Melanoma Image Classification Based on Multivariate Parametric Statistical Tests of Hypothesis

2019 ◽  
pp. 89-109
Author(s):  
K. Seetharaman
Keyword(s):  
Skin Lesion ◽  
Statistical Tests ◽  
Normal Skin ◽  
Region Of Interest ◽  
Feature Space ◽  
Advanced Stage ◽  
Initial Stage ◽  
Advanced Stages ◽  
Novel Method ◽  
Tests Of Hypotheses

This chapter proposes a novel method, based on the multivariate parametric statistical tests of hypotheses, which classifies the normal skin lesion images and the various stages of the melanoma images. The melanoma images are categorized into two classes, such as initial stage and advanced stage, based on the degree of aggressiveness of the cancer. The region of interest is identified and segmented from the input skin melanoma image. The features, such as HSV color, shape, and texture, are extracted from the region of interest. The features are treated as a feature space, which is assumed to be a multivariate normal random field. The proposed statistical tests are employed to identify and classify the melanoma images. The proposed method yields an average correct classification up to 91.55% for the normal skin lesion versus the initial and the advanced stages of the melanoma images, up to 91.39% for initial stage melanoma versus the normal skin lesion and the advanced stages melanoma, and up to 92.27% for the advanced stage melanoma versus the normal skin lesion and the initial stage melanoma. The proposed method yields better results.

scholarly journals Vascular phenotyping of brain tumors using magnetic resonance microscopy (μMRI)

2011 ◽  
Vol 31 (7) ◽  
pp. 1623-1636 ◽  
Author(s):  
Eugene Kim ◽  
Jiangyang Zhang ◽  
Karen Hong ◽  
Nicole E Benoit ◽  
Arvind P Pathak
Keyword(s):  
Brain Tumor ◽  
Magnetic Resonance ◽  
Brain Tumors ◽  
Three Dimensional ◽  
Region Of Interest ◽  
Tumor Model ◽  
Magnetic Resonance Microscopy ◽  
Vascular Phenotype ◽  
Brain Tumor Model ◽  
Novel Method

Abnormal vascular phenotypes have been implicated in neuropathologies ranging from Alzheimer's disease to brain tumors. The development of transgenic mouse models of such diseases has created a crucial need for characterizing the murine neurovasculature. Although histologic techniques are excellent for imaging the microvasculature at submicron resolutions, they offer only limited coverage. It is also challenging to reconstruct the three-dimensional (3D) vasculature and other structures, such as white matter tracts, after tissue sectioning. Here, we describe a novel method for 3D whole-brain mapping of the murine vasculature using magnetic resonance microscopy (μMRI), and its application to a preclinical brain tumor model. The 3D vascular architecture was characterized by six morphologic parameters: vessel length, vessel radius, microvessel density, length per unit volume, fractional blood volume, and tortuosity. Region-of-interest analysis showed significant differences in the vascular phenotype between the tumor and the contralateral brain, as well as between postinoculation day 12 and day 17 tumors. These results unequivocally show the feasibility of using μMRI to characterize the vascular phenotype of brain tumors. Finally, we show that combining these vascular data with coregistered images acquired with diffusion-weighted MRI provides a new tool for investigating the relationship between angiogenesis and concomitant changes in the brain tumor microenvironment.

scholarly journals Embedding Undersampling Rotation Forest for Imbalanced Problem

2018 ◽  
Vol 2018 ◽  
pp. 1-15 ◽  
Author(s):  
Huaping Guo ◽  
Xiaoyu Diao ◽  
Hongbing Liu
Keyword(s):  
Imbalanced Data ◽  
Feature Space ◽  
Original Data ◽  
Training Set ◽  
Data Set ◽  
Minority Class ◽  
Rotation Forest ◽  
Novel Method ◽  
Individual Classifier ◽  
The Cost

Rotation Forest is an ensemble learning approach achieving better performance comparing to Bagging and Boosting through building accurate and diverse classifiers using rotated feature space. However, like other conventional classifiers, Rotation Forest does not work well on the imbalanced data which are characterized as having much less examples of one class (minority class) than the other (majority class), and the cost of misclassifying minority class examples is often much more expensive than the contrary cases. This paper proposes a novel method called Embedding Undersampling Rotation Forest (EURF) to handle this problem (1) sampling subsets from the majority class and learning a projection matrix from each subset and (2) obtaining training sets by projecting re-undersampling subsets of the original data set to new spaces defined by the matrices and constructing an individual classifier from each training set. For the first method, undersampling is to force the rotation matrix to better capture the features of the minority class without harming the diversity between individual classifiers. With respect to the second method, the undersampling technique aims to improve the performance of individual classifiers on the minority class. The experimental results show that EURF achieves significantly better performance comparing to other state-of-the-art methods.

