Patients with a high visceral fat area are more likely to develop an intra-abdominal infection after gastrectomy. Assessment of fat area, in particular visceral fat area, should alert surgeons to increased postoperative risks.

The present study demonstrated that increased fat content and large body shape have little effect on the number of lymph nodes retrieved and blood loss in LADG. However, for patients undergoing conventional open distal gastrectomy, increased fat content and large body shape do impact on the amount of blood lost and the number of lymph nodes retrieved.

CNN is more accurate than conventional machine learning methods that utilize the manual feature extraction.

During angiographic examinations, interventional radiologists need various image manipulations such as paging, roaming, zooming, windowing, fusion, summation, subtraction, image enhancement, and reference image display in order to assess forward vessels, nidus conditions, and sources of bleeding. However, radiologists cannot touch the screen by hand because their gloves must always be kept clean. Usually, radiologists provide verbal instructions to technologists, who may operate the workstation to display appropriate images with various types of image processing. However, if a technologist is not at the console because of preparation of the catheter or contrast agent, a quick operation may not be performed. Therefore, we developed an image operation system that uses a motion sensor for angiography. With this motion-sensor workstation, radiologists can give instructions for image display and manipulation merely with movements of the fingers. This is a new technology for angiography in operating rooms. For proper display of simulated viewing for the sequence of clinical digital subtraction angiography (DSA) images, we investigated a comparison of the average time required by the conventional method using the mouse and the finger operating method with a motion sensor. After practice of a total of 30 min, the average operation time by the finger method was significantly shorter than that by the mouse method. This motion sensor method with appropriate training about finger movements can provide a better operating performance than the conventional mouse method.

The method to lump a general kinematic wave equation of the slope runoff system is described. The method is as follows: At first, the basin topography is modeled as a set of slope units using digital topographic information. From the topographic model, the spatial distribution of the upslope contributing area is easily obtained. Secondly, the spatial distributions of the discharge and the flow area are calculated assuming that the rainfall runoff system is under a steady state. Finally, the storage volume of the slope system is calculated by spatially integrating the flow area and then the lumped relation between the storage volume and the outflow discharge is obtained. The flow area can not be always analytically integrated and some numerical computations are required. Then the technique to reduce the numerical computations is also given.