper limbs contralateral to epileptogenic cortex were kept in abduction in all 123 (100%) seizures and extension in 118 (95.9%) seizures. This type of asymmetry became visible and remained stable in the initial three epochs of the tonic phase in 107 out of 123 (87.0%) seizures. In each epoch, especially the initial one, the contralateral upper limbs were involved more frequently than those ipsilateral to the epileptogenic cortex (p < 0.05). By pairwise comparison, an earlier involvement of the contralateral side to epileptogenic cortex was visually observed in 53 out of 280 (18.9%) limb pairs, in which the ipsilateral limbs were preceded by the contralateral ones 4.6 (0.1-16.0) seconds earlier. Both of the proximal and distal segments were simultaneously involved in 260 out of 298 (87.2%) limb pairs, although the former were 4.3 (0.5-16.0) earlier than the latter in 34 out of 298 (11.4%) limb pairs. Conclusions: This study demonstrated that by studying the restricted epileptogenic lesion limited to pure SSMA, unilateral extension and abduction posturing in upper limb were the most prominent and valuable sign for the lesion lateralization in SSMA neurosurgery decision-making.
Article
Background: Epilepsy is a chronic disease with transient brain dysfunction that results from the sudden abnormal discharge of neurons in the brain. Since electroencephalogram (EEG) is a harmless and noninvasive detection method, it plays an important role in the detection of neurological diseases. However, the process of analyzing EEG to detect neurological diseases is often difficult because the brain electrical signals are random, non-stationary and nonlinear. Objective: In order to overcome such difficulty, this study aims to develop a new computer-aided scheme for automatic epileptic seizure detection in EEGs based on multi-fractal detrended fluctuation analysis (MF-DFA) and support vector machine (SVM). Methods: New scheme first extracts features from EEG by MF-DFA during the first stage. Then, the scheme applies a genetic algorithm (GA) to calculate parameters used in SVM and classify the training data according to the selected features using SVM. Finally, the trained SVM classifier is exploited to detect neurological diseases. The algorithm utilizes MLlib from library of SPARK and runs on cloud platform. Results: Applying to a public dataset for experiment, the study results show that the new feature extraction method and scheme can detect signals with less features and the accuracy of the classification reached up to 99%. Conclusions: MF-DFA is a promising approach to extract features for analyzing EEG, because of its simple algorithm procedure and less parameters. The features obtained by MF-DFA can represent samples as well as traditional wavelet transform and Lyapunov exponents. GA can always find useful parameters for SVM with enough execution time. The results illustrate that the classification model can achieve comparable accuracy, which means that it is effective in epileptic seizure detection.
Article
Background: Surface electromyography (sEMG) signal is the combined effect of superficial muscle EMG and neural electrical activity. In recent years, researchers did large amount of human-machine system studies by using the physiological signals as control signals. Objective: To develop and test a new multi-classification method to improve performance of analyzing sEMG signals based on public sEMG dataset. Methods: First, ten features were selected as candidate features. Second, a genetic algorithm (GA) was applied to select representative features from the initial ten candidates. Third, a multi-layer perceptron (MLP) classifier was trained by the selected optimal features. Last, the trained classifier was used to predict the classes of sEMG signals. A special graphics processing unit (GPU) was used to speed up the learning process. Results: Experimental results show that the classification accuracy of the new method reached higher than 90%. Comparing to other previously reported results, using the new method yielded higher performance. Conclusions: The proposed features selection method is effective and the classification result is accurate. In addition, our method could have practical application value in medical prosthetics and the potential to improve robustness of myoelectric pattern recognition.
Article
Background: The computer mouse is an important human-computer interaction device. But patients with physical finger disability are unable to operate this device. Surface EMG (sEMG) can be monitored by electrodes on the skin surface and is a reflection of the neuromuscular activities. Therefore, we can control limbs auxiliary equipment by utilizing sEMG classification in order to help the physically disabled patients to operate the mouse. Objective: To develop a new a method to extract sEMG generated by finger motion and apply novel features to classify sEMG. Methods: A window-based data acquisition method was presented to extract signal samples from sEMG electordes. Afterwards, a two-dimensional matrix image based feature extraction method, which differs from the classical methods based on time domain or frequency domain, was employed to transform signal samples to feature maps used for classification. In the experiments, sEMG data samples produced by the index and middle fingers at the click of a mouse button were separately acquired. Then, characteristics of the samples were analyzed to generate a feature map for each sample. Finally, the machine learning classification algorithms (SVM, KNN, RBF-NN) were employed to classify these feature maps on a GPU. Results: The study demonstrated that all classifiers can identify and classify sEMG samples effectively. In particular, the accuracy of the SVM classifier reached up to 100%. Conclusions: The signal separation method is a convenient, efficient and quick method, which can effectively extract the sEMG samples produced by fingers. In addition, unlike the classical methods, the new method enables to extract features by enlarging sample signals' energy appropriately. The classical machine learning classifiers all performed well by using these features.
