Search

With the advent of cloud computing, it has become more and more popular to outsource various services to the cloud for releasing the burden of local data storage and maintenance. However, it may cause serious privacy problems because the cloud may be untrusted. In this paper, we study the privacy-preserving reverse nearest neighbor (PPRNN) query over encrypted spatial data. First, we introduce the concept of reference-locked order-preserving encryption (RL-OPE) with its construction and security proof, which reveals less information than traditional order-preserving encryption (OPE). Then, we present a novel PPRNN scheme in static setting based on structured encryption (SE) and the proposed RL-OPE, called sPPRNN. After that, we design a generic method that extends a PPRNN scheme in static setting to the counterpart in dynamic setting, called dPPRNN. Furthermore, we present a thorough privacy analysis of our proposal. Finally, we demonstrate its efficiency and effectiveness for practical deployment through extensive experiments.

Clinically significant portal hypertension is associated with an increased risk of developing gastroesophageal varices and hepatic decompensation. Hepatic venous pressure gradient measurement and esophagogastroduodenoscopy are the gold-standard methods for assessing clinically significant portal hypertension and gastroesophageal varices, respectively. However, invasiveness, cost, and feasibility limit their widespread use, especially if repeated and serial evaluations are required to assess the efficacy of pharmacotherapy. Artificial intelligence describes a range of techniques that allow machines to perform tasks typically thought to require human reasoning and problem-solving skills. Artificial intelligence has made great strides in the field of medicine, and is also involved in portal hypertension diagnosis. Artificial intelligence tools will potentially transform our practice by leveraging massive amounts of data to personalize care to the right patient, in the right amount, at the right time. This review focuses on the recent advances in artificial intelligence for the noninvasive diagnosis of portal hypertension and gastroesophageal varices and monitoring of risk assessment of its complications in clinical practice.

It is essential to release annealing induced strain during the crystallization process to realize efficient and stable perovskite solar cells (PSCs), which does not seem achievable using the conventional annealing process. Here we report a novel and facile thermal gradient assisted crystallization strategy by simply introducing a slant angle between the preheated hot plate and the substrate. A distinct crystallization sequence resulted along the in-plane direction pointing from the hot side to the cool side, which effectively reduced the crystallization rate, controlled the perovskite grain growth, and released the in-plane tensile strain. Moreover, this strategy enabled uniform strain distribution in the vertical direction and assisted in reducing the defects and aligning the energy bands. The corresponding device demonstrated champion power conversion efficiencies (PCEs) of 23.70% and 21.04% on the rigid and flexible substrates, respectively. These highly stable rigid devices retained 97% of the initial PCE after 1097 h of storage and more than 80% of the initial PCE after 1000 h of continuous operation at the maximum power point. This novel strategy opens a simple and effective avenue to improve the quality of perovskite films and photovoltaic devices via strain modulation and defect passivation.

Background and Purpose: Several pieces of evidence suggest that serum lactate hydrogenase (LDH) level is associated with the pathological process of delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH). This research aimed to investigate the associations of serum LDH level with the occurrence of DCI in aSAH patients. Methods: A total of 122 patients diagnosed with aSAH within 72h of onset were retrospectively enrolled. The serum levels of LDH between 7:00-8:00 am on day 1, day 3 and day 7, patients’ demographics, and clinical features were collected. Computed tomography perfusion was performed within 7 days after aSAH. The occurrence of DCI was recorded during the hospitalization. Results: Among all the enrolled patients, 43 (35.2%) developed DCI during hospitalization. Patients occurred DCI were always accompanied by more serious clinical features and found with higher serum LDH levels. LDH levels on day 3 and day 7 after onset were independently associated with the occurrence of DCI and showed high predictive value according to the receiver operating characteristic (ROC) curve. Moreover, there was a strong correlation between LDH and mean cerebral blood flow, transit time, and mean time to peak. Conclusions: Serum LDH level on day 3 and day 7 may be a valuable, convenient, and rapid predictive indicator for the occurrence of DCI in aSAH patients.

