Classification of radiographic lung pattern based on texture analysis and machine learning

This study evaluated the feasibility of using texture analysis and machine learning to distinguish radiographic lung patterns. A total of 1200 regions of interest (ROIs) including four specific lung patterns (normal, alveolar, bronchial, and unstructured interstitial) were obtained from 512 thoracic radiographs of 252 dogs and 65 cats. Forty-four texture parameters based on eight methods of texture analysis (first-order statistics, spatial gray-level-dependence matrices, gray-level-difference statistics, gray-level run length image statistics, neighborhood gray-tone difference matrices, fractal dimension texture analysis, Fourier power spectrum, and Law's texture energy measures) were used to extract textural features from the ROIs. The texture parameters of each lung pattern were compared and used for training and testing of artificial neural networks. Classification performance was evaluated by calculating accuracy and the area under the receiver operating characteristic curve (AUC). Forty texture parameters showed significant differences between the lung patterns. The accuracy of lung pattern classification was 99.1% in the training dataset and 91.9% in the testing dataset. The AUCs were above 0.98 in the training set and above 0.92 in the testing dataset. Texture analysis and machine learning algorithms may potentially facilitate the evaluation of medical images.

Efeito da oclusão de informações espaciais na cortada do voleibol sobre a tomada de decisão defensiva em atletas com diferentes níveis de experiência / The effect of spatial information occlusion in volleyball spike on defensive decision-making in athletes with different levels of experience

O presente estudo analisou o efeito da oclusão de informações espaciais na cortada do voleibol sobre a tomada de decisão defensiva em atletas com diferentes níveis de experiência. Os participantes foram divididos em grupo adulto (GAD; n=16), infanto/mirim (GIM; n=16) e adulto novato (GNO; n=16). Foram ocluídas 5 informações espaciais: bola (OE1), braço e mão (OE2), cabeça (OE3), tronco (OE4) e MMII como condição controle (OE5). Foi mensurada a precisão na predição da trajetória da bola e a confiança da resposta. O GAD foi mais preciso que os demais grupos na condição OE2 (P's<0,008), que proporcionou o pior desempenho dos grupos (P's<0,001). O GAD apresentou maior confiança que o GNO em todas as condições (P's<0,003), mas sem diferença em relação ao GIM (Bonferroni P's>0,036). O GIM apenas foi mais confiante que GNO em OE4 (P=0,01). Assim, as informações OE1 e OE2 demonstraram afetar mais o melhor desempenho dos participantes.

The aim of this study was to analyze the effect of the occlusion of spatial information in volleyball spike on defensive decision-making in athletes with different levels of experience. Participants were divided into adult (GAD; n=16), juvenile (GIM; n=16) and novice (GNO; n=16) groups. Five types of spatial information were occluded: ball (OE1), arm and hand (OE2), head (OE3), trunk (OE4), and lower limbs as a control condition (OE5). We measured the accuracy in predicting the ball's trajectory prediction and the confidence of the response. GAD was more precise than the other groups in the OE2 condition (P's<0.008), which provided the worst performance of the groups (P's<0.001). GAD showed more confidence than GNO in all conditions (P's<0.003), but with no difference compared to GIM (P's>0,036). GIM was only more confident than GNO in OE4 (P=0.01). Therefore, the OE1 and OE2 proved to have greater effect on the performance of the best participants.