A Hybrid Approach Based on SVM and Bernoulli Mixture Model for Binary Vectors Classification

2020 ◽  
Author(s):  
Fahdah Alalyan ◽  
Nuha Zamzami ◽  
Nizar Bouguila
Keyword(s):  
Mixture Model ◽  
Hybrid Approach ◽  
Binary Vectors

scholarly journals Discriminative Learning Approach Based on Flexible Mixture Model for Medical Data Categorization and Recognition

Sensors ◽  
2021 ◽  
Vol 21 (7) ◽  
pp. 2450
Author(s):  
Fahd Alharithi ◽  
Ahmed Almulihi ◽  
Sami Bourouis ◽  
Roobaea Alroobaea ◽  
Nizar Bouguila
Keyword(s):  
Mixture Model ◽  
Hybrid Approach ◽  
Medical Data ◽  
Support Vector ◽  
Discriminative Learning ◽  
Learning Approach ◽  
X Rays ◽  
Intrinsic Nature ◽  
Discriminative Models ◽  
Dirichlet Mixture

In this paper, we propose a novel hybrid discriminative learning approach based on shifted-scaled Dirichlet mixture model (SSDMM) and Support Vector Machines (SVMs) to address some challenging problems of medical data categorization and recognition. The main goal is to capture accurately the intrinsic nature of biomedical images by considering the desirable properties of both generative and discriminative models. To achieve this objective, we propose to derive new data-based SVM kernels generated from the developed mixture model SSDMM. The proposed approach includes the following steps: the extraction of robust local descriptors, the learning of the developed mixture model via the expectation–maximization (EM) algorithm, and finally the building of three SVM kernels for data categorization and classification. The potential of the implemented framework is illustrated through two challenging problems that concern the categorization of retinal images into normal or diabetic cases and the recognition of lung diseases in chest X-rays (CXR) images. The obtained results demonstrate the merits of our hybrid approach as compared to other methods.

The Research on Image Segmentation for Fingerprint Identification

2011 ◽  
Vol 217-218 ◽  
pp. 396-401
Author(s):  
Xiao Jie Xu ◽  
Xi Yan Dong
Keyword(s):  
Image Segmentation ◽  
Mixture Model ◽  
Hybrid Approach ◽  
Identification Accuracy ◽  
Fingerprint Identification ◽  
Image Feature ◽  
Background Information ◽  
Bayesian Decision ◽  
Fingerprint Image ◽  
Directional Information

As the precondition of fingerprint identification, the effective image segmentation plays the significant role in the following image processing. Unlike other images, the fingerprint images are obviously directional. Aiming at this feature, in this paper, an image segmentation method based on the directional information of fingerprint image is introduced, which sufficiently utilizes the directional information of fingerprint image and succeeds in separating the background information. However, owing to the absence of directional information in some local areas of fingerprint image, this method will produce large segmentation errors, even fail. Therefore, for these local regions without directional information, it is proposed to apply Bayesian decision-making theory based on minimum error probability to realize image segmentation. On the assumption that the gray values accord with the probability distribution of Gaussian finite mixture model in image feature space, EM algorithm is used to estimate the parameters of mixture model. The mixture application of two methods can effectively separate the background information from fingerprint image while saving the preprocessing time and ensuring the following identification accuracy of fingerprint. The experiments illustrate the feasibility of the hybrid approach.

Surgical and endovascular hybrid approach in peripheral arterial disease of the lower limbs

VASA ◽  
2016 ◽  
Vol 45 (5) ◽  
pp. 417-422 ◽  
Author(s):  
Anouk Grandjean ◽  
Katia Iglesias ◽  
Céline Dubuis ◽  
Sébastien Déglise ◽  
Jean-Marc Corpataux ◽  
...  
Keyword(s):  
Mortality Rate ◽  
Peripheral Arterial Disease ◽  
Limb Salvage ◽  
Hybrid Approach ◽  
Arterial Disease ◽  
Early Mortality ◽  
Secondary Patency ◽  
Limb Salvage Rate ◽  
Peripheral Arterial ◽  
Salvage Rate