Organic-inorganic hybrid semiconducting (OIHS) materials, which can detect broader spectral regions, are highly desired in several applications including biomedical imaging, night vision, and optical communications. Although lead (Pb)-halide perovskites have reached a mature research stage, high toxicity of Pb hinders their large-scale viability. Tin (Sn)-based perovskites are the most common OIHS broadband light absorbers that replace toxic Pb; however, they are extremely unstable due to the notorious Sn

With the increasing popularity of Internet of Energy (IoE), more and more physical devices connected to IoE depend on the information system for energy control and coordination. Under this condition, how to achieve information flow control in IoE becomes a great challenge. Access control encryption (ACE) is a promising technology to address the problem. However, existing ACE requires a centralized sanitizer, hindering its successful application in IoE. In this paper, we construct a new kind of ACE without sanitizers for IoE. We first construct a basic ACE scheme based on number-theoretic assumptions (i.e., DBDH assumption), and this scheme can control not only what users can read but also what they can write. To resist against the quantum attacks, we further construct a more secure ACE scheme based on learning with errors (LWE). We formally prove that our proposed two ACE schemes are secure under the proposed security definition, and we also evaluate the applicability and the efficiency of them in experiments.

A deficiency in the photoelectrical dynamics at the interface due to the surface traps of the TiO2 electron transport layer (ETL) has been the critical factor for the inferiority of the power conversion efficiency (PCE) in the perovskite solar cells. Despite its excellent energy level alignment with most perovskite materials, its large density of surface defect as a result of sub lattice vacancies has been the critical hurdle for an efficient photovoltaic process in the device. Here, we report that atoms thick 2D TiS2 layer grown on the surface of a (001) faceted and single-crystalline TiO2 nanograss (NG) ETL have effectively passivated the defects, boosting the charge extractability, carrier mobility, external quantum efficiency, and the device stability. These properties allow the perovskite solar cells (PSCs) to produce a PCE as high as 18.73% with short-circuit current density (Jsc), open-circuit voltage (Voc), and fill-factor (FF) values as high as 22.04 mA/cm2, 1.13 V, and 0.752, respectively, a 3.3% improvement from the pristine TiO2-NG-based PSCs. The present approach should find an extensive application for controlling the photoelectrical dynamic deficiency in perovskite solar cells.

When performing endoluminal surgery inside the colon using a flexible endoscopic robot, there is a problem of losing visualization and task space due to the collapsing of the surrounding wall. This happens mainly because of the intra-abdominal pressure and peristalsis of the colon. Although insufflation is commonly used for expanding the colon, it does not function when a full-thickness incision is created on the colon wall. To support the collapsing colon and ensure sufficient visualization and task space even with a hole, we developed a deployable colon support structure (CSS) that can be seamlessly adapted to the existing procedures. While the CSS is designed to be small and flexible enough to pass through an endoscopic channel that can be tortuous, it becomes steady enough to hold the collapsing/squeezing colon after being deployed. Also, the CSS is collapsible after task completion, for retraction through an endoscopic channel. Through the ex-vivo and in-vivo studies with a swine, we have successfully demonstrated the feasibility of supporting the colon wall during endoluminal interventions with the CSS. We confirmed that the CSS was easily deliverable and deployable and the created space was large enough to perform surgical tasks using robotic arms. National Research Foundation (NRF) Published version This work was supported by the National Research Foundation (NRF) Singapore under Grant NRFI2016-07

Subarachnoid hemorrhage (SAH) is a devastating cerebral vascular disease which causes neurological deficits including long-term cognitive deficit. Demyelination of white matter is correlated with cognitive deficit in SAH. Electroacupuncture (EA) is a traditional Chinese medical treatment which protects against cognitive deficit in varies of neurological diseases. However, whether EA exerts protective effect on cognitive function in SAH has not been investigated. The underlying mechanism of remyelination regulated by EA remains unclear. This study aimed to investigate the protective effects of EA on cognitive deficit in a rat model of SAH. SAH was induced in SD rats (n = 72) by endovascular perforation. Rats in EA group received EA treatment (10 min per day) under isoflurane anesthesia after SAH. Rats in SAH and sham groups received the same isoflurane anesthesia with no treatment. The mortality rate, neurological score, cognitive function, cerebral blood flow (CBF), and remyelination in sham, SAH and EA groups were assessed at 21 d after SAH.EA treatment alleviated cognitive deficits and myelin injury of rats compared with that in SAH group. Moreover, EA treatment enhanced remyelination in white matter and promoted the differentiation of OPCs after SAH. EA treatment inhibited the expression of Id2 and promoted the expression of SOX10 in oligodendrocyte cells. Additionally, the cerebral blood flow (CBF) of rats was increased by EA compared with that in SAH group. EA treatment exerts protective effect against cognitive deficit in the late phase of SAH. The underlying mechanisms involve promoting oligodendrocyte progenitor cell (OPC) differentiation and remyelination in white matter via regulating the expression of Id2 and SOX10. The improvement of CBF may also account for the protective effect of EA on cognitive function. EA treatment is a potential therapy for the treatment of cognitive deficit after SAH.