Article
Background: Brain tumor segmentation is a challenging task for its variation in intensity. The phenomenon is caused by the inhomogeneous content of tumor tissue and the choice of imaging modality. In 2010 Zhang developed the Selective Binary Gaussian Filtering Regularizing Level Set (SBGFRLS) model that combined the merits of edge-based and region-based segmentation. Objective: To improve the SBGFRLS method by modifying the singed pressure force (SPF) term with multiple image information and demonstrate effectiveness of proposed method on clinical images. Methods: In original SBGFRLS model, the contour evolution direction mainly depends on the SPF. By introducing a directional term in SPF, the metric could control the evolution direction. The SPF is altered by statistic values enclosed by the contour. This concept can be extended to jointly incorporate multiple image information. The new SPF term is expected to bring a solution for blur edge problem in brain tumor segmentation. The proposed method is validated with clinical images including pre- and post-contrast magnetic resonance images. The accuracy and robustness is compared with sensitivity, specificity, DICE similarity coefficient and Jaccard similarity index. Results: Experimental results show improvement, in particular the increase of sensitivity at the same specificity, in segmenting all types of tumors except for the diffused tumor. Conclusion: The novel brain tumor segmentation method is clinical-oriented with fast, robust and accurate implementation and a minimal user interaction. The method effectively segmented homogeneously enhanced, non-enhanced, heterogeneously-enhanced, and ring-enhanced tumor under MR imaging. Though the method is limited by identifying edema and diffuse tumor, several possible solutions are suggested to turn the curve evolution into a fully functional clinical diagnosis tool.
Article
Objective: To compare a full-automated software to quantify 3D transthoracic echocardiography namely, 3DE-HM (three-dimensional echocardiography HeartModel, Philips Healthcare) with the traditional manual quantitative method (3DE-manual) for assessing volumes of left atrial and ventricular volumes, and left ventricular ejection fraction (LVEF). Methods: 3D full volume images acquired from 156 subjects were collected and divided into 3 groups, which include 70 normal control cases (Group A), 17 patients with left ventricular remodeling after acute myocardial infarction (AMI) (Group B), and 69 patients with left atrial remodeling secondary to hypertension (Group C). The 3DE-HM method was used to quantify left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), left atrial end-systolic volume (LAESV), and left ventricular ejection fraction (LVEF), respectively. The results were compared with those obtained with the 3DE-manual method for correlation and consistency analyses. The reproducibility of the 3DE-HM method was also evaluated. Results: There was a high correlation between LVEDV, LVESV, LAESV and LVEF values obtained with the 3DE-HM method and those obtained using the 3DE-manual method (r = 0.72 to 0.97). The correlation was strongest for Group B, patients with left ventricular remodeling post-AMI also demonstrated the greatest degree of morphologic changes. There was a significant difference in all parameters measured with the 3DE-HM method in different groups (P < 0.05). The difference in the measurements of LVEDV and LVESV between the two methods was greatest in patients in Group B compared with patients with hypertension-induced left ventricular remodeling (Group C) and in normal controls (Group A) (P < 0.05). Lastly, the difference in the measurement of LAESV between the two methods was greater in patients with hypertension-induced left ventricular remodeling (Group C) than that in the control group (Group A) (P < 0.05). The post-processing time of the 3DE-HM data was significantly shorter than that using the 3DE-manual method (P < 0.05). There was no significant variability in repeated measurements at different time points using the 3DE-HM method either between subjects in different groups or within the same subject. Conclusion: 3DE-HM is a quick and feasible method for left ventricular quantification and is clinically applicable for evaluating patients with left atrial and left ventricular remodeling.
Article
Purpose: Thoracic aortic dissection (TAD) is considered one of the most catastrophic and non-traumatic cardiovascular diseases associated with high morbidity and mortality rates in clinical treatment. The purpose of this paper is to investigate the pulsatile hemodynamics changes throughout a cardiac cycle in a Stanford Type B TAD model with the aid of computational fluid dynamics (CFD) method. Methods: A patient-specific dissected aorta geometry was reconstructed from the three-dimensional (3D) computed tomography angiography (CTA) scanning. The realistic time-dependent pulsatile boundary conditions were prescribed for our 3D patient-specific TAD model. Blood was considered to be an incompressible, Newtonian fluid. The aortic wall was assumed to be rigid, and a no-slip boundary condition was applied at the wall. CFD simulations were processed using the finite volume (FV) method to investigate the pulsatile hemodynamics in terms of blood flow velocity, aortic wall pressure, wall shear stress and flow vorticity. In the experiments, blood velocity, pressure, wall shear stress and vorticity distributions were analyzed qualitatively and quantitatively. Results: The experimental results demonstrated a high wall shear stress and strong vertical flow at dissection initiation. The results also indicated that wall shear progressed along the false lumen, which is a possible cause of blood flow between aortic wall layers.