Abstract. Background: Multilevel peripheral arterial disease is frequently observed in patients with intermittent claudication or critical limb ischemia. This report evaluates the efficacy of one-stage hybrid revascularization in patients with multilevel arterial peripheral disease. Patients and methods: A retrospective analysis of a prospective database included all consecutive patients treated by a hybrid approach for a multilevel arterial peripheral disease. The primary outcome was the patency rate at 6 months and 1 year. Secondary outcomes were early and midterm complication rate, limb salvage and mortality rate. Statistical analysis, including a Kaplan-Meier estimate and univariate and multivariate Cox regression analyses were carried out with the primary, primary assisted and secondary patency, comparing the impact of various risk factors in pre- and post-operative treatments. Results: 64 patients were included in the study, with a mean follow-up time of 428 days (range: 4 − 1140). The technical success rate was 100 %. The primary, primary assisted and secondary patency rates at 1 year were 39 %, 66 % and 81 %, respectively. The limb-salvage rate was 94 %. The early mortality rate was 3.1 %. Early and midterm complication rates were 15.4 % and 6.4 %, respectively. The early mortality rate was 3.1 %. Conclusions: The hybrid approach is a major alternative in the treatment of peripheral arterial disease in multilevel disease and comorbid patients, with low complication and mortality rates and a high limb-salvage rate.

The association of Raynaud’s phenomenon/syndrome with scleroderma and Sjögren’s syndrome – a meta-analysis

VASA ◽  
2008 ◽  
Vol 37 (Supplement 73) ◽  
pp. 26-32 ◽  
Author(s):  
Schlattmann ◽  
Höhne ◽  
Plümper ◽  
Heidrich
Keyword(s):  
Quantitative Analysis ◽  
Sjögren’S Syndrome ◽  
Mixture Model ◽  
Sjögren's Syndrome ◽  
Meta Analysis ◽  
Sjogren’S Syndrome ◽  
Sjogren's Syndrome ◽  
Raynaud’S Syndrome ◽  
Raynaud's Syndrome ◽  
Sjögren‘S Syndrome

Background: In order to analyze the prevalence of Raynaud’s syndrome in diseases such as scleroderma and Sjögren’s syndrom – a meta-analysis of published data was performed. Methods: The PubMed data base of the National Library of Medicine was used for studies dealing with Raynaud’s syndrome and scleroderma or Raynaud’s syndroem and Sjögren’s syndrom respectively. The studies found provided data sufficient to estimate the prevalence of Raynaud’s syndrome. The statistical analysis was based on methods for a fixed effects meta-analysis and finite mixture model for proportions. Results: For scleroderma a pooled prevalence of 80.9% and 95% CI (0.78, 0.83) was obtained. A mixture model analysis found four latent classes. We identified a class with a very low prevalence of 11%, weighted with 0.15. On the other hand there is a class with a very high prevalence of 96%. Analysing the association with Sjögren’s syndrome, the pooled analysis leads to a prevalence of Raynaud’s syndrome of 32%, 95% CI(26.7%, 37.7%). A mixture model finds a solution with two latent classes. Here, 38% of the studies show a prevalence of 18.8% whereas 62% observe a prevalence of 38.3%. Conclusion: There is strong variability of studies reporting the prevalence of Raynaud’s syndrome in patients suffering from scleroderma or Sjögren’s syndrome. The available data are insufficient to perform a proper quantitative analysis of the association of Raynaud’s phenomenon with scleroderma or Sjögren’s syndrome. Properly planned and reported epidemiological studies are needed in order to perform a thorough quantitative analysis of risk factors for Raynaud’s syndrome.

Delayed Closure of Multiple Muscular Ventricular Septal Defects in an Infant after Coarctation Repair and a Hybrid Procedure�A Case Report

2011 ◽  
Vol 14 (1) ◽  
pp. 67 ◽  
Author(s):  
Ireneusz Haponiuk ◽  
Maciej Chojnicki ◽  
Radosaw Jaworski ◽  
Jacek Juciski ◽  
Mariusz Steffek ◽  
...  
Keyword(s):  
Septal Defect ◽  
Color Doppler ◽  
Hybrid Approach ◽  
Ventricular Septal Defects ◽  
Color Doppler Echocardiography ◽  
Septal Defects ◽  
Coarctation Repair ◽  
Step Procedure ◽  
The Right ◽  
Delayed Closure

There are several strategies of surgical approach for the repair of multiple muscular ventricular septal defects (mVSDs), but none leads to a fully predictable, satisfactory therapeutic outcome in infants. We followed a concept of treating multiple mVSDs consisting of a hybrid approach based on intraoperative perventricular implantation of occluding devices. In this report, we describe a 2-step procedure consisting of a final hybrid approach for multiple mVSDs in the infant following initial coarctation repair with pulmonary artery banding in the newborn. At 7 months, sternotomy and debanding were performed, the right ventricle was punctured under transesophageal echocardiographic guidance, and the 8-mm device was implanted into the septal defect. Color Doppler echocardiography results showed complete closure of all VSDs by 11 months after surgery, probably via a mechanism of a localized inflammatory response reaction, ventricular septum growth, and implant endothelization.

Sign in / Sign up

Export Citation Format

Share Document