scholarly journals Skin Lesion Classification Based on Surface Fractal Dimensions and Statistical Color Cluster Features Using an Ensemble of Machine Learning Techniques

Cancers ◽  
2021 ◽  
Vol 13 (21) ◽  
pp. 5256
Author(s):  
Simona Moldovanu ◽  
Felicia Anisoara Damian Michis ◽  
Keka C. Biswas ◽  
Anisia Culea-Florescu ◽  
Luminita Moraru
Keyword(s):  
Neural Network ◽  
Fractal Dimension ◽  
Skin Cancer ◽  
Skin Lesion ◽  
Fractal Dimensions ◽  
Color Distribution ◽  
Surface Fractal Dimension ◽  
Color Features ◽  
Surface Fractal ◽  
Lesion Classification

(1) Background: An approach for skin cancer recognition and classification by implementation of a novel combination of features and two classifiers, as an auxiliary diagnostic method, is proposed. (2) Methods: The predictions are made by k-nearest neighbor with a 5-fold cross validation algorithm and a neural network model to assist dermatologists in the diagnosis of cancerous skin lesions. As a main contribution, this work proposes a descriptor that combines skin surface fractal dimension and relevant color area features for skin lesion classification purposes. The surface fractal dimension is computed using a 2D generalization of Higuchi’s method. A clustering method allows for the selection of the relevant color distribution in skin lesion images by determining the average percentage of color areas within the nevi and melanoma lesion areas. In a classification stage, the Higuchi fractal dimensions (HFDs) and the color features are classified, separately, using a kNN-CV algorithm. In addition, these features are prototypes for a Radial basis function neural network (RBFNN) classifier. The efficiency of our algorithms was verified by utilizing images belonging to the 7-Point, Med-Node, and PH2 databases; (3) Results: Experimental results show that the accuracy of the proposed RBFNN model in skin cancer classification is 95.42% for 7-Point, 94.71% for Med-Node, and 94.88% for PH2, which are all significantly better than that of the kNN algorithm. (4) Conclusions: 2D Higuchi’s surface fractal features have not been previously used for skin lesion classification purpose. We used fractal features further correlated to color features to create a RBFNN classifier that provides high accuracies of classification.

scholarly journals Skin Lesion Classification With Deep Convolutional Neural Network: Process Development and Validation

10.2196/18438 ◽  
2020 ◽  
Vol 3 (1) ◽  
pp. e18438
Author(s):  
Arnab Ray ◽  
Aman Gupta ◽  
Amutha Al
Keyword(s):  
Neural Network ◽  
Skin Cancer ◽  
Skin Lesion ◽  
Convolutional Neural Network ◽  
Process Development ◽  
Early Stage ◽  
Deep Convolutional Neural Network ◽  
Fine Tuning ◽  
Network Process ◽  
Lesion Classification

Background Skin cancer is the most common cancer and is often ignored by people at an early stage. There are 5.4 million new cases of skin cancer worldwide every year. Deaths due to skin cancer could be prevented by early detection of the mole. Objective We propose a skin lesion classification system that has the ability to detect such moles at an early stage and is able to easily differentiate between a cancerous and noncancerous mole. Using this system, we would be able to save time and resources for both patients and practitioners. Methods We created a deep convolutional neural network using an Inceptionv3 and DenseNet-201 pretrained model. Results We found that using the concepts of fine-tuning and the ensemble learning model yielded superior results. Furthermore, fine-tuning the whole model helped models converge faster compared to fine-tuning only the top layers, giving better accuracy overall. Conclusions Based on our research, we conclude that deep learning algorithms are highly suitable for classifying skin cancer images.

