1. Pathological changes in endovascular fistulae The establishment of AVF in the distal limb is associated with thrombosis or poor maturation with a 20-50% chance. The conventional approach to avoid this problem is to use ultrasound to examine the limb vessels and screen for suitable vessels and locations. In general, the growth of blood flow after AVF establishment is dependent on many factors, including vascular geography, such as vessel diameter and length; vascular network structure, such as the number of venous collateral branches and the number of vessels; the resistance of the capillary bed and the structure of the anastomosis. Allon and Robbin studied the surgical success of arteriovenous endovascular fistulas from 1977 to mid-2002 and found a 25% early failure rate and a mean one-year original survival rate of 70%. The main problem affecting the function of arteriovenous endovascular fistulas is stenosis, which arises from hyperplasia of the endothelium. The proliferation mechanism is divided into upstream and downstream events. Upstream events include the response of endothelial cells and smooth muscle cells to stimulation and injury, and downstream events include the movement of adhesion molecules, chemokines and inflammatory mediators, leukocyte adhesion, and migration of smooth muscle cells induced by the response. In uremic patients, advanced age, diabetes mellitus, and obesity and cardiovascular disease can lead to atherosclerotic changes in the vasculature, which increases the risk of arteriovenous endovascular fistula procedure failure. Thus, the vascular status of the patient prior to fistula determines the prognosis of AVF. allon et al. found, through a study of vascular sampling in failed endovascular fistulas, that intimal hyperplasia occurred right after the AVF procedure. When intravascular hemodynamics act on endothelial cell receptors and smooth muscle cells, the vessel wall thickens, and usually when the vessel is injured, the artery is remodeled, and this remodeling is divided into 2 types, benign and non-benign remodeling. Roca-Tey et al. found that the main site of stenosis in AVF was the venous outlet followed by the anastomotic site by measuring the blood flow in patients with internal fistula. In contrast, Yong He et al. found that the lumen area at the venous end of the AVF gradually expanded over time by imaging an ESRD hemodialysis patient’s endovascular fistula from the initial establishment using MRI imaging, velocimetry, and later hydrodynamic analysis using CFD, while the venous area at the anastomosis did not change or even shrank. The KDOQI guidelines state that AVF should be performed in at least 50% of patients, while the percentage of AVF should be at 40%. The criteria for evaluating the maturity of the endovascular fistula are not uniform, and there are two types, one considers that the endovascular fistula flow can reach a minimum of 350-450 ml/min and a single dialysis session can be maintained for 3-5 hours. The other is that after 4 months of fistula establishment, the fistula flow should reach a minimum of 300 ml/min and be able to perform 8 consecutive effective dialysis sessions. The increase in blood flow after the establishment of the endovascular fistula is related to the pressure step difference and total vascular resistance. Statistically, it has been found that when an endovascular fistula is established, the flow in the flexural artery increases from an average of 20 ml/min to a flow of 300 ml/min in the fistula if it is a flexural arterial cephalic anastomosis. Although it was previously thought that the smaller the vessel diameter, the lower the chance of successful fistula, some studies have found that fistula failure is possible even when the artery diameter is sometimes 3 mm. Wong et al. found that fistula success was related to fistula flow on the first postoperative day, while Dixon et al. suggested that arteriovenous vasodilation and the chance of rapid increase in blood flow are key factors in determining fistula maturation. In contrast, Roy-Chaudhury et al. found that moderate vasodilation was effective in inhibiting the extent of endothelial hyperplasia, and that in the absence of vasodilation, even minor endothelial hyperplasia resulted in endovascular stenosis. tronc et al. found that endothelial cells lost their endothelial cell properties when stimulated by changes in blood flow, resulting in decreased vasodilatation. Thus, vascular diastolicity rather than diameter is the determining factor in the success of endovascular fistulae. At three weeks of endovascular fistula establishment, there was a significant increase in flexural artery flow shear, at which time, although both arteriovenous diameters tended to increase, radial artery intima-media thickness became thinner compared to cephalic veins, and cephalic vein intima-media thickness decreased to some extent. One year after the fistula was established, the diameter of the cephalic vein continued to increase, while the diameter of the radial artery did not change from three weeks; at the same time, the wall thickness of the radial artery continued to increase, while the wall thickness of the cephalic vein did not change; at the same time, the intima-media thickness of the radial artery also tended to increase, while the intima-media thickness of the cephalic vein did not change. However, there was also no significant change in the intra-arterial shear force from three weeks. Usually endovascular fistula maturation fails in three ways, 1 arterial dilation failure, mainly due to arterial endothelial calcification or peripheral vascular disease; 2 venous dilation failure 3 accelerated