The observed differences in the clinical and radiological features of psRLS and psPLMS suggest that the pathophysiologies of the two conditions are distinct. Further research is needed to understand the pathophysiologies of primary RLS and PLMS.

We examined the effects of ankle and knee joint cooling on 20-m sprint times and maximal vertical jump heights during high-intensity intermittent exercise. 21 healthy collegiate male basketball (n=14) and handball players (n=7) underwent 3 experimental sessions. Each session consisted of four 15-min quarters of high-intensity intermittent exercises including various intensities of 20-m shuttle running and jumping. A 20-min bilateral joint cooling (ankle, knee, or control-no cooling: in a counterbalanced order) was applied before quarters 1 and 3. After joint cooling, no warm-up activity other than the exercise protocol was given. The 20-m sprint times and maximal vertical jump heights in each experimental session were recorded at baseline (prior to quarter-1) and during each quarter. To test joint cooling effects over time, we performed 3×5 mixed model ANOVAs. Neither ankle nor knee joint cooling changed 20-m sprint times (F8,280=1.45; p=0.18) or maximal vertical jump heights (F8,280=0.76; p=0.64). However, a trend was observed in which joint cooling immediately decreased (quarters 1 and 3) but active warm-up for approximately 20 min improved 20-min sprint times (quarters 2 and 4). Our study suggests that athletic performance such as sprinting and jumping are not altered by joint cooling applied prior to or during high-intensity intermittent exercise.

In deep learning architectures, rectified linear unit based functions are widely used as activation functions of hidden layers, and the softmax is used for the output layers. Two critical problems of the softmax are introduced, and an improved softmax method to resolve the problems is proposed. The proposed method minimises instability of the softmax while reducing its losses. Moreover, this method is straightforward so its computation complexity is low, but it is substantially reasonable and operates robustly. Therefore, the proposed method can replace the softmax functions.

A novel estimation of the collision response between a virtual object and an arbitrarily-shaped real object is presented. By analysing the collision regions, the boundary of the arbitrarily-shaped object is approximated as a sphere for simplicity, and then the collision responds as a sphere-to-sphere collision. Simulations between segmented hand regions and virtual balls show that the proposed method is more appropriate than other methods for augmented reality and serious games, because the motion of real objects can be directly applied.

No abstract