A Microcontact Model Developed for Asperity Heights With a Variable Profile Fractal Dimension, a Surface Fractal Dimension, Topothesy, and Non-Gaussian Distribution

2008 ◽  
Author(s):  
Jen Luen Liou ◽  
Jen Fin Lin
Keyword(s):  
Fractal Dimension ◽  
Cross Sections ◽  
Fractal Dimensions ◽  
Surface Fractal Dimension ◽  
Surface Fractal ◽  
Deformation Regime ◽  
Circular Contour ◽  
The Mean ◽  
Non Gaussian ◽  
The Relationship

The cross sections formed by the contact asperities of two rough surfaces at an interference are island-shaped, rather than having the commonly assumed circular contour. These island-shaped contact surface contours show fractal behavior with a profile fractal dimension Ds. The surface fractal dimension for the asperity heights is defined as D and the topothesy is defined as G. In the study of Mandelbrot, the relationship between D and Ds was given as D = Ds+1 if these two fractal dimensions are obtained before contact deformation. In the present study, D, G, and Ds are considered to be varying with the mean separation (or the interference at the rough surface) between two contact surfaces. The D-Ds relationships for the contacts at the elastic, elastoplastic, and fully plastic deformations are derived and the inceptions of the elastoplastic deformation regime and the fully plastic deformation regime are redefined using the equality of two expressions established in two different ways for the number of contact spots (N).

scholarly journals TECHNIQUE FOR DETERMINATION OF SURFACE FRACTAL DIMENSION AND MORPHOLOGY OF MESOPOROUS TITANIA USING DYNAMIC FLOW ADSORPTION AND ITS CHARACTERIZATION

2010 ◽  
Vol 9 (1) ◽  
pp. 13-17
Author(s):  
Silvester Tursiloadi
Keyword(s):  
Fractal Dimension ◽  
Adsorption Isotherm ◽  
Surface Morphology ◽  
Average Pore Size ◽  
Fractal Dimensions ◽  
Nano Particles ◽  
Dynamic Flow ◽  
N2 Adsorption ◽  
Surface Fractal Dimension ◽  
Surface Fractal

A technique to determine the surface fractal dimension of mesoporous TiO­2 using a dynamic flow adsorption instrument is described. Fractal dimension is an additional technique to characterize surface morphology. Surface fractal dimension, a quantitative measurement of surface ruggedness, can be determined by adsorbing a homologous series of adsorbates onto an adsorbent sample of mesoporous TiO­2. Titania wet gel prepared by hydrolysis of Ti-alkoxide was immersed in the flow of supercritical CO2 at 60 °C and the solvent was extracted.  Mesoporous TiO­2 consists of anatase nano-particles, about 5nm in diameter, have been obtained. After calcination at 600 °C, the average pore size of the extracted gel, about 20nm in diameter, and the pore volume, about 0.35cm3g-1, and the specific surface area, about 58 m2g-1. Using the N2 adsorption isotherm, the surface fractal dimension, DS, has been estimated according to the Frenkel-Halsey-Hill (FHH) theory. The N2 adsorption isotherm for the as-extracted aerogel indicates the mesoporous structure. Two linear regions are found for the FHH plot of the as-extracted aerogel. The estimated surface fractal dimensions are about 2.49 and 2.68. Both of the DS  values indicate rather complex surface morphology. The TEM observation shows that there are amorphous and crystalline particles. Two values of DS may be attributed to these two kinds of particles. The two regions are in near length scales, and the smaller DS, DS =2.49, for the smaller region. This result indicates that there are two kinds of particles, probably amorphous and anatase particles as shown by the TEM observation.     Keywords: surface fractal dimensions, CO2 supercritically extraction, sol-gel, aerogel, titania

scholarly journals Fractal analysis of bentonite modified with heteropoly acid using nitrogen sorption and mercury intrusion porosimetry

2011 ◽  
Vol 76 (10) ◽  
pp. 1403-1410 ◽  
Author(s):  
Srdjan Petrovic ◽  
Zorica Vukovic ◽  
Tatjana Novakovic ◽  
Zoran Nedic ◽  
Ljiljana Rozic
Keyword(s):  
Fractal Dimension ◽  
Specific Surface ◽  
Surface Area ◽  
Specific Surface Area ◽  
Heteropoly Acid ◽  
Mercury Intrusion Porosimetry ◽  
Fractal Dimensions ◽  
Surface Fractal Dimension ◽  
Mercury Intrusion ◽  
Surface Fractal

