Semi-automatic segmentation of skin lesions based on superpixels and hybrid texture information.

Dermoscopic images are commonly used in the early diagnosis of skin lesions, and several computational systems have been proposed to analyze them. The segmentation of the lesions is a fundamental step in many of these systems. Therefore, a semi-automatic segmentation method is proposed here, which begins by building the superpixels of the image under analysis based on the zero parameter version of the simple linear iterative clustering (SLIC0) algorithm. Then, each superpixel is represented using a descriptor built by combining the grey-level co-occurrence matrix and Tamura texture features. Afterward, the gain ratios of the features are used to select the input for the semi-supervised seeded fuzzy C-means clustering algorithm. Hence, from a few specialist-selected superpixels, this clustering algorithm groups the built superpixels into lesion or background regions. Finally, the segmented image undergoes a post-processing step to eliminate sharp edges. The experiments were performed on 1380 images: 401 images from the PH2 and DermIS datasets, which were used to establish the parameters of the method, and 3,573 images from the ISIC 2016, ISIC 2017 and ISIC 2018 datasets were used for the analysis of the method's performance. The findings suggest that, by manually identifying just a few of the generated superpixels, the method can achieve an average segmentation accuracy of 96.78%, which confirms its superiority to the ones in the literature.

Automatic segmentation of retinal layers in OCT images with intermediate age-related macular degeneration using U-Net and DexiNed.

Age-related macular degeneration (AMD) is an eye disease that can cause visual impairment and affects the elderly over 50 years of age. AMD is characterized by the presence of drusen, which causes changes in the physiological structure of the retinal pigment epithelium (RPE) and the boundaries of the Bruch's membrane layer (BM). Optical coherence tomography is one of the main exams for the detection and monitoring of AMD, which seeks changes through the evaluation of successive sectional cuts in the search for morphological changes caused by drusen. The use of CAD (Computer-Aided Detection) systems has contributed to increasing the chances of correct detection, assisting specialists in diagnosing and monitoring disease. Thus, the objective of this work is to present a method for the segmentation of the inner limiting membrane (ILM), retinal pigment epithelium, and Bruch's membrane in OCT images of healthy and Intermediate AMD patients. The method uses two deep neural networks, U-Net and DexiNed to perform the segmentation. The results were promising, reaching an average absolute error of 0.49 pixel for ILM, 0.57 for RPE, and 0.66 for BM.

Home e-Care: monitoramento de quedas em idosos através de smartwatches / Home e-Care: fall monitoring using smartwatches

Pessoas da terceira idade normalmente possuem a saúde mais frágil, elas ficam cada vez mais dependentes de ajuda na medida em que vão envelhecendo, principalmente na ocorrência de acidentes como quedas. Por esse motivo, torna-se necessário a existência de métodos automatizados de monitoramento e notificação da ocorrência de tais acidentes. OBJETIVO: Este trabalho apresenta uma metodologia para detecção automatizada de quedas em idosos, através do uso do acelerômetro presente em smartwatches com Android Wear e de detecção de quedas baseada em limiares. As ocorrências de queda são notificadas automaticamente para contatos de emergência. RESULTADOS: A aplicação criada alcançou resultados satisfatórios, obtendo 89,29% de especificidade, 75% de sensibilidade e 83,33% de acurácia. CONCLUSÃO: Concluiu-se que a abordagem adotada caracteriza uma alternativa robusta e viável para a detecção automatizada de quedas que contribui para a qualidade de vida para pessoas da terceira idade.

The elderly population is growing worldwide, and with that the concerns about their quality of life becomemore important. As the elderly normally have frail health, which tends to get worse as they age, they grow even more dependent on someone else, specifically in the event of an accident such as a fall. For that reason, an automated monitoring and notification mechanism for accidents is necessary. OBJECTIVE: This work aims to present an automated fall detection and notification methodology through the use of the android smartwatches embedded accelerometer andthreshold based algorithms. Notifications of falls will be sent to emergency contacts. RESULTS: The developed system reached satisfactory results, with 89,29% of specificity, 75% of recall and 83,33% of accuracy. CONCLUSION: We could observe that Home e-Care and the TBA are a robust and viable solution for automated fall detection, and can improve the elderly's quality of life.