Article
Simulation of blood flow in a stenosed artery using Smoothed Particle Hydrodynamics (SPH) is a new research field, which is a particle-based method and different from the traditional continuum modelling technique such as Computational Fluid Dynamics (CFD). Both techniques harness parallel computing to process hemodynamics of cardiovascular structures. The objective of this study is to develop and test a new robust method for comparison of arterial flow velocity contours by SPH with the well-established CFD technique, and the implementation of SPH in computed tomography (CT) reconstructed arteries. The new method was developed based on three-dimensional (3D) straight and curved arterial models of millimeter range with a 25% stenosis in the middle section. In this study, we employed 1,000 to 13,000 particles to study how the number of particles influences SPH versus CFD deviation for blood-flow velocity distribution. Because further increasing the particle density has a diminishing effect on this deviation, we have determined a critical particle density of 1.45 particles/mm2 based on Reynolds number (Re = 200) at the inlet for an arterial flow simulation. Using this critical value of particle density can avoid unnecessarily big computational expenses that have no further effect on simulation accuracy. We have particularly shown that the SPH method has a big potential to be used in the virtual surgery system, such as to simulate the interaction between blood flow and the CT reconstructed vessels, especially those with stenosis or plaque when encountering vasculopathy, and for employing the simulation results output in clinical surgical procedures.
Article
Background and objective: Effective treatment of Uterus Cervical Carcinoma (UCC) rely heavily on the precise pre-surgical staging. The conventional International Federation of Gynecology and Obstetrics (FIGO) system based on clinical examination is being applied worldwide for UCC staging. Yet its performance just appears passable. Thus, this study aims to investigate the value of applying Magnetic Resonance Imaging (MRI) with clinical examination in staging of UCC. Materials and methods: A retrospective dataset involving 164 patients diagnosed with UCC was enrolled in this study. The mean age of this study population was 46.1 years (range, 28-#x2013;75 years). All patients underwent operations and UCC types were confirmed by pathological examinations. The tumor stages were determined by two experienced Gynecologist independently based on FIGO examinations and MRI. The diagnostic results were also compared with the post-operative pathologic reports. Statistical data analysis on diagnostic performance was then done and reported. Results: The study results showed that the overall accuracy of applying MRI in UCC staging was 82.32%, while using FIGO staging method, the staging accuracy was 59.15%. Conclusions: MRI is suitable to evaluate tumor extent with high accuracy, and it can offer more objective information for the diagnosis and staging of UCC. Compared with clinical examinations based on FIGO, MRI illustrated relatively high accuracy in evaluating UCC staging, and is worthwhile to be recommended in future clinical practice.
Article
The goal of this study was to test the feasibility and accuracy of an automated algorithm that simultaneously quantifies 3-dimensional (3D) transthoracic echocardiography (TTE)-derived left atrial (LA) and left ventricular (LV) volumes and left ventricular ejection fraction (LVEF). Conventional manual 3D TTE tracings and cardiac magnetic resonance (CMR) images were used as a reference for comparison. Cardiac chamber quantification from 3D TTE is superior to 2D TTE measurements. However, integration of 3D quantification into clinical practice has been limited by time-consuming workflow and the need for 3D expertise. A novel automated software was developed that provides LV and LA volumetric quantification from 3D TTE datasets that reflect real-life manual 3-dimensional echocardiography measurements and values comparable to CMR. A total of 159 patients were studied in 2 separate protocols. In protocol 1, 94 patients underwent 3D TTE imaging (EPIQ, iE33, X5-1, Philips Healthcare, Andover, Massachusetts) covering the left atrium and left ventricle. LA and LV volumes and LVEF were obtained using the automated software (HeartModel, Philips Healthcare) with and without contour correction, and compared with the averaged manual 3D volumetric measurements from 3 readers. In protocol 2, automated measurements from 65 patients were compared with a CMR reference. The Pearson correlation coefficient, Bland-Altman analysis, and paired Student _t_ tests were used to assess inter-technique agreement. Correlations between the automated and manual 3D TTE measurements were strong (r = 0.87 to 0.96). LVEF was underestimated and automated LV end-diastolic, LV end-systolic, and LA volumes were overestimated compared with manual measurements. Agreement between the automated analysis and CMR was also strong (r = 0.84 to 0.95). Test–retest variability was low. Automated simultaneous quantification of LA and LV volumes and LVEF is feasible and requires minimal 3D software analysis training. The automated measurements are not only comparable to manual measurements but also to CMR. This technique is highly reproducible and timesaving, and it therefore promises to facilitate the integration of 3D TTE-based left-heart chamber quantification into clinical practice.
