When a car crashes, the aorta may rupture by pressing the chest with a seat belt. Therefore, a tool that can evaluate injury of the aorta is needed. The purpose of this study is to develop an analytical model that can express the injured state of the aorta by MPS method. At first, a cylindrical model with a simple shape was created to represent the pressure by blood, and an appropriate time step at which pressure propagated was determined. Next, a model of a dumbbell shape was prepared, a tensile test simulation was carried out, and the maximum stress of the aorta was determined by comparison with the experimental value based on the literature.

Many studies are conducted on the mechanism of cerebral aneurysm development, which is the main factor of subarachnoid hemorrhage, with a high mortality rate and severe sequelae. In particular, although the relationship between Wall Shear Stress (WSS) and vascular endothelial cells has been pointed out, quantitative Has not been evaluated. Therefore, we developed a vascular damage assessment system combining in vivo and in vitro. As a result, it became possible to quantitatively evaluate biological tissues by fluorescence microscopy. In the previous study, reduction and detachment of living cells was regarded as a factor of the occurrence of aneurysm, but from the viewpoint of pathology, cell death called apoptosis is considered to be the cause. Therefore, by observing the mechanical response of vascular endothelial cells to WSS, the relationship between WSS and cell death was examined. Based on this fact, the characteristics of each staining solution used in the experiment were clarified. In addition, it was found from the experimental results that it is difficult to observe apoptosis and apoptotic corpuscle, but the condition that was roughly positioned as cell death in the previous study was considered to be secondary necrosis.

Endovascular abdominal aortic aneurysm repair (EVAR) is less invasive and has brought good short-term outcomes. In EVAR, a stent graft is guided and placed in abdominal aortic aneurysms (AAA) using catheter, and spread out to create an internal bypass in the AAA. However, the distal end of stent graft sometimes migrates upward and occurs an endoleak especially in case of AAA with severe angulation. This means that aneurysm rupture after EVAR. On the basis of these backgrounds the purpose of this project is to clarify the mechanism of stent graft migration. To discuss biomechanical mechanism of stent graft migration, we identified the bending stiffness of the stent graft by 4-point bending test. We measure reaction force to forced load in 0 to 90 degrees in several types of stent graft. In addition to the results showed the bending properties of stent graft depend on the structure of stent graft. By using the elucidated mechanical properties of stent graft, we developed finite element models of stent graft and validate the models. We analysis 4point bending test of stent graft under the same condition. Moreover, we analysis geometry of stent graft for 3 types, calculated the optimal geometry of stent graft analysis model. Those validated model will be used for stent graft migration simulation to clarify the criteria to avoid stent graft migration after EVAR.

Small-diameter artificial blood vessels with an inner diameter of less than 6 mm are obstacles such as thrombus formation, and are not currently in clinical use. One of the causes of thrombus formation is the difference in mechanical properties between artificial blood vessels and native blood vessels. A native blood vessel has a characteristic that it hardly expands its inner diameter as the internal pressure of the vessel rises. In this study, monolayer biodegradable polymer scaffolds were fabricated by electrospinning method and examined the relation between the internal pressure and the outer diameter change rate was investigated. As a result, it was confirmed that the relaxed polymer scaffold could reproduce the mechanical properties of native blood vessel.