Background and Aims. Gastroesophageal variceal bleeding is a serious complication of portal hypertension in cirrhotic patients and could be predicted by hepatic venous pressure gradient (HVPG). However, whether the presence of ascites affects the prognostic value of HVPG for patients with acute variceal bleeding is still unknown. This retrospective study is aimed at investigating the influence of ascites on predictive performance of HVPG for early rebleeding in cirrhotic patients with acute variceal bleeding. Methods. In this retrospective study, a total of 148 patients with cirrhosis hospitalized for acute variceal bleeding who underwent HVPG measurement and endoscopic variceal ligation (EVL) for the prevention of rebleeding were included. The receiver operating characteristic curve (ROC) and logistical regression method were employed to analyze the predictive performance of HVPG for early rebleeding. The locally weighted scatterplot smoothing approach was adopted to assess the monotonicity between bleeding risk and HVPG. Results. A significantly higher HVPG level was observed in patients with early rebleeding compared to patients without rebleeding in the nonascites cohort. When using HVPG to predict early rebleeding, there was a lower area under curve in the ascites cohort compared to the nonascites cohort. HVPG was recognized as a risk factor for early rebleeding by a logistic regression model only in the nonascites cohort. An overall monotonicity in the trend of change in HVPG and risk for early rebleeding was observed in the nonascites cohort solely. Conclusion. The predictive value of HVPG for early rebleeding in patients with cirrhosis that developed acute variceal bleeding is hindered by the presence of ascites.

BACKGROUND AND AIM Gastroesophageal varices (GEV) present in compensated advanced chronic liver disease (cACLD) and can develop into high-risk varices (HRV). The gold standard for diagnosing GEV is esophagogastroduodenoscopy (EGD). However, EGD is invasive and less tolerant. This study aimed to develop and validate radiomics signatures based on noncontrast-enhanced computed tomography (CT) images for non-invasive diagnosis of GEV and HRV in patients with cACLD. METHODS The multicenter trial enrolled 161 patients with cACLD from six university hospitals in China between January 2015 and September 2019, who underwent both EGD and noncontrast-enhanced CT examination within 14 days prior to the endoscopy. Two radiomics signatures, termed rGEV and rHRV, respectively, were built based on CT images in a training cohort of 129 patients and validated in a prospective validation cohort of 32 patients (ClinicalTrials. gov identifier: NCT03749954). RESULTS In the training cohort, both rGEV and rHRV exhibited high discriminative abilities on determining the existence of GEV and HRV with the area under receiver operating characteristic curve (AUC) of 0.941 (95% confidence interval [CI] 0.904-0.978) and 0.836 (95% CI 0.766-0.905), respectively. In validation cohort, rGEV and rHRV showed high discriminative abilities with AUCs of 0.871 (95% CI 0.739-1.000) and 0.831 (95% CI 0.685-0.978), respectively. CONCLUSIONS This study demonstrated that rGEV and rHRV could serve as the satisfying auxiliary parameters for detection of GEV and HRV with good diagnostic performance.

If a perforation occurs as a result of a flexible endoscopic procedure, suturing through urgent laparoscopy or open surgery may be required to repair the perforation because suturing is normally stronger than closure using existing endoscopic devices. Suturing with stitches and knots, widely adopted in open or laparoscopic surgery, is still not possible in flexible endoscopy. This is because of the confined space of the natural orifice and target area, high levels of motion dexterity and force needed for stitching and knot-tying, and critical size and strength requirements of wound closure. We present a novel flexible endoscopic robotic suturing system that is able to suture gastrointestinal defects without opening up the patient’s body like in open or laparoscopic surgery. This system features a robotic needle driver and a robotic grasper, both of which are flexible, through-the-scope (small in sizes), and dexterous with five degrees of freedom. The needle driver, facilitated by the grasper, enables the surgeon to control a needle through teleoperation to make stitches and knots in flexible endoscopy. Successful in vivo trials were conducted in the rectum of a live pig to confirm the feasibility of endoscopic suturing and knot-tying using the system in a realistic surgical scenario (not possible with existing devices which are all manually driven). This new technology will change the way how surgeons close gastrointestinal defects. NRF (Natl Research Foundation, S’pore) Accepted version