Experimental adsorption isotherms were used to evaluate the specific surface area and the surface fractal dimensions of acid-activated bentonite samples modified with a heteropoly acid (HPW). The aim of the investigations was to search for correlations between the specific surface area and the geometric heterogeneity, as characterized by the surface fractal dimension and the content of added acid. In addition, mercury intrusion was employed to evaluate the porous microstructures of these materials. The results from the Frankel-Halsey-Hill method showed that, in the p/p0 region from 0.75 to 0.96, surface fractal dimension increased with increasing content of heteropoly acid. The results from mercury intrusion porosimetry (MIP) data showed the generation of mesoporous structures with important topographical modifications, indicating an increase in the roughness (fractal geometry) of the surface of the solids as a consequence of the modification with the heteropoly acid. By comparison, MIP is preferable for the characterization because of its wide effective probing range.

scholarly journals Skin Lesion Classification With Deep Convolutional Neural Network: Process Development and Validation (Preprint)

2020 ◽  
Author(s):  
Arnab Ray ◽  
Aman Gupta ◽  
Amutha Al
Keyword(s):  
Neural Network ◽  
Skin Cancer ◽  
Skin Lesion ◽  
Convolutional Neural Network ◽  
Process Development ◽  
Early Stage ◽  
Deep Convolutional Neural Network ◽  
Fine Tuning ◽  
Network Process ◽  
Lesion Classification

BACKGROUND Skin cancer is the most common cancer and is often ignored by people at an early stage. There are 5.4 million new cases of skin cancer worldwide every year. Deaths due to skin cancer could be prevented by early detection of the mole. OBJECTIVE We propose a skin lesion classification system that has the ability to detect such moles at an early stage and is able to easily differentiate between a cancerous and noncancerous mole. Using this system, we would be able to save time and resources for both patients and practitioners. METHODS We created a deep convolutional neural network using an Inceptionv3 and DenseNet-201 pretrained model. RESULTS We found that using the concepts of fine-tuning and the ensemble learning model yielded superior results. Furthermore, fine-tuning the whole model helped models converge faster compared to fine-tuning only the top layers, giving better accuracy overall. CONCLUSIONS Based on our research, we conclude that deep learning algorithms are highly suitable for classifying skin cancer images.

scholarly journals Skin Lesion Classification Using Convolutional Neural Network for Melanoma Recognition

2020 ◽  
Author(s):  
Aishwariya Dutta ◽  
Md. Kamrul Hasan ◽  
Mohiuddin Ahmad
Keyword(s):  
Neural Network ◽  
Skin Cancer ◽  
Skin Lesion ◽  
Convolutional Neural Network ◽  
Characteristic Curve ◽  
Experimental Studies ◽  
Visual Assessment ◽  
Low Contrast ◽  
High Degree ◽  
Lesion Classification

AbstractSkin cancer, also known as melanoma, is generally diagnosed visually from the dermoscopic images, which is a tedious and time-consuming task for the dermatologist. Such a visual assessment, via the naked eye for skin cancers, is a challenging and arduous due to different artifacts such as low contrast, various noise, presence of hair, fiber, and air bubbles, etc. This article proposes a robust and automatic framework for the Skin Lesion Classification (SLC), where we have integrated image augmentation, Deep Convolutional Neural Network (DCNN), and transfer learning. The proposed framework was trained and tested on publicly available IEEE International Symposium on Biomedical Imaging (ISBI)-2017 dataset. The obtained average area under the receiver operating characteristic curve (AUC), recall, precision, and F1-score are respectively 0.87, 0.73, 0.76, and 0.74 for the SLC. Our experimental studies for lesion classification demonstrate that the proposed approach can successfully distinguish skin cancer with a high degree of accuracy, which has the capability of skin lesion identification for melanoma recognition.