Cross-Border Mergers and Acquisitions and the Competitive Positioning of European National Industries

1997 ◽  
Vol 2 (4) ◽  
pp. 359-390
Author(s):  
Massimo G. Colombo ◽  
Sergio Mariotti
Keyword(s):  
Mergers And Acquisitions ◽  
Statistical Tests ◽  
Foreign Direct Investments ◽  
Competitive Advantages ◽  
Competitive Position ◽  
Cross Border ◽  
Target Firms ◽  
Eclectic Paradigm ◽  
International Oligopoly ◽  
Tests Of Hypotheses

This paper relies on the eclectic paradigm of foreign direct investments and Porter's theory on the competitive advantages of nations to study the localisation of target firms of international M&As by European enterprises. Firms' propensities towards extra- and infra-European acquisitions are correlated with the competitive position of European national industries in the international arena. Strategic groups of national industries are created through a cluster analysis based on the Fortune lists of the 500 world largest enterprises. Logit econometric estimates and statistical tests of hypotheses suggest that the share of extra-European acquisitions is greater in a) sectors where European large firms have achieved leadership of the world oligopoly, and b) sectors where the competitive position of Europe is rather weak though stable. Instead, firms belonging to national industries which have been rapidly increasing their share of the international oligopoly during the ′80s concentrate their M&As within Europe. The same holds true for declining weak competitors.

Boosting Automated Sleep Staging Performance in Big Datasets using Population Sub-grouping

SLEEP ◽  
2021 ◽  
Author(s):  
Samaneh Nasiri ◽  
Gari D Clifford
Keyword(s):  
Deep Learning ◽  
Feature Space ◽  
Learning Approaches ◽  
Sleep Staging ◽  
Learning Framework ◽  
Sensitivity Score ◽  
Instrument Noise ◽  
Novel Method ◽  
Clustering Approach ◽  
Statistical Relationships

Abstract Current approaches to automated sleep staging from the electroencephalogram (EEG) rely on constructing a large labeled training and test corpora by aggregating data from different individuals. However, many of the subjects in the training set may exhibit changes in the EEG that are very different from the subjects in the test set. Training an algorithm on such data without accounting for this diversity can cause underperformance. Moreover, test data may have unexpected sensor misplacement or different instrument noise and spectral responses. This work proposes a novel method to learn relevant individuals based on their similarities effectively. The proposed method embeds all training patients into a shared and robust feature space. Individuals that share strong statistical relationships and are similar based on their EEG signals are clustered in this feature space before being passed to a deep learning framework for classification. Using 994 patient EEGs from the 2018 Physionet Challenge (≈ 6,561 hours of recording), we demonstrate that the clustering approach significantly boosts performance compared to state-of-the-art deep learning approaches. The proposed method improves, on average, a precision score from 0.72 to 0.81, a sensitivity score from 0.74 to 0.82, and a Cohen’s Kappa coefficient from 0.64 to 0.75 under 10-fold cross-validation.

Both Hands’ Fingers’ Angle Calculation from Live Video

2012 ◽  
Vol 2 (2) ◽  
pp. 1-11 ◽  
Author(s):  
Ankit Chaudhary ◽  
Jagdish L. Raheja ◽  
Karen Das ◽  
Shekhar Raheja
Keyword(s):  
Gesture Recognition ◽  
Color Space ◽  
Three Dimensional ◽  
Region Of Interest ◽  
Natural Computing ◽  
Robotic Hand ◽  
Hsv Color Space ◽  
Binary Format ◽  
Novel Method ◽  
Angle Calculation