The efficient electron transfer at the interface of the electron transport layer (ETL) and perovskite, and the transportation along the ETL and collection at the electrode are characteristics that are critically governed by the properties of the ETL. Both characteristics are fundamental for the fabrication of high-performance perovskite solar cells (PSC). So far, mesoporous TiO2 might be considered a versatile ETL as its structure can provide an enormous heterojunction with the perovskite absorber. However, its inert (101) pore facet, massive dead-end pore structure, and extensive grain boundary with the underlying layers in the PSC device have made them unlikely to meet those basic criteria. Here, we have come up with a new design of TiO2 ETL that provides a high carrier dynamic in the device. It is a 3 nm thick high-energy (001) faceted single-crystalline nanograss-like structure (TNG) that is vertically grown on the ITO surface. With these unique structural properties, we achieved an enhanced carrier extraction from the perovskite layer, transportation along the ETL and collection at the electrode. Our results indicate that the charge-transfer dynamic at the TNG–perovskite interface and transport in the device are up to more than 2 times faster than those in the control device or in the recently reported results. With this property, we recorded a power conversion efficiency (PCE) as high as 21.60% (or at the level of 20.84% on average) from the TNG-based PSC with the device area as high as ∼0.17 cm2 using triple cation Cs0.05[MA0.13FA0.87]0.95Pb(I0.87Br0.13)3 perovskites, an improvement as high as 18% from the control device. The TNG with its unique structure and surface chemistry may become a potential platform for the fabrication of a high-performance PSC device via combination with a high-performance perovskite system.

Perceptual image encryption provides an efficient and effective way to preserve the confidentiality of visual information, and the measurement of content leakage is of fundamental importance for perceptually encrypted images. Numerous visual security indexes (VSIs) have been proposed to evaluate visual content leakage. Due to the lack of perceptually encrypted image databases, image quality assessment (IQA) databases are widely adopted to evaluate the performance of existing VSIs. However, there are huge differences between VSIs and IQAs. The misuse of databases may lead to an inaccurate evaluation. In this paper, we propose a perceptually encrypted image database (PEID) which contains 1080 encrypted images from 20 plain images with 10 well-known perceptual encryption techniques. Both visual quality and content leakage scores of the encrypted images are obtained through a comprehensive subjective evaluation. We also propose a systemic methodology to accurately evaluate the monotonicity, fitness, and accuracy of VSIs. We conduct extensive experiments on the proposed PEID to evaluate the performance of existing state-of-the-art VSIs. We have made the database publicly available for download and hope that the proposed PEID can facilitate the research of visual security evaluation and beyond.

Two-dimensional (2D) perovskite materials are a promising platform to construct high performance photodetectors due to their novel structure, high stability, resistance to ion migration and decent light harvesting ability. However, the detection range of 2D perovskite photodetectors is limited due to their large bandgap. Here, self-trapped states with a lower bandgap were created by introducing Sn into a 2D perovskite ((C6H5C2H4NH3)2PbI4, (PEA)2PbI4). Although previously the excitons in the self-trapped states have been regarded as localized excitons, an obvious photocurrent generated by the dissociation of excitons in self-trapped states has been observed. With this property, the detection range of 2D perovskite photodetectors has been successfully extended. Additionally, the special 2D structure of (PEA)2PbI4 endows the device with high stability and great resistance for ion migration to endure long-term (over 1000 s) consistent work without degradation in performance. This facile doping method for extending the detection range of 2D perovskite photodetectors will greatly broaden their application potential.