scholarly journals Relevance of coral geometry in the outcomes of the coral-algal benthic war

2018 ◽  
Author(s):  
Emma E. George ◽  
James Mullinix ◽  
Fanwei Meng ◽  
Barbara Bailey ◽  
Clinton Edwards ◽  
...  
Keyword(s):  
Fractal Dimension ◽  
Surface Area ◽  
Essential Element ◽  
Fractal Dimensions ◽  
Environmental Conservation ◽  
Microbial Composition ◽  
Surface Area Ratio ◽  
Surface Fractal Dimension ◽  
Surface Fractal ◽  
The Mean

AbstractCorals have built reefs on the benthos for millennia, becoming an essential element in marine ecosystems. Climate change and human impact, however, are favoring the invasion of non-calcifying benthic algae and reducing coral coverage. Corals rely on energy derived from photosynthesis and heterotrophic feeding, which depends on their surface area, to defend their outer perimeter. But the relation between geometric properties of corals and the outcome of competitive coral-algal interactions is not well known. To address this, 50 coral colonies interacting with algae were sampled in the Caribbean island of Curaçao. 3D and 2D digital models of corals were reconstructed to measure their surface area, perimeter, and polyp sizes. A box counting algorithm was applied to calculate their fractal dimension. The perimeter and surface dimensions were statistically non-fractal, but differences in the mean surface fractal dimension captured relevant features in the structure of corals. The mean fractal dimension and surface area were negatively correlated with the percentage of losing perimeter and positively correlated with the percentage of winning perimeter. The combination of coral perimeter, mean surface fractal dimension, and coral species explained 19% of the variability of losing regions, while the surface area, perimeter, and perimeter-to-surface area ratio explained 27% of the variability of winning regions. Corals with surface fractal dimensions smaller than two and small perimeters displayed the highest percentage of losing perimeter, while corals with large surface areas and low perimeter-to-surface ratios displayed the largest percentage of winning perimeter. This study confirms the importance of fractal surface dimension, surface area, and perimeter of corals in coral-algal interactions. In combination with non-geometrical measurements such as microbial composition, this approach could facilitate environmental conservation and restoration efforts on coral reefs.

DISTINCT FRACTAL CHARACTERISTICS OF MONOMER AND MULTIMER PROTEINS

Fractals ◽  
2010 ◽  
Vol 18 (02) ◽  
pp. 207-214 ◽  
Author(s):  
DANA CRACIUN ◽  
ADRIANA ISVORAN ◽  
R. D. REISZ ◽  
N. M. AVRAM
Keyword(s):  
Surface Roughness ◽  
Fractal Dimension ◽  
Structural Characteristics ◽  
Fractal Dimensions ◽  
The Other ◽  
Mean Values ◽  
Surface Fractal Dimension ◽  
Surface Fractal ◽  
Fractal Characteristics ◽  
The Mean

Within this study we have calculated the surface fractal dimension (Ds) and the backbone fractal dimensions associated to the local folding (D1) and to the global folding (D2) for two unbiased sets of 50 proteins each, one for monomer and the other for homo- multimer proteins. The mean surface fractal dimension is Ds = 2.29 ± 0.02 for monomers and Ds = 2.21 ± 0.01 for multimers, the two means being significantly different. The mean backbone fractal dimensions associated to the local folding are D1 = 1.34 ± 0.14 for monomers and D1 = 1.33 ± 0.11 for multimers and those associated to the global folding are D2 = 1.33 ± 0.05 for monomers and D2 = 1.29 ± 0.04 for multimers, respectively. There are not significant differences between the mean values of the backbone fractal dimensions corresponding to monomers and multimers. These results suggest that there are different structural characteristics between monomer and multimer proteins only concerning their surface roughness, with multimers being smoother than monomers.

Sign in / Sign up

Export Citation Format

Share Document