In the last few years gesture recognition and gesture-based human computer interaction has gained a significant amount of popularity amongst researchers all over the world. It has a number of applications ranging from security to entertainment. Gesture recognition is a form of biometric identification that relies on the data acquired from the gesture depicted by an individual. This data, which can be either two-dimensional or three-dimensional, is compared against a database of individuals or is compared with respective thresholds based on the way of solving the riddle. In this paper, a novel method for angle calculation of both hands’ bended fingers is discussed and its application to a robotic hand control is presented. For the first time, such a study has been conducted in the area of natural computing for calculating angles without using any wired equipment, colors, marker or any device. The system deploys a simple camera and captures images. The pre-processing and segmentation of the region of interest is performed in a HSV color space and a binary format respectively. The technique presented in this paper requires no training for the user to perform the task.

scholarly journals Programmed Death 1 Ligand Expression in the Monocytes of Patients with Hepatocellular Carcinoma Depends on Tumor Progression

Cancers ◽  
2020 ◽  
Vol 12 (8) ◽  
pp. 2286
Author(s):  
Akira Asai ◽  
Hidetaka Yasuoka ◽  
Masahiro Matsui ◽  
Yusuke Tsuchimoto ◽  
Shinya Fukunishi ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Cancer Progression ◽  
Tumor Regression ◽  
Advanced Stage ◽  
Tumoricidal Activity ◽  
Programmed Death ◽  
Early Hcc ◽  
Advanced Stages ◽  
Programmed Death 1 ◽  
After Cancer

Monocytes (CD14+ cells) from advanced-stage hepatocellular carcinoma (HCC) patients express programmed death 1 ligand (PD-L)/PD-1 and suppress the host antitumor immune response. However, it is unclear whether cancer progression is associated with CD14+ cells. We compared CD14+ cell properties before and after cancer progression in the same HCC patients and examined their role in antitumor immunity. CD14+ cells were isolated from 15 naïve early-stage HCC patients before treatment initiation and after cancer progression to advanced stages. Although CD14+ cells from patients at early HCC stages exhibited antitumor activity in humanized murine chimera, CD14+ cells from the same patients after progression to advanced stages lacked this activity. Moreover, CD14+ cells from early HCC stages scantly expressed PD-L1 and PD-L2 and produced few cytokines, while CD14+ cells from advanced stages showed increased PD-L expression and produced IL-10 and CCL1. CD14+ cells were also isolated from five naïve advanced-stage HCC patients before treatment as well as after treatment-induced tumor regression. The CD14+ cells from patients with advanced-stage HCC expressed PD-L expressions, produced IL-10 and CCL1, and exhibited minimal tumoricidal activity. After treatment-induced tumor regression, CD14+ cells from the same patients did not express PD-Ls, failed to produce cytokines, and recovered tumoricidal activity. These results indicate that PD-L expression as well as CD14+ cell phenotype depend on the tumor stage in HCC patients. PD-L expressions of monocytes may be used as a new marker in the classification of cancer progression in HCC.

scholarly journals Flood Mapping in Vegetated Areas Using an Unsupervised Clustering Approach on Sentinel-1 and -2 Imagery

2020 ◽  
Vol 12 (21) ◽  
pp. 3611
Author(s):  
Lisa Landuyt ◽  
Niko E. C. Verhoest ◽  
Frieke M. B. Van Coillie
Keyword(s):  
Optimal Parameter ◽  
Complex Mixture ◽  
Region Of Interest ◽  
Region Growing ◽  
Feature Space ◽  
Flood Mapping ◽  
Dry Land ◽  
Object Based ◽  
Forested Areas ◽  
Sar Imagery

The European Space Agency’s Sentinel-1 constellation provides timely and freely available dual-polarized C-band Synthetic Aperture Radar (SAR) imagery. The launch of these and other SAR sensors has boosted the field of SAR-based flood mapping. However, flood mapping in vegetated areas remains a topic under investigation, as backscatter is the result of a complex mixture of backscattering mechanisms and strongly depends on the wave and vegetation characteristics. In this paper, we present an unsupervised object-based clustering framework capable of mapping flooding in the presence and absence of flooded vegetation based on freely and globally available data only. Based on a SAR image pair, the region of interest is segmented into objects, which are converted to a SAR-optical feature space and clustered using K-means. These clusters are then classified based on automatically determined thresholds, and the resulting classification is refined by means of several region growing post-processing steps. The final outcome discriminates between dry land, permanent water, open flooding, and flooded vegetation. Forested areas, which might hide flooding, are indicated as well. The framework is presented based on four case studies, of which two contain flooded vegetation. For the optimal parameter combination, three-class F1 scores between 0.76 and 0.91 are obtained depending on the case, and the pixel- and object-based thresholding benchmarks are outperformed. Furthermore, this framework allows an easy integration of additional data sources when these become available.

