Ten-year experiences and outcomes of bypass surgery and endovascular therapy in the management of infrarenal aortic occlusion: a single-center retrospective cohort study.

Background: Infrarenal aortic occlusion (IAO) is a life-threatening condition that often causes lower limb ischemia. Although open surgery is the current recommendation for first-line treatment, recent technological innovations have made endovascular treatment (EVT) a promising alternative. This study aims to compare the clinical outcomes of bypass surgery and EVT in the treatment of IAO. Methods: This study is a single-center retrospective observative study at Peking Union Medical College Hospital. Consecutive 92 patients with chronic and atherosclerotic IAO were treated with either EVT (n=40) or bypass surgery (n=52) between 2011 and 2021. The baseline clinical factors (including demographic data and comorbidities), perioperative data (including Rutherford classification changes, technical success) and complication rates were evaluated. The mid-term patency and overall survival of EVT and bypass were assessed. Follow-up was defined as the time from surgery to the last outpatient visit. Continuous variables and category variables were statistically compared, respectively. Kaplan-Meier survival analyses were conducted for vascular patency. Results: The study found that the demographics and pre-operative Rutherford classification were evenly distributed between the two groups (P>0.05). As for technical success, clinical success, comorbidities, mortality, complication rate, and Rutherford classification after procedures, no significant differences were observed (P>0.05). The average post-procedure hospital stay was 5.15 days in the EVT group and was significantly shorter than that of the bypass group, which was 11.83 days (P<0.0001). As for short-term and long-term results, the 1-, 3-, and 5-year primary patency rates were 81.8%, 73.1%, and 73.1% in the EVT group and 97.8%, 80.6%, and 80.6% in the bypass group. The bypass group had significantly better primary patency (P=0.034). There was a significant difference in the secondary patency rate (Bypass 100% vs. EVT 81.6%; P=0.005). Moreover, survival rates were higher in the bypass surgery group than in the EVT group (P=0.035). Conclusions: Although EVT's primary patency rate was lower than that with the bypass surgery, its safety and efficacy were still comparable to anatomic bypass surgery for IAO with less severe perioperative complications and shorter hospital stay. Therefore, EVT could be a feasible option for IAO.

A contrast-enhanced CT-based radiomic nomogram for the differential diagnosis of intravenous leiomyomatosis and uterine leiomyoma.

Objective: Uterine intravenous leiomyomatosis (IVL) is a rare and unique leiomyoma that is difficult to surgery due to its ability to extend into intra- and extra-uterine vasculature. And it is difficult to differentiate from uterine leiomyoma (LM) by conventional CT scanning, which results in a large number of missed diagnoses. This study aimed to evaluate the utility of a contrast-enhanced CT-based radiomic nomogram for preoperative differentiation of IVL and LM. Methods: 124 patients (37 IVL and 87 LM) were retrospectively enrolled in the study. Radiomic features were extracted from contrast-enhanced CT before surgery. Clinical, radiomic, and combined models were developed using LightGBM (Light Gradient Boosting Machine) algorithm to differentiate IVL and LM. The clinical and radiomic signatures were integrated into a nomogram. The diagnostic performance of the models was evaluated using the area under the curve (AUC) and decision curve analysis (DCA). Results: Clinical factors, such as symptoms, menopausal status, age, and selected imaging features, were found to have significant correlations with the differential diagnosis of IVL and LM. A total of 108 radiomic features were extracted from contrast-enhanced CT images and selected for analysis. 29 radiomics features were selected to establish the Rad-score. A clinical model was developed to discriminate IVL and LM (AUC=0.826). Radiomic models were used to effectively differentiate IVL and LM (AUC=0.980). This radiological nomogram combined the Rad-score with independent clinical factors showed better differentiation efficiency than the clinical model (AUC=0.985, p=0.046). Conclusion: This study provides evidence for the utility of a radiomic nomogram integrating clinical and radiomic signatures for differentiating IVL and LM with improved diagnostic accuracy. The nomogram may be useful in clinical decision-making and provide recommendations for clinical treatment.

Renal perfusion improvement in the perioperative period after unilateral endovascular revascularization in patients with atherosclerotic renal artery stenosis.