Cerebral blood flow is constantly fluctuating, but physiological source of the fluctuation remains uncertain. Previous studies have suggested that high-frequency component of the flow fluctuation reflects periodical oscillation of systemic physiology, such as respiration and heartbeat, while that low-frequency fluctuation originates from localized neural and/or glial activity in the brains. It is therefore expected that continuous non-invasive monitoring of the low-frequency fluctuation of cerebral blood flow allows us to know the physiologic states of the brains. In the present study, we aimed to clarify the origin of low-frequency fluctuation of the cerebral blood flow using animal models. To directly visualize and quantify spatiotemporal fluctuations of cerebral blood flow, we used genetically modified rats that express fluorescent proteins in the erythrocyte, and vessel structure and erythrocyte flow were imaged with two-photon microscopy in vivo. Dynamic imaging was conducted in the parenchyma at depths up to 0.2 mm from the cortical surface, and a linear discriminant analysis was applied to determine pixel identifies, such as blood plasma labeled sulforhodamine 101, red blood cells (RBCs) expressing the fluorescent proteins, and non-fluorescent backgrounds in each frame. Finally, time-dependent changes in pixel occupancy with blood plasma and RBCs in the vessel area were characterized, and apparent dwell time of the RBCs in each pixel was calculated. Our results showed that slow fluctuations of the RBC movements mainly arise from a branch of the capillary networks, while the high-frequency fluctuation was dominant in large arteries and veins.

This study investigates spatial distributions of red blood cells (RBCs) in a simplified capillary network, using a three-dimensional numerical simulation for a fluid-capsule interaction problem. The number of RBCs in the capillary network (called Hematocrit) is roughly controlled and varied by changing the period supplying RBCs into the system, and the RBC distributions in middle channels are investigated through introducing a distribution index. The results show that variations of the RBCs going through different channels are reduced with increasing the overall Hematocrit, owing to a passive regulation mechanism in fluid-mechanical point of view. This outcome infers an important role in oxygen transport and feeding into a tissue region in microcirculation systems.

Blast furnaces are large industrial structures to produce iron. Since scaffolding and wearing of furnace walls are caused in their long lifecycle, furnaces have to be repeatedly inspected and renovated. In conventional diagnosis for blast furnaces, inspectors visually estimate the amount of scaffold and wear to decide whether restoration is required. To precisely and efficiently estimate the amount of scaffold and wear, dense point-clouds of furnaces would be useful. In this paper, we discuss methods for detecting scaffolding and wearing on furnace walls using point-clouds. We detect scaffolding and wearing as differences between the original wall surface and measured points. However, it is not easy to estimate wall shapes with no scaffolding and wearing, because most blast furnaces were not precisely built as designed in drawings. Therefore, we estimate the original wall surfaces as the reference, and verify whether scaffolding and wearing regions can be detected. We also discuss methods for detecting small deterioration such as cracks. Since detectable deterioration depends on the resolution of reference surfaces, we introduce reference surfaces with multiple resolutions, and detect large and small scale deterioration.

For the simulations in industrial facilities, 3D models generated based on the current status are required. In this case, point-clouds captured by terrestrial laser scanner are effective for creating 3D models. So far, many researchers have proposed methods for shape reconstruction of industrial facilities based on the extraction of cylinders and planes. However, in manufacturing plants for liquids and gases, rotating surfaces and generalized cylinder surfaces are often used as well as cylinders and planes. It is difficult to extract these shapes from noisy and incomplete point-clouds, because these shapes include surfaces which have many-degrees of freedom. In this paper, we discuss methods for precisely reconstructing rotating surfaces and generalized cylinders. For extracting rotating surfaces and generalized cylinders, point-clouds are converted to wireframe models, and section points are calculated using section planes. Then, ellipses are extracted on each section planes, and they are connected for reconstructing rotating surfaces and generalized cylinders. We also discuss 3D shape reconstruction combining our method with voxel-based methods. Our method is robust to missing points caused by occlusion. Voxel-based methods can generate general shapes, but they are unstable to missing points. We show that the quality of shape reconstruction can be improved by combining these methods.

Nowadays, machine learning has attracted in the industrial field. However, there is a risk when we use machine learning as a black box in the production system and the like. Consequently, it is important to develop the rule extraction system which extracts rules from machine learning engines. Therefore, the proposition of this research is the development of the rule extraction system which provides certainty domain to improve reliability. In the proposed method, though there is the area not recognized near the boundary, in the recognized area we confirm the improvement of the accuracy. Also, the rules were simplified when the certainty domain is low.