scholarly journals Factors Related to Advanced Stage of Cancer Presentation in Botswana

2018 ◽  
pp. 1-9 ◽  
Author(s):  
Chidinma Anakwenze ◽  
Rohini Bhatia ◽  
William Rate ◽  
Lame Bakwenabatsile ◽  
Kebatshabile Ngoni ◽  
...  
Keyword(s):  
Universal Access ◽  
Stage Iv ◽  
Stage Iii ◽  
Advanced Stage ◽  
Care Seeking ◽  
Free Health Care ◽  
Symptom Awareness ◽  
Advanced Stages ◽  
Female Specific ◽  
Advanced Stage Cancer

Purpose Botswana, a country with a high prevalence of HIV, has an increasing incidence of cancer-related mortality in the post–antiretroviral therapy era. Despite universal access to free health care, the majority of Botswana patients with cancer present at advanced stages. This study was designed to explore the factors related to advanced-stage cancer presentation in Botswana. Methods Patients attending an oncology clinic between December 2015 and January 2017 at Princess Marina Hospital in Gaborone, Botswana, completed a questionnaire on sociodemographic and clinical factors as well as cancer-related fears, attitudes, beliefs, and stigma. Odds ratios (ORs) were calculated to identify factors significantly associated with advanced stage (stage III and IV) at diagnosis. Results Of 214 patients, 18.7% were men and 81.3% were women. The median age at diagnosis was 46 years, with 71.9% of patients older than 40 years. The most commonly represented cancers included cervical (42.3%), breast (16%), and head and neck (15.5%). Cancer stages represented in the study group included 8.4% at stage I, 19.2% at stage II, 24.1% at stage III, 11.9% at stage IV, and 36.4% at an unknown stage. Patients who presented at advanced stages were significantly more likely to not be afraid of having cancer (OR, 3.48; P < .05), believe that their family would not care for them if they needed treatment (OR, 6.35; P = .05), and believe that they could not afford to develop cancer (OR, 2.73; P < .05). The perception that symptoms were less serious was also significantly related to advanced stage ( P < .05). Patients with non–female-specific cancers were more likely to present in advanced stages (OR, 5.67; P < .05). Conclusion Future cancer mortality reduction efforts should emphasize cancer symptom awareness and early detection through routine cancer screening, as well as increasing the acceptability of care-seeking, especially among male patients.

A Novel Method of Visual Attention for Targets Detection

2013 ◽  
Vol 347-350 ◽  
pp. 3764-3768 ◽  
Author(s):  
Zhuo Zhang ◽  
Xin Nan Fan ◽  
Xue Wu Zhang ◽  
Hai Yan Xu ◽  
Min Li
Keyword(s):  
Visual Attention ◽  
Information Extraction ◽  
Feature Space ◽  
Gain Factor ◽  
Top Down ◽  
Bottom Up ◽  
Attention Model ◽  
Proposed Model ◽  
Novel Method ◽  
Statistical Prior

Inspired by the research of human visual system in neuroanatomy and psychology, the paper proposes a two-way collaborative visual attention model for target detection.In this new method , bottom-up attention information cooperates with top-down attention information to detect a target rapidly and accuractly. Firstly,the statistical prior knowledge of target and background is applied to optimize bottom-up attention information in different feature space and scale space.Secondly, after the SNR of salience difference between target and interference is computed ,the bottom-up gain factor is obtained.Thirdly, the gain factor is applied to adjust bottom up attention information extraction and then to maximize the salience contrast of target and background.Finally, target is detected by adjusted saliency.Experimental results shows that the proposed model in this paper can improve the real-time capability and reliability of target detection.

Sign in / Sign up

Export Citation Format

Share Document