Background: The clinical benefits of endovascular treatment in renal artery stenosis (RAS) remain controversial. This study used an intraoperative renal perfusion imaging technique, called flat-panel detector parenchymal blood volume imaging (FD-PBV), to observe the change in renal perfusion after endovascular treatment in RAS. Materials and methods: In a prospective, single-center study, we assigned 30 patients with atherosclerotic RAS who underwent endovascular treatment between March 2016 and March 2021. The preoperative and postoperative results of renal perfusion, blood pressure, and renal function, were compared. Results: Both median kidney volume (p < 0.001) and median preoperative mean density of contrast medium (MDCM) (p = 0.028) increased significantly after endovascular treatment. The ratio of postoperative and preoperative MDCM differed greatly among the patients. For patients with preoperative MDCM <304.0 HU (Subgroup A, 15 cases), MDCM significantly increased after treatment (p = 0.001) and 12 (80.0%) patients had more than 10% increase in renal perfusion. For patients who had relatively high preoperative renal perfusion (MDCM ≥304.0 HU, Subgroup B, 15 cases), preoperative and postoperative MDCM were similar (p = 0.776). On the other hand, the serum creatinine levels significantly decreased in Subgroup A (p = 0.033) and fewer antihypertensive drugs were used after endovascular revascularization (p = 0.041). The preoperative and postoperative creatinine levels and number of antihypertensive drugs were similar in Subgroup B. Conclusions: During the perioperative period, RAS patients with relatively low preoperative renal perfusion levels had greater improvement in renal perfusion, renal function, and blood pressure control after endovascular treatment. The improvement of renal function needs to be confirmed by long-term follow-up.

Midterm Results of a Surgeon-Modified Device to Preserve the Flow of the Internal Iliac Artery During Endovascular Repair of Aneurysm: Single-Center Experiences.

BACKGROUND: During endovascular aneurysm repair (EVAR), commercial iliac branch devices (IBDs) have become an inescapable alternative for preserving antegrade internal iliac artery (IIA) blood flow. Due to the different morphological features of aneurysms, commercial IBDs may not be suitable for all patients. Reported experience with the implantation of the new surgeon-modified IBD (sm IBD) is limited. This investigation describes the indications, efficacy, and safety of the sm IBD. METHODS: Data from consecutive elective implantations of IBDs in patients between March 2011 and May 2021 in a single center were incorporated. The sm IBDs were indicated in patients with common iliac artery aneurysms (CIAAs) and with a challenging anatomy and in those patients with or without abdominal aortic aneurysm (AAA). RESULTS: Fifteen patients (15 male, mean age 67.6 ± 7.9 years) were included. Fifteen sm IBDs were implanted in 1 procedure (100%). Fourteen (93.3%) patients had simultaneous endovascular aneurysm repair (EVAR) and 1 (6.7%) patient previously had a bilateral CIAAs repair by EVAR. The mean common iliac artery (CIA) diameter was 36.6 ± 12.5 mm. Technical success was obtained in all patients (100%). The median operation time was 189.7 ± 78.6 min, with a median fluoroscopy time of 45.3 ± 15.9 min. Axillary artery access was used in 11 (73.3%) procedures. The mean total hospital stay was 5.6 ± 2.8 days, and the postoperative follow-up was 35.4 months (range 2-120). The estimated IIA bridge stent patency at 1 year after operation was 100% and 85.7% ± 13.2% 5 years postoperatively. One (6.7%) IIA branch was occluded, and this patient remained asymptomatic. One patient (6.7%) needed reintervention, and another (6.7%) patient had type II leakage, which is currently under close surveillance. CONCLUSIONS: Using an IBD to maintain the pelvic blood flow is an effective and feasible intravascular technique, especially for patients with an abnormal iliac artery anatomy. This novel technique has similar midterm procedural success rate compared to the use of commercial IBDs. Therefore, these devices are more suitable for patients with certain anatomic challenges and can be used as an alternative treatment.

Angiosome-directed endovascular intervention and infrapopliteal disease: Intraoperative evaluation of distal hemodynamic changes and foot blood volume of lower extremity.