Article
Objective: To perform whole-brain morphometry in patients with frontal lobe epilepsy and evaluate the utility of group-level patterns for individualized diagnosis and prognosis. Methods: We compared MRI-based cortical thickness and folding complexity between 2 frontal lobe epilepsy cohorts with histologically verified focal cortical dysplasia (FCD) (13 type I; 28 type II) and 41 closely matched controls. Pattern learning algorithms evaluated the utility of group-level findings to predict histologic FCD subtype, the side of the seizure focus, and postsurgical seizure outcome in single individuals. Results: Relative to controls, FCD type I displayed multilobar cortical thinning that was most marked in ipsilateral frontal cortices. Conversely, type II showed thickening in temporal and postcentral cortices. Cortical folding also diverged, with increased complexity in prefrontal cortices in type I and decreases in type II. Group-level findings successfully guided automated FCD subtype classification (type I: 100%; type II: 96%), seizure focus lateralization (type I: 92%; type II: 86%), and outcome prediction (type I: 92%; type II: 82%). Conclusion: FCD subtypes relate to diverse whole-brain structural phenotypes. While cortical thickening in type II may indicate delayed pruning, a thin cortex in type I likely results from combined effects of seizure excitotoxicity and the primary malformation. Group-level patterns have a high translational value in guiding individualized diagnostics.
Article
The development of targeted anti-cancer therapies through the study of cancer genomes is intended to increase survival rates and decrease treatment-related toxicity. We treated a transposon-driven, functional genomic mouse model of medulloblastoma with 'humanized' in vivo therapy (microneurosurgical tumour resection followed by multi-fractionated, image-guided radiotherapy). Genetic events in recurrent murine medulloblastoma exhibit a very poor overlap with those in matched murine diagnostic samples (<5%). Whole-genome sequencing of 33 pairs of human diagnostic and post-therapy medulloblastomas demonstrated substantial genetic divergence of the dominant clone after therapy (<12% diagnostic events were retained at recurrence). In both mice and humans, the dominant clone at recurrence arose through clonal selection of a pre-existing minor clone present at diagnosis. Targeted therapy is unlikely to be effective in the absence of the target, therefore our results offer a simple, proximal, and remediable explanation for the failure of prior clinical trials of targeted therapy.
Article
Medically refractory epilepsy is associated with significant morbidity and mortality. Surgery is a safe and effective option for some patients, however the opportunity exists to develop less invasive and more effective surgical options. To this end, multiple minimally invasive, image-guided techniques have been applied to the treatment of epilepsy. These techniques can be divided into thermoablative and disconnective techniques. Each has been described in the treatment of epilepsy only in small case series. Larger series and longer follow up periods will determine each option's place in the surgical armamentarium for the treatment of refractory epilepsy but early results are promising.
Article
In this paper, a comprehensive method using symmetric normal inverse Gaussian (NIG) parameters of the sub-bands of EEG signals calculated in the dual-tree complex wavelet transformation domain is proposed for classifying EEG data. The suitability of the NIG probability distribution function is illustrated using statistical measures. A support vector machine is employed as the classifier of the EEG signals, wherein the NIG parameters are used as features. The performance of the proposed method is studied using a publicly available benchmark EEG database for various classification cases that include healthy, inter-ictal (seizure-free interval) and ictal (seizure), non-seizure and seizure, healthy and seizure, and inter-ictal and ictal, and compared with that of several recent methods. It is shown that in almost all the cases, the proposed method can provide 100 % accuracy with 100 % sensitivity and 100 % specificity while being faster as compared to the time–frequency analysis-based and EMD techniques.
Article
In this paper, a comprehensive method using symmetric normal inverse Gaussian (NIG) parameters of the sub-bands of EEG signals calculated in the dual-tree complex wavelet transformation domain is proposed for classifying EEG data. The suitability of the NIG probability distribution function is illustrated using statistical measures. A support vector machine is employed as the classifier of the EEG signals, wherein the NIG parameters are used as features. The performance of the proposed method is studied using a publicly available benchmark EEG database for various classification cases that include healthy, inter-ictal (seizure-free interval) and ictal (seizure), non-seizure and seizure, healthy and seizure, and inter-ictal and ictal, and compared with that of several recent methods. It is shown that in almost all the cases, the proposed method can provide 100 % accuracy with 100 % sensitivity and 100 % specificity while being faster as compared to the time–frequency analysis-based and EMD techniques.
Article
Aim: To evaluate the impact of staging FDG PET-CT on the initial management of patients with locally advanced cervical carcinoma (LACC) and any prognostic variables predicting survival. Materials and methods: Retrospective analysis of consecutive patients undergoing FDG PET-CT for staging of LACC in a single tertiary referral centre, between April 2008 and August 2011. Comparison was made between MRI and PET-CT findings and any subsequent impact on treatment intent or radiotherapy planning was evaluated. Results: Sixty-three patients underwent FDG PET-CT for initial staging of LACC. Major impact on management was found in 20 patients (32%), a minor impact in five (8%), and no impact in 38 (60%). In those patients where PET-CT had a major impact, 12 had more extensive local nodal involvement, five had occult metastatic disease, two had synchronous tumours, and one patient had equivocal lymph nodes on MRI characterized as negative. PET-positive nodal status at diagnosis was found to be a statistically significant predictor of relapse-free survival (p < 0.05). Conclusion: Staging FDG PET-CT has a major impact on the initial management of approximately one-third of patients with LACC by altering treatment intent and/or radiotherapy planning. PET-defined nodal status is a poor prognostic indicator.
Article
The aim of this meta-analysis is to demonstrate whether diffusion-weighted magnetic resonance imaging (DWI) could assist in the precise diagnosis of cervical cancer or not. Both English and Chinese electronic databases were searched for potential relevant studies followed by a comprehensive literature search without any language restriction. Two reviewers independently assessed the methodological quality of the included trials. Standardized mean difference (SMD) and its corresponding 95 % confidence interval (95 % CI) were calculated in this meta-analysis. We chose Version 12.0 STATA statistical software to analyze our statistical data. Thirteen eligible cohort studies were selected for statistical analysis, including 645 tumor tissues and 504 normal tissues. Combined SMD of apparent diffusion coefficient (ADC) suggested that the ADC value in cervical cancer tissues was significantly lower than that of normal tissue (SMD = 2.80, 95 % CI = 2.64 ~ 2.96, P < 0.001). Subgroup analysis stratified by ethnicity indicated a higher ADC value in the normal tissues compared to the cancer tissues in both the Asian and Caucasian subgroups (Asians: SMD = 2.83, 95 % CI = 2.64 ~ 3.02, P < 0.001; Caucasians: SMD = 2.73, 95 % CI = 2.45 ~ 3.01, P < 0.001, respectively). The results from the subgroup analysis by MRI machine type revealed a statistically significant difference in ADC value between normal cervical tissue and tumor tissues among all of the six MRI machine type subgroups (all P < 0.05). The main finding from our meta-analysis revealed that increased signal intensity on DWI and decreased signal on ADC seem to be useful in the diagnosis of cervical cancer. DWI could therefore be an important imaging tool in potentially identifying patients with cervical cancer.
Article
Artificial neural networks are appearing as useful alternatives to traditional statistical modelling techniques in many scientific disciplines. This paper presents a general introduction and discussion of recent applications of the multilayer perceptron, one type of artificial neural network, in the atmospheric sciences.
Article
Three-dimensional echocardiographic (3DE) analysis provides better measurements of left ventricular (LV) volumes, ejection fraction, myocardial deformation, and dyssynchrony. Many studies have shown that this technique has high intrainstitutional reproducibility. However, interinstitutional reproducibility is low, limiting its adoption. The aim of this study was to determine if standardization of training could reduce the interinstitutional variability in 3DE data analysis. In total, 50 full-volume, transthoracic 3DE data sets of the left ventricle were analyzed by two readers. Measurements obtained included LV volumes, ejection fraction, global longitudinal strain, and two dyssynchrony indices. The cases represented a wide spectrum of ejection fraction. After initial analysis of 21 studies, readers formally met to standardize their analytic approach on six additional cases. Five months after the intervention, 23 new cases were analyzed. Paired t tests were performed to identify systematic institutional differences in measurements. Interinstitutional variability was quantified using intraclass correlation coefficients and variability. Before the intervention, there was a systematic bias in LV volumes, which was eliminated after intervention. Intraclass correlation coefficients showed that the intervention improved agreement in measurements of LV volumes, strain, and dyssynchrony between the two centers and decreased variability. A simple intervention to standardize analysis can reduce interinstitutional variability of measurements obtained from 3DE analysis. This intervention is needed before the use of 3DE measurement in multicenter trials and to increase the reproducibility of such measurements in routine clinical practice.
Article
Support vector machine (SVM) is an extensively used machine learning method with many biomedical signal classification applications. In this study, a novel PSO-SVM model has been proposed that hybridized the particle swarm optimization (PSO) and SVM to improve the EMG signal classification accuracy. This optimization mechanism involves kernel parameter setting in the SVM training procedure, which significantly influences the classification accuracy. The experiments were conducted on the basis of EMG signal to classify into normal, neurogenic or myopathic. In the proposed method the EMG signals were decomposed into the frequency sub-bands using discrete wavelet transform (DWT) and a set of statistical features were extracted from these sub-bands to represent the distribution of wavelet coefficients. The obtained results obviously validate the superiority of the SVM method compared to conventional machine learning methods, and suggest that further significant enhancements in terms of classification accuracy can be achieved by the proposed PSO-SVM classification system. The PSO-SVM yielded an overall accuracy of 97.41% on 1200 EMG signals selected from 27 subject records against 96.75%, 95.17% and 94.08% for the SVM, the k-NN and the RBF classifiers, respectively. PSO-SVM is developed as an efficient tool so that various SVMs can be used conveniently as the core of PSO-SVM for diagnosis of neuromuscular disorders.
Article
A new technique for three-dimensional (3D) camera calibration for machine vision metrology using off-the-shelf TV cameras and lenses is described. The two-stage technique is aimed at efficient computation of camera external position and orientation relative to object reference coordinate system as well as the effective focal length, radial lens distortion, and image scanning parameters. The two-stage technique has advantage in terms of accuracy, speed, and versatility over existing state of the art. A critical review of the state of the art is given in the beginning. A theoretical framework is established, supported by comprehensive proof in five appendixes, and may pave the way for future research on 3D robotics vision. Test results using real data are described. Both accuracy and speed are reported. The experimental results are analyzed and compared with theoretical prediction. Recent effort indicates that with slight modification, the two-stage calibration can be done in real time.