In patients with compensated advanced chronic liver disease (cACLD), the invasive measurement of hepatic venous pressure gradient is the best predictor of hepatic decompensation. This study aimed at developing an alternative risk prediction model to provide a decompensation risk assessment in cACLD.Patients with cACLD were retrospectively included from 9 international centers within the Portal Hypertension Alliance in China (CHESS) network. Baseline variables from a Japanese cohort of 197 patients with cACLD were examined and fitted a Cox hazard regression model to develop a specific score for predicting hepatic decompensation. The novel score was validated in an external cohort (n = 770) from 5 centers in China, Singapore, Korea, and Egypt, and was further assessed for the ability of predicting clinically significant portal hypertension in a hepatic venous pressure gradient cohort (n = 285).In the derivation cohort, independent predictors of hepatic decompensation were identified including Stiffness of liver, Albumin, Varices, and platElets and fitted to develop the novel score, termed "SAVE" score. This score performed significantly better (all P0.05) than other assessed methods with a time-dependent receiver operating characteristic curve of 0.89 (95% confidence interval [CI]: 0.83-0.94) and 0.83 (95% CI: 0.73-0.92) in the derivation and validation cohorts, respectively. The decompensation risk was best stratified by the cutoff values at -6 and -4.5. The 5-year cumulative incidences of decompensation were 0%, 24.9%, and 69.0% in the low-risk, middle-risk, and high-risk groups, respectively ( P0.001). The SAVE score also accurately predicted clinically significant portal hypertension (AUC, 0.85 95% CI: 0.80-0.90).The SAVE score can be readily incorporated into clinical practice to accurately predict the individual risk of hepatic decompensation in cACLD.

Realization of remote wearable health monitoring (RWHM) technology for the flexible photodiodes is highly desirable in remote-sensing healthcare systems used in space stations, oceans, and forecasting warning, which demands high external quantum efficiency (EQE) and detectivity in NIR region. Traditional inorganic photodetectors (PDs) are mechanically rigid and expensive while the widely reported solution-processed mixed tin-lead (MSP) perovskite photodetectors (PPDs) exhibit a trade-off between EQE and detectivity in the NIR region. Herein, a novel functional passivating antioxidant (FPA) strategy has been introduced for the first time to simultaneously improve crystallization, restrain Sn

The hepatic venous pressure gradient (HVPG) is the gold standard for cirrhotic portal hypertension (PHT), but it is invasive and specialized. Alternative non-invasive techniques are needed to assess the hepatic venous pressure gradient (HVPG). Here, we develop an auto-machine-learning CT radiomics HVPG quantitative model (aHVPG), and then we validate the model in internal and external test datasets by the area under the receiver operating characteristic curves (AUCs) for HVPG stages (≥10, ≥12, ≥16, and ≥20 mm Hg) and compare the model with imaging- and serum-based tools. The final aHVPG model achieves AUCs over 0.80 and outperforms other non-invasive tools for assessing HVPG. The model shows performance improvement in identifying the severity of PHT, which may help non-invasive HVPG primary prophylaxis when transjugular HVPG measurements are not available.

Face recognition (FR) has become increasingly significant in many computer vision applications. However, with the rapid deployment of FR, the privacy of face images has been a growing concern, especially when FR is performed in cloud platforms. Several privacy-preserving face recognition (PPFR) schemes have been proposed to outsource FR without divulging private information. However, most existing schemes are built by employing partially or fully homomorphic cryptosystems, which contain a series of computationally expensive operations and require interactions between users and service providers. In this paper, we propose a novel and efficient scheme to achieve privacy-preserving face recognition in the cloud. Using randomness techniques, we propose an affine transformation, which consists of permutation, diffusion and shift transformations, to protect the privacy of faces. Both projection (feature extraction) and recognition procedures are performed in encrypted domain without interaction. We also propose an optimization technique to increase the efficiency of encryption. Extensive experimental analysis is performed to demonstrate the correctness and effectiveness of the proposed scheme. Theoretical comparison is also conducted to illustrate the efficiency of the proposed scheme.

Accurate haptic feedback is highly challenging for flexible endoscopic surgical robots due to space limitation for sensors on small end-effectors and critical force hysteresis of their tendon-sheath mechanisms (TSMs). This paper proposes a deep learning approach to predicting the distal force of TSMs when manipulating a biological tissue based on only proximal-end measurements. Both Multilayer Perceptron (MLP) and Recurrent Neural Network (RNN) were investigated to study their capabilities of making sequential distal force predictions. The results were compared with those of the conventional modelling approach. It was observed that, when sufficient data was provided for training, RNN achieved the most accurate prediction (RMSE = 0.0219 N) in experiments with constant system velocity. The effects of insufficient training data, varying system velocity and irregular motion trajectories on the performance of RNN were further studied. Notably, RNN could precisely identify the current system phase in the force hysteresis profile and can be applied to TSMs with realistic non-periodic movement such as manual manipulation trajectory (RSME = 0.2287 N). The proposed approach can be applied to any TSM-driven robotic systems for accurate haptic feedback without requiring sensors at the distal ends of the robots. Accepted version