venous intimal hyperplasia, because local low shear after endovascular fistula establishment induces intimal hyperplasia, while local venous stenosis also leads to intimal hyperplasia; in addition, venous intimal injury during surgery also induces venous stenosis, which leads to intimal hyperplasia. Robbin et al. found by continuous observation of vessel diameter in patients for 2 months after endovascular fistula establishment that 95% of patients were able to meet dialysis requirements with venous internal diameter >10 px and endovascular blood flow >500 ml/min. However, these parameters do not continue to increase with time. Venous stenosis and the presence of collateral veins are two important factors contributing to AVF failure. And early application of endovascular intervention techniques can save AVF in time to reopen it. 2. Hydrodynamic changes in endovascular fistula Monica Sigovan et al. by observing 3 patients with established arteriovenous endovascular fistula using MRI digital imaging and velocity monitoring system for 3 months found that the mean shear force (WSS) tended to decrease gradually during the observation period, while endovascular shear force in patients with end-lateral anastomosis did not change during the observation period. In contrast, Sanjay Misra et al. performed an intracarotid arteriovenous anastomosis of PTFE vascular grafts by using pigs, followed by 14 days of observation using MRA and MRI techniques, mainly measuring blood flow in the distal and proximal sides of the artery exiting the anastomosis, the distal and proximal sides of the vein, shear force changes, and contralateral arteriovenous blood flow and shear force changes. It was found that shear forces were greatest at the anastomosis at 14 days, along with stenosis and increased blood flow velocity. In contrast, arteriovenous flow distal to the anastomosis was lower than arterial flow on the control side. Ruben Dammers et al. used ultrasound to observe arteriovenous diameters before and after fistula in 16 ESRD patients and calculated shear forces and found a significant increase in cephalic venous flow on the first day after endovascular fistula establishment, but no significant change in peak shear forces, considering that the increase in flow at this time was related to dilatation of the vein itself. At three weeks after the establishment of the internal fistula, the shear force of radial artery blood flow increased significantly, and at this time, although the arteriovenous diameter tended to increase, the intima-media thickness of the radial artery became thinner than that of the cephalic vein, and the intima-media thickness of the cephalic vein also decreased to some extent. One year after the fistula was established, the diameter of the cephalic vein continued to increase, while the diameter of the radial artery did not change from three weeks; at the same time, the wall thickness of the radial artery continued to increase, while the wall thickness of the cephalic vein did not change; at the same time, the intima-media thickness of the radial artery also tended to increase, while the intima-media thickness of the cephalic vein did not change. However, there was also no significant change in intra-arterial shear force from three weeks. Most studies use the Poiseuille formula for calculating shear force, but it has some drawbacks, such as the fact that the arteries are pulsatile and the vascular system is dilated; also the viscosity of the blood and the changes in the increase in blood flow can affect the accuracy of the calculation. Conventional ultrasound measurement of blood vessels may lead to poor prediction of preoperative and postoperative AVF prognosis due to ignoring various physiological parameters as well as pulse waves. There are three reasons for this: 1 the accuracy of the measurements, data entry and environmental parameter settings can limit the connection between the simulated and real environment; 2 the simulated model ignores the adaptive function of the vessel itself and the regulatory role of the peripheral vascular bed; and 3 some physical test descriptions do not fully reflect the expansion of pressure and pulse. Therefore, specific parameters such as blood flow viscosity coefficient and pulse wave intensity need to be entered to correct the hydrodynamic simulations. after AVF is established, blood flow increases to 30 times the original level. AVF is usually started 3 months before hemodialysis is established. Difficulties in maturation occur in 50% of low endovascular fistulas, mainly due to distal limb ischemia and heart failure. Therefore, predictive tools are needed to optimize the selection of the location of the endovascular fistula prior to AVF, as well as to be able to individually predict the increase in endovascular flow to the patient. In general, distal limb ischemia and heart failure occur when flow is greater than 30% of cardiac output after endovascular fistula creation, and W. Huberts et al. used the creation of an ex vivo simulation of increased vascular flow after AVF to verify the accuracy of the description in the actual physiological state. Andrea Remuzzi et al. found through hydrodynamic modeling that the smaller the AVF pinch angle, the smaller the region where low shear occurs. They also came to use MRI combined with CFD to image the phenomenon of endovascularization in 2 AVF patients at 4 weeks postoperatively and found that the actual measured results were similar to those simulated using the technique of CFD alone, thus confirming that CFD can simulate the actual situation. The peak shear force plays a dominant role in the factors leading to AVF vasodilation, rather