Article
A novel region-based active contour model (ACM) is proposed in this paper. It is implemented with a special processing named Selective Binary and Gaussian Filtering RegularizedLevel Set(SBGFRLS) method, which first selectively penalizes the level set function to be binary, and then uses a Gaussian smoothing kernel to regularize it. The advantages of our method are as follows. First, a new region-based signed pressure force (SPF) function is proposed, which can efficiently stop the contours at weak or blurred edges. Second, the exterior and interior boundaries can be automatically detected with the initial contour being anywhere in the image. Third, the proposed ACM with SBGFRLS has the property of selective local or global segmentation. It can segment not only the desired object but also the other objects. Fourth, the level set function can be easily initialized with a binary function, which is more efficient to construct than the widely used signed distance function (SDF). The computational cost for traditional re-initialization can also be reduced. Finally, the proposed algorithm can be efficiently implemented by the simple finite difference scheme. Experiments on synthetic and real images demonstrate the advantages of the proposed method over geodesic active contours (GAC) and Chan–Vese (C–V) active contours in terms of both efficiency and accuracy.
Article
Over the last few decades pattern classification has been one of the most challenging area of research. In the present-age pattern classification problems, the support vector machines (SVMs) have been extensively adopted as machine learning tools. SVM achieves higher generalization performance, as it utilizes an induction principle called structural risk minimization (SRM) principle. The SRM principle seeks to minimize the upper bound of the generalization error consisting of the sum of the training error and a confidence interval. SVMs are basically designed for binary classification problems and employs supervised learning to find the optimal separating hyperplane between the two classes of data. The main objective of this paper is to introduce a most promising pattern recognition technique called cross-correlation aided SVM based classifier. The idea of using cross-correlation for feature extraction is relatively new in the domain of pattern recognition. In this paper, the proposed technique has been utilized for binary classification of EEG signals. The binary classifiers employ suitable features extracted from crosscorrelograms of EEG signals. These cross-correlation aided SVM classifiers have been employed for some benchmark EEG signals and the proposed method could achieve classification accuracy as high as 95.96% compared to a recently proposed method where the reported accuracy was 94.5%.
Article
The goal of this study was to prospectively analyze the diagnostic performances of magnetic resonance imaging (MRI) and positron emission tomography (PET)/computed tomography (CT) in predicting pathologically assessed residual disease in a large, single-institution series of locally advanced cervical cancer (LACC) patients triaged to neoadjuvant treatments followed by radical surgery. Between April 2007 and March 2010, 96 patients with histologically documented cervical cancer (any histology) and FIGO stage IB2-IVA were enrolled. MRI and PET/CT were recommended to be performed within 4-6 weeks from the end of treatment, and histology was the reference standard. Sensitivity, specificity, and accuracy were compared using the McNemar test. For residual disease in the cervix, sensitivity was higher for MRI than for PET/CT (86.1% vs 63.1%; P = .002), while specificity was significantly higher for PET/CT compared with MRI (P = .002). There was no difference in accuracy values between the 2 imaging modalities. For MRI analysis of lymph node groups, sensitivity, specificity, and accuracy were 35.7%, 95.9%, and 88.0%, respectively. Conversely, sensitivity, specificity, and accuracy for PET/CT were 28.6%, 97.8%, and 88.7%, respectively. Absence of follicular structures replaced by prevalent sclerosis and/or sinus histiocytosis was the most frequently documented morphological pattern in false-positive cases. Neither MRI nor PET/CT accurately detected residual disease in LACC patients triaged to radical surgery after neoadjuvant treatment, disallowing the option of avoiding or modulating completion surgery.
Article
This paper proposes and evaluates the application of support vector machine (SVM) to classify upper limb motions using myoelectric signals. It explores the optimum configuration of SVM-based myoelectric control, by suggesting an advantageous data segmentation technique, feature set, model selection approach for SVM, and postprocessing methods. This work presents a method to adjust SVM parameters before classification, and examines overlapped segmentation and majority voting as two techniques to improve controller performance. A SVM, as the core of classification in myoelectric control, is compared with two commonly used classifiers: linear discriminant analysis (LDA) and multilayer perceptron (MLP) neural networks. It demonstrates exceptional accuracy, robust performance, and low computational load. The entropy of the output of the classifier is also examined as an online index to evaluate the correctness of classification; this can be used by online training for long-term myoelectric control operations.