Objectives: To evaluate foot blood volume and hemodynamics and explore whether quantitative techniques can guide revascularization. Materials and methods: A prospective single-center cohort study included thirty-three patients with infrapopliteal artery occlusion who underwent percutaneous transluminal angioplasty (PTA) between November 2016 and May 2020. The time-to-peak (TTP) from color-coded quantitative digital subtraction angiography (CCQ-DSA) and parenchymal blood volume (PBV) were used to evaluate the blood volume and hemodynamic changes in different regions of the foot before and after the operation. Results: After the intervention procedure, the overall blood volume significantly increased from 25.15 ± 21.1 ml/1,000 ml to 72.33 ± 29.3 ml/1,000 ml (p < 0.001, with an average increase of 47.18 ml/1,000 ml. The overall TTP decrease rate, postoperative blood flow time significantly faster than those preoperatively, from 22.93 ± 7.83 to 14.85 ± 5.9 s (p < 0.001, with an average decrease of 8.08 s). Direct revascularization (DR) resulted in significant blood volume improvement than compared with indirect revascularization (IR) [188% (28, 320) vs.51% (10, 110), p = 0.029]. Patients with DR had a significantly faster blood flow time than those with IR [80% (12, 180) vs. 26% (5, 80), p = 0.032]. The ankle-brachial index (ABI) of the affected extremity also showed an significant change from 0.49 ± 0.3 to 0.63 ± 0.24 (p < 0.001) after the intervention. The relative values of ΔTTP and ΔABI showed a weak correlation (r = -0.330). Conclusions: The quantitative measurement results based on PBV and CCQ-DSA techniques showed that the overall blood volume increased significantly and that the foot distal hemodynamics were significantly improved after endovascular treatment. DR in the ischemic area could r improve foot perfusion.

Using closed-circuit television cameras to analyze traffic safety at intersections based on vehicle key points detection.

In the within-intersection area, vehicles from different approaches make turning movements resulting in many conflict points. Hence, drivers are more prone to make mistakes in that area, which leads to severe crash outcomes. In the current roadway system, the Closed-Circuit Television (CCTV) cameras could be a cost-effective sensor to monitor the safety condition in the within-intersection area. This study proposed a framework named "Near Miss Event Detection System (NMEDS)" for road safety diagnostics using video data collected from CCTV cameras. The proposed framework combined the Mask-RCNN bounding box detection and Occlusion-Net detection algorithm to reconstruct vehicles' key points in a 3D view. Vehicles' key points including right-front headlight, left-front headlight, right-back taillight, and left-back taillight could be identified and transformed into a 2D bird's-eye view (i.e., real-world coordinate system) for safety analysis. A method was proposed to modify the occluded key points, which could not be observed by cameras due the turning movements in the within-intersection area. The post-encroachment time (PET) was calculated by using the trajectory data in the 2D view. The proposed framework was compared with two counterparts (i.e., bounding box detection only and key point detection only) by conducting an empirical study at a 4-leg intersection. The results suggested that the proposed framework could obtain more accurate vehicle trajectory and better autocorrelation analytics was conducted to identify the significantly dangerous locations in the within-intersection area. It is expected that the proposed methods could help diagnose road safety problems using CCTV cameras. Moreover, the proposed method could be incorporated with Connected Vehicle Systems and provide information to nearby drivers based on Infrastructure-to-Vehicle (I2V) technologies.

Effect of Endovascular Treatment on Urinary Obstruction Caused by Iliac Artery Aneurysm: A Case Series and Review of the Literature.

BACKGROUND: Endovascular aneurysm repair is superior to open surgical repair (OSR) in multiple aspects and is the mainstay for noninflammatory iliac artery aneurysm (IAA) treatment. However, for noninflammatory IAAs with neighboring structures compressed, the experience of endovascular treatment is limited. This case series study aimed to describe the experience of endovascular aneurysm repair of noninflammatory IAAs with urinary obstruction and to review the limited reports in the literature. METHODS: From December 2012 to July 2020, we reported 7 cases of noninflammatory IAAs combined with urinary obstