Human activity recognition based on the inertial information and convolutional neural network / 生物医学工程学杂志

With the rapid improvement of the perception and computing capacity of mobile devices such as smart phones, human activity recognition using mobile devices as the carrier has been a new research hot-spot. The inertial information collected by the acceleration sensor in the smart mobile device is used for human activity recognition. Compared with the common computer vision recognition, it has the following advantages: convenience, low cost, and better reflection of the essence of human motion. Based on the WISDM data set collected by smart phones, the inertial navigation information and the deep learning algorithm-convolutional neural network (CNN) were adopted to build a human activity recognition model in this paper. The K nearest neighbor algorithm (KNN) and the random forest algorithm were compared with the CNN network in the recognition accuracy to evaluate the performance of the CNN network. The classification accuracy of CNN model reached 92.73%, which was much higher than KNN and random forest. Experimental results show that the CNN algorithm model can achieve more accurate human activity recognition and has broad application prospects in predicting and promoting human health.

Value of dual-time-point 18FDG PET-CT imaging on involved-field radiotherapy for hilar and mediastinal metastatic lymph nodes in non-small cell lung cancer / 中华放射肿瘤学杂志

Objective To discuss the value of dual-time-point 18FDG PET-CT imaging on involved field radiotherapy for hilar and mediastinal metastatic lymph nodes in patients with non-small cell lung cancer (NSCLC).Methods Fifty-four patients with NSCLC were included in this analysis,including 34 men and 20 women with mean age of 59(34-76)years.Two sequential PET-CT scans given 3-5 days before surgery were standard single-time-point imaging for the whole body and delayed imaging for the thorax.The pathologic data were used as golden standard to determine the difference between the standard single-time-point and dual-time-point FET-CT imaging in the definition of gross target volume(GTV)of involved-field radiotherapy for metastatic lymph nodes. Results For hilar metastatic lymph nodes,the GTV defined by single-time-point imaging was consistent with pathologic GTV in 21 patients(39%),comparing with 31 patients(57%) by dual-time-point imaging.Using pathologic data as golden standard,GTV alteration defined by single-time-point imaging had statisticaly significant difference comparing with that defined by dual-time-point imaging(u=519.00,P=0.023).For mediastinal metastatic lymph nodes,the GTV defined by single-time-point imaging was consistent with pathologic GTV in 30 patients(56%),comparing with 36 patients(67%)by dual-time-point imaging.Using pathologic data as golden standard.GTV alteration defined by single-time-point imaging had no statisticaly significant difference comparing with that defined by dual-time-point imaging(u=397.50,P=0.616).Conclusions For patients with NSCLC receiving involved-field radiotherapy,GTV definition for hilar and mediastinal metastatic lymph nodes by dual-time-point imaging is more consistent with that by pathologic data.Dual-time-point imaging has a larger value in terms of target delineation for hilar and mediastinal metastatic lymph nodes.