Article
Aortic dissecting aneurysm is one of the most catastrophic cardiovascular emergencies that carries high mortality. It was pointed out from clinical observations that the aneurysm development is likely to be related to the hemodynamics condition of the dissected aorta. In order to gain more insight on the formation and progression of dissecting aneurysm, hemodynamic parameters including flow pattern, velocity distribution, aortic wall pressure and shear stress, which are difficult to measure in vivo, are evaluated using numerical simulations. Pulsatile blood flow in patient-specific dissecting aneurismal aortas before and after the formation of lumenal aneurysm (pre-aneurysm and post-aneurysm) is investigated by computational fluid dynamics (CFD) simulations. Realistic time-dependent boundary conditions are prescribed at various arteries of the complete aorta models. This study suggests the helical development of false lumen around true lumen may be related to the helical nature of hemodynamic flow in aorta. Narrowing of the aorta is responsible for the massive recirculation in the poststenosis region in the lumenal aneurysm development. High pressure difference of 0.21 kPa between true and false lumens in the pre-aneurismal aorta infers the possible lumenal aneurysm site in the descending aorta. It is also found that relatively high time-averaged wall shear stress (in the range of 4-8 kPa) may be associated with tear initiation and propagation. CFD modeling assists in medical planning by providing blood flow patterns, wall pressure and wall shear stress. This helps to understand various phenomena in the development of dissecting aneurysm.
Article
Atherosclerosis, a disease of large- and medium-size arteries, is the chief cause of death in the United States and in most of the western world. Severe atherosclerosis interferes with blood flow; however, even in the early stages of the disease, i.e. during atherogenesis, there is believed to be an important relationship between the disease processes and the characteristics of the blood flow in the arteries. Atherogenesis involves complex cascades of interactions among many factors. Included in this are fluid mechanical factors which are believed to be a cause of the highly focal nature of the disease. From in vivo studies, there is evidence of hemodynamic influences on the endothelium, on intimal thickening, and on monocyte recruitment. In addition, cell culture studies have demonstrated the important effect of a cell's mechanical environment on structure and function. Most of this evidence is for the endothelial cell, which is believed to be a key mediator of any hemodynamic effect, and it is now well documented that cultured endothelial monolayers, in response to a fluid flow-imposed laminar shear stress, undergo a variety of changes in structure and function. In spite of the progress in recent years, there are many areas in which further work will provide important new information. One of these is in the engineering of the cell culture environment so as to make it more physiologic. Animal studies also are essential in our efforts to understand atherogenesis, and it is clear that we need better information on the pattern of the disease and its temporal development in humans and animal models, as well as the specific underlying biologic events.(ABSTRACT TRUNCATED AT 250 WORDS)
Article
During the past 7 yr, definitive corrective surgery for acute as well as chronic dissecting aneurysms has developed to the point where, in experienced hands, successful results in the range of 75% can be anticipated. The same 7 yr has seen the introduction and establishment of intensive drug therapy as the initial therapy of choice in all dissecting aneurysms, the primary therapy of choice in many dissecting aneurysms, and a useful adjunctive type of therapy in those patients where definitive surgery is indicated. Definitive diagnosis utilizing aortography to identify the site of the intimal tear and the extent of the dissecting hematoma plus evaluation of the general condition of the patient are the keys to the best use of those two modes of therapy and should lead to a success rate of 90% or better in the near future.
Article
Arterial remodeling in response to atherosclerosis may be outward (positive) or inward (negative) and is an important mechanism in the clinical manifestations of atherosclerosis and restenosis after percutaneous coronary interventions. Postmortem and intravascular ultrasound studies of arterial remodeling do not allow serial and noninvasive data to be obtained. In a rabbit model of atherosclerosis, we sought to validate MRI as a new tool for documentation of arterial remodeling. Watanabe heritable hyperlipidemic rabbits underwent serial MRI at baseline and 6 months after aortic balloon denudation. The lumen area had a small but significant (P=0.006) increase, from 4.36+/-0.16 to 4. 89+/-0.12 mm(2). There was a large, significant (P<0.0001) increase in the outer wall area, from 7.96+/-0.19 to 10.46+/-0.19 mm(2). The vessel wall area (a marker of atherosclerotic burden) increased significantly (P<0.0001), from 3.61+/-0.07 to 5.57+/-0.09 mm(2). Thus, the increase in atherosclerotic burden over time was completely accounted for by positive arterial remodeling. The subgroup used for histopathological validation confirmed a significant (P<0.0001) agreement between histopathology and MRI for assessment of all 3 parameters. MRI can provide serial and noninvasive data about the arterial wall, allowing assessment of arterial remodeling in this rabbit model. Thus, MRI appears to be a useful tool for the investigation of arterial remodeling both in native atherosclerosis and after percutaneous coronary intervention.
Article
Without prompt diagnosis and treatment, aortic dissection is rapidly fatal. While standard chest radiography may give clues to the diagnosis of aortic dissection, suspected dissection can be confirmed by only 4 imaging techniques: aortography, echocardiography, computed tomography, and magnetic resonance imaging. The following review discusses each of these methods. It also explains why aortography, the previous diagnostic benchmark, has been replaced by newer techniques and why magnetic resonance imaging has become the diagnostic method of choice for imaging aortic dissection.