Background & Aims The development of COVID-19 vaccines has progressed with encouraging safety and efficacy data. Concerns have been raised about SARS-CoV-2 vaccine responses in the large population of patients with non-alcoholic fatty liver disease (NAFLD). The study aimed to explore the safety and immunogenicity of COVID-19 vaccination in NAFLD. Methods This multicenter study included patients with NAFLD without a history of SARS-CoV-2 infection. All patients were vaccinated with 2 doses of inactivated vaccine against SARS-CoV-2. The primary safety outcome was the incidence of adverse reactions within 7 days after each injection and overall incidence of adverse reactions within 28 days, and the primary immunogenicity outcome was neutralizing antibody response at least 14 days after the whole-course vaccination. Results A total of 381 patients with pre-existing NAFLD were included from 11 designated centers in China. The median age was 39.0 years (IQR 33.0–48.0 years) and 179 (47.0%) were male. The median BMI was 26.1 kg/m2 (IQR 23.8–28.1 kg/m2). The number of adverse reactions within 7 days after each injection and adverse reactions within 28 days totaled 95 (24.9%) and 112 (29.4%), respectively. The most common adverse reactions were injection site pain in 70 (18.4%), followed by muscle pain in 21 (5.5%), and headache in 20 (5.2%). All adverse reactions were mild and self-limiting, and no grade 3 adverse reactions were recorded. Notably, neutralizing antibodies against SARS-CoV-2 were detected in 364 (95.5%) patients with NAFLD. The median neutralizing antibody titer was 32 (IQR 8-64), and the neutralizing antibody titers were maintained. Conclusions The inactivated COVID-19 vaccine appears to be safe with good immunogenicity in patients with NAFLD. Lay summary The development of vaccines against coronavirus disease 2019 (COVID-19) has progressed rapidly, with encouraging safety and efficacy data. This study now shows that the inactivated COVID-19 vaccine appears to be safe with good immunogenicity in the large population of patients with non-alcoholic fatty liver disease.

Objective To investigate the effect of exosomes derived from human placental mesenchymal stem cells (hPMSC-exs) on lipopolysaccharide (LPS)-induced injury of human pulmonary microvascular endothelial cells (HPMECs) and its possible mechanism. Methods hPMSCs were expanded and cultured in vitro and the cell culture supernatant was collected. The hPMSC-exs in the supernatant was separated and purified by ExoQuick exosomes extraction and purification kit. The morphological characteristics of exosomes were observed by transmission electron microscopy, and the expression of specific markers CD9 and CD63 on the surface of exosomes was detected by Western blotting. A non-contact co-culture system of hPMSCs and HPMECs was constructed. The experiment included a control group, an LPS injury group, an hPMSC group and an hPMSC-exs group. After 12 hours of co-cultivation, the fluorescence intensity of FITC-dextran from the upper chamber into the lower chamber was detected to reflect the permeability of single-layer pulmonary vascular endothelium; DiI-labeled hPMSC-exs was observed under a fluorescence microscope to judge the migration effect of hPMSC-exs on the damaged endothelium; CCK-8 assay was used to detect the proliferation level of the upper chamber cells in each group; JC-1 staining was used to detect the mitochondrial membrane potential of upper ventricular cells in each group; Western blotting was performed to detect the expression of autophagy-related protein LC3 and beclin-1 of HPMECs in the lower chamber. Results The diameter of hPMSC-exs was 30-120 nm, and the expression of CD9 and CD63 was positive, which matched the phenotypic characteristics of hPMSC-exs. Compared with the LPS injury group, the fluorescence intensity of FITC-dextran decreased; LC3II/I ratio and cell proliferation rate increased; beclin-1 protein expression was up-regulated in the hPMSC group and the hPMSC-exs group. Compared with the hPMSC group, the structure of endothelial cells was roughly intact under the fluorescence microscope; the FITC-dextran fluorescence intensity, endothelial cell proliferation rate, mitochondrial membrane potential, expression levels of LC3-II/I and beclin-1 did not change significantly in the hPMSC-exs group. Conclusion hPMSC-exs can alleviate the damage of HPMECs induced by LPS and improves mitochondrial function in the cells. Its mechanism may be related to enhance the autophagy of HPMECs.