In this paper, a new Computer-Aided Detection (CAD) system for the detection and classification of dangerous skin lesions (melanoma type) is presented, through a fusion of handcraft features related to the medical algorithm ABCD rule (Asymmetry Borders-Colors-Dermatoscopic Structures) and deep learning features employing Mutual Information (MI) measurements. The steps of a CAD system can be summarized as preprocessing, feature extraction, feature fusion, and classification. During the preprocessing step, a lesion image is enhanced, filtered, and segmented, with the aim to obtain the Region of Interest (ROI); in the next step, the feature extraction is performed. Handcraft features such as shape, color, and texture are used as the representation of the ABCD rule, and deep learning features are extracted using a Convolutional Neural Network (CNN) architecture, which is pre-trained on Imagenet (an ILSVRC Imagenet task). MI measurement is used as a fusion rule, gathering the most important information from both types of features. Finally, at the Classification step, several methods are employed such as Linear Regression (LR), Support Vector Machines (SVMs), and Relevant Vector Machines (RVMs). The designed framework was tested using the ISIC 2018 public dataset. The proposed framework appears to demonstrate an improved performance in comparison with other state-of-the-art methods in terms of the accuracy, specificity, and sensibility obtained in the training and test stages. Additionally, we propose and justify a novel procedure that should be used in adjusting the evaluation metrics for imbalanced datasets that are common for different kinds of skin lesions.
Written communication is pivotal for societies to develop. However, lexicon and depth of information vary greatly among texts according to their purpose. Scientific texts, diffusion of science reports, general and area-specific news are all written differently. Thus, we explore the characterization of different text categories through a nature-inspired feature known as the Hurst parameter. We contend that the Hurst exponent is useful to unveil the rhetorical structure within written documents. We collected and processed texts in five categories: scientific articles, diffusion of science reports, business news, entertainment news, and random texts. Each category contains 350 documents. We found that the median for scientific texts has the highest value of the Hurst parameter (0.575), followed by business news (0.54); the median for randomly-generated texts is 0.48, which lies in the region associated with random walks. The median value for diffusion texts is 0.49, and for entertainment texts is 0.53. However, these two categories present high dispersion. We conclude that the Hurst parameter is a measure that quantifies the structure of communication in the selected categories of texts. Application of our finding in the field of e-research is discussed.
El cáncer de piel es una enfermedad que afecta a las personas con tono de piel oscuro o claro. Por otra parte, cada vez más personas tienden a emplear camas de bronceado o permanecer periodos prolongados de tiempo ante los rayos del sol, provocando que esta enfermedad sea más frecuente. Como complemento en el diagnóstico de esta enfermedad existe la inteligencia artificial, la cual, permite emplear algoritmos de clasificación como árboles de decisión, máquinas de soporte vectorial, regresión logística, entre otros; además, del uso de algoritmos de aprendizaje profundo como las redes neuronales convolucionales, ayudando a realizar un pre-diagnostico de cáncer de piel. En este artículo se explicara el desarrollo de un método, en el cual, utilizando la base de datos de imágenes dermatológicas publicada por el Internacional Skin Imaging Collaboration (ISIC) \citep{ISIC:12} se considera un conjunto de imágenes las cuales ya han sido caracterizadas por especialistas y que se encuentran en un grupo de benignas y malignas.
No abstract
Gait analysis is one of the most important challenging research areas in clinical and computing settings. Gait biomechanics and gait human recognition are two major areas of interest. Alterations in walking can cause physical and metal health problems in people, so diagnoses and treatments derived from optimal gait analysis are very useful in clinical settings. This paper surveys the gait analysis methods, applications and platforms, gait biomechanics, as well as, gait recognition approaches, and datasets. Then, we describe contributions in gait forward kinematics, useful to assess gaits such as crouched and normal. Also, a framework for antalgic gait recognition based on human activity, using the gyroscope embedded in a smartphone is described. Different algorithms and metrics were used to perform the gait recognition, highlighting Support Vector Machines, Naive Bayes, k- Nearest Neighbours, and Accuracy and F-measure, respectively. Finally, we discuss the challenges and future perspectives on gait recognition.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.