Visualização imersiva e interação gestual com dados médicos / Immersive visualization and gestural interaction withmMedical data

As tecnologias de Realidade Virtual vêm se desenvolvendo bastante nos últimos anos e com elas a sua utilização em diversas áreas, dentre as quais, a medicina. Testes, treinamentos, e alguns tipos de tratamento que seriam complicados de serem ser feitos com abordagens tradicionais agora podem ser produzidos graças aos elementos disponíveis nas tecnologias de realidade virtual. OBJETIVO: Propor uma ferramenta de visualização volumétrica em realidade virtual que possua interação gestual e ferramentas de segmentação de imagens e que facilite o processo de análise de dados médicos. MÉTODOS: Aquisição das imagens, geração dos dados volumétricos, desenvolvimento das ferramentas de interação e desenvolvimento da interface gestual. RESULTADOS: O sistema obteve êxito na geração e visualização de dados médicos tendo bom desempenho em testes realizados na avaliação de usabilidade de sua interface gestual. CONCLUSÃO: O sistemas e mostra como uma promissora alternativa para a visualização de dados médicos em realidade virtual.

The Virtual Reality technologies have been developing greatly in recent years and with them their use in various fields, among which the medicine. Some testings, trainings, and some types of treatments that would be complicated to be made with traditional approaches can now be produced thanks to the elements available in the virtual reality technologies. OBJECTIVE: To propose a volume visualization tool in virtual reality that has gestural interaction and image segmentation tools and facilitates the process of analysis of medical data. METHODS: Image acquisition volumetric data generation, development of the interaction tools and development of the gestural interface. RESULTS: The system was successful in the generation and visualization of medical data, having good performance in usability tests of its gestural interface. CONCLUSION: The system is a promising alternative for viewing medical data in virtual reality.

Texture analysis of masses in digitized mammograms using Gleason and Menhinick diversity indexes

INTRODUCTION: Breast cancer is the second most common type of cancer in the world, being more common among women and representing 22% of all new cancer cases every year. The sooner it is diagnosed, the better the chances of a successful treatment are. Mammography is one way to detect non-palpable tumors that cause breast cancer. However, it is known that the sensitivity of this exam can vary considerably due to factors such as the specialist's experience, the patient's age and the quality of the images obtained in the exam. The use of computational techniques involving artificial intelligence and image processing has contributed more and more to support the specialists in obtaining a more precise diagnosis. METHODS: This paper proposes a methodology that exclusively uses texture analysis to describe features of masses in digitized mammograms. To increase the efficiency of texture feature extraction, the diversity index's capability to detect patterns of species co-occurrence is used. For this purpose, the Gleason and Menhinick indexes are used. Finally, the extracted texture is classified using the Support Vector Machine, looking to differentiate the malignant masses from the benign. RESULTS: The best result was obtained using the Gleason index, with 86.66% accuracy, 90% sensitivity, 83.33% specificity and an area under the ROC Curve (Az) of 0.86. CONCLUSION: Both indexes showed statistically similar performance; however, the Gleason index was slightly superior.

Classification of breast tissues using Moran's index and Geary's coefficient as texture signatures and SVM.

Female breast cancer is the major cause of cancer-related deaths in western countries. Efforts in computer vision have been made in order to help improving the diagnostic accuracy by radiologists. In this paper, we present a methodology that uses Moran's index and Geary's coefficient measures in breast tissues extracted from mammogram images. These measures are used as input features for a support vector machine classifier with the purpose of distinguishing tissues between normal and abnormal cases as well as classifying them into benign and malignant cancerous cases. The use of both proposed techniques showed to be very promising, since we obtained an accuracy of 96.04% and Az ROC of 0.946 with Geary's coefficient and an accuracy of 99.39% and Az ROC of 1 with Moran's index to discriminate tissues in mammograms as normal or abnormal. We also obtained accuracy of 88.31% and Az ROC of 0.804 with Geary's coefficient and accuracy of 87.80% and Az ROC of 0.89 with Moran's index to discriminate tissues in mammograms as benign and malignant.