than the average shear force. This is because 4 weeks after AVF establishment, peak shear force did not change, while mean shear force increased more than fivefold. Meanwhile Aron S. Bode et al. used a computational hemodynamic model to simulate vascular conditions in 25 patients to predict the location of AVF establishment, as well as postoperative blood flow, with a 75% compliance. From a certain point of view, computerized hemodynamic simulations can help those young physicians to perform preoperative assessment of patients’ vascular access, but validation in large randomized controlled studies is still needed. The increase in endovascular flow occurs within 48 h after the establishment of the AVF, and S. Manini et al. showed a high correlation in estimating the postoperative vessel diameter and flow changes in patients with AVF vessels using computer simulated blood flow pulse vibration network data. It could be useful in the future for surgical modality selection, although of course it is necessary to fit the data to different surgical modalities. This simulation may also help to further investigate the important role of endothelial cell dysfunction in endovascular fistula malfunction. This study was able to simulate changes in arterial diameter and blood flow over time after AVF, so it was hypothesized that there were no significant endothelial cell changes at the proximal end of the endovascular fistula despite high flow. hull et al. used CFD simulations to create AVF and increasing arteriovenous diameter based on 3 mm to simulate blood flow after creation of the endovascular fistula. The model was divided into two types of procedures: lateral anastomosis (STS) and end-lateral anastomosis (ETS). It was found that there was a pressure step difference in the internal shear force between the lateral anastomosis and the 90° end-lateral anastomosis. When the ETS was 45°, the pressure step difference was the lowest, the venous outflow was the lowest, and the shear strength was moderate. The use of CFD to simulate the AVF anastomosis angle can be useful for preoperative selection of the internal fistula anastomosis to avoid intimal hyperplasia, arterial regurgitation and blood theft. And it was also found in a porcine model that when the arterial length was extended 1.5 to 3 times at 90° ETS and STS, blood flow at the venous outflow end was increased 5 to 10 times. In contrast, at ETS between 3° and 58°, no arterial regurgitation occurs and arterial blood flow reaches 900 ml/min. Computer simulation of blood flow pulse waves allows calculation of blood flow within the branch of the vessel after the establishment of AVF occurs and is more patient specific than the 3D model. It can incorporate nonlinear equations to calculate the relationship between cross-sectional area and pressure, as well as pressure step differences at stenoses, bends, or anastomoses. This method involves summarizing the characteristic parameters of the arterial and branch systems into different nodes, each node representing a site of the arterial system, in which specific physiological parameters are incorporated, combined with some mathematical formulae algorithms to simulate and calculate the pressure and vessel wall shear and resistance conditions at that site in a lighted manner, and the various nodes are automatically analyzed and summarized so that a complete set of vascular system can be simulated blood flow pulse situation. This model is then applied to the clinic to present predicted values, which are then compared with the actual vascular parameters measured in the clinic, and finally the model is improved to guide the clinic. This approach, however, ignores the geographical map of the patient’s own vascular distribution, the latter being important for surgeons to plan the location of the procedure. Recently, it has been proposed to use boundary theory and velocity patterns to rapidly model vascular remodeling as well as blood flow during endovascular fistula maturation, while making this model more accurate in predicting endovascular maturation and changes due to screening for more patient-specific individualized parameters. To verify the accuracy of this model in clinical work, an ARCH study was initiated in 93 hemodialysis patients from four European centers over a 2-year period, in which vessel diameters and blood flow were routinely measured in the flexural, ulnar and cephalic veins of each patient, and events of endovascular fistula malfunction were recorded. It was found that in diabetic patients, the diameter of the brachial artery in the high fistula did not change significantly 40 days after the creation of the internal fistula, while the diameter of the flexural artery in the low fistula changed significantly. And the flow changes predicted using computer simulations were very close to those actually measured 40 days later. And those genetic and systemic factors do not affect the accuracy of the measurements. Anders Koustrup Niemann et al. observed changes in endovascular flow velocity and shear at different times by performing MRI imaging of the patient’s endovascular fistula, followed by simulations using CFD software. They found no clear correlation between endovascular morphology and shear force, but the model can help reveal the relationship between endovascular shear force and endovascular morphological changes, and can help healthcare professionals better understand the management of endovascular complications. In the future, image CFD techniques are needed to investigate the mechanism of the effect of the elasticity of the vessel wall on the turbulence generated within the AVF during simulated blood pulse flow.