Article
Atherosclerotic disease and its thrombotic complications remain the leading causes of mortality and morbidity in Western society. In Australia, cardiovascular disease is responsible for one in every 2.4 (41%) deaths and is the leading single cause of mortality. The crucial final common process for the conversion of a non-occlusive, often clinically silent, atherosclerotic lesion to a potentially fatal condition is plaque disruption. The mortality associated with atherosclerotic disease relates to the acute coronary syndromes, including acute myocardial infarction, unstable angina pectoris and sudden cardiac death. There is substantial clinical, experimental and postmortem evidence demonstrating the role acute thrombosis upon a disrupted atherosclerotic plaque plays in the onset of acute coronary syndromes. Atherosclerotic plaque composition, rather than the stenotic severity, appears to be central in determining risk of both plaque rupture and subsequent thrombogenicity. In particular, a large lipid core and a thin fibrous cap render an atherosclerotic lesion susceptible or vulnerable to these complications. We are currently limited in our ability to accurately identify patients at risk for an acute coronary event. The armamentarium of diagnostic investigations, both non-invasive and invasive, currently clinically available is only able to provide us with data related to the stenotic severity of a coronary artery. The non-invasive testing includes stress-induced (exercise or pharmacological) ischaemic changes in electrical repolarisation, wall motion or myocardial radioactive-tracer uptake. The invasive test of coronary angiography, although the current 'gold standard' for the detection of coronary atherosclerotic disease, provides us with no data about the composition of the atherosclerotic lesion. However, the vast majority of acute coronary events involve a non-critically stenosed atherosclerotic lesion, and thus with currently available means of identification, these lesions would be undetected by stress testing/imaging techniques. Given the critical role that atherosclerotic lesion composition has been shown to play in the risk of both plaque rupture and subsequent thrombogenicity and, consequently, an acute coronary event, new detection techniques need to be investigated for the task of documenting atherosclerotic lesion composition. In the present review we will focus on the status of imaging modalities available for coronary artery imaging and how they may advance our understanding and management of patients with and at risk of coronary artery disease in the new millennium.
A visually guided mobile platform acting in a dynamic environment needs information, delivered from cues such as stereo and motion. In this paper a binocular stereo system, capable of dynamic vergence and ego-motion estimation, will be presented. The ability to dynamically verge the cameras is essential for visual tasks such as tracking, recognition and manipulation, since the computation of 3D geometry will be considerably simplified at fixation. We will show that the epipolar geometry of the presented system can be estimated in real-time. The number of degrees of freedom has been minimized, without affecting the flexibility of the system. Reconstructed 3D data is used in order to determine the ego-motion of the platform, at a very low computational cost. Finally, it will be shown how disparity maps and estimated ego-motion can facilitate the localization of independently moving objects
Article
The purpose of this paper is to explore the representation capability of radial basis function (RBF) neural networks. The main results are: 1) the necessary and sufficient condition for a function of one variable to be qualified as an activation function in RBF network is that the function is not an even polynomial, and 2) the capability of approximation to nonlinear functionals and operators by RBF networks is revealed, using sample data either in frequency domain or in time domain, which can be used in system identification by neural networks
  • S K Bandt
  • E C Leuthardt
S.K. Bandt and E.C. Leuthardt, Minimally Invasive Neurosurgery for Epilepsy Using Stereotactic MRI Guidance, Neurosurgery Clinics of North America 27 (2016), 51–58.
A real-time system for epipolar geometry and ego-motion estimation, in Computer Vision and Pattern Recognition, 2000. Proceedings
  • Bjorkman
  • S G Worthley
S.G. Worthley et al., Coronary artery imaging in the new millennium, Heart, Lung and Circulation 11 (2002), 193 19-25.
Real-time monitoring system with accelerator controlling: An improvement of radiotherapy monitoring 169 based on binocular location and classification
  • L Chai
L. Chai et al., Real-time monitoring system with accelerator controlling: An improvement of radiotherapy monitoring 169 based on binocular location and classification, Journal of X-ray Science and Technology 2017. DOI: 10.3233/XST-
Electroclinical semiology of the bilateral asymmetric tonic seizures observed in patients with supplemen-204 tary sensorimotor area epilepsy confirmed by pre-and post-operative MRI
  • G Cai
G. Cai et al., Electroclinical semiology of the bilateral asymmetric tonic seizures observed in patients with supplemen-204 tary sensorimotor area epilepsy confirmed by pre-and post-operative MRI, Journal of X-ray Science and Technology 205 2017. DOI: 10.3233/XST-17157
Automatic epileptic seizure detection in EEGs using MF-DFA, SVM based on cloud computing, 212 Journal of X-ray Science and Technology
  • Z Zhang
Z. Zhang et al., Automatic epileptic seizure detection in EEGs using MF-DFA, SVM based on cloud computing, 212 Journal of X-ray Science and Technology 2017. DOI: 10.3233/XST-17158
Features extraction and multi-classification of sEMG using a GPU-Accelerated GA/MLP hybrid 220 algorithm
  • W Luo
W. Luo et al., Features extraction and multi-classification of sEMG using a GPU-Accelerated GA/MLP hybrid 220 algorithm, Journal of X-ray Science and Technology 2017. DOI: 10.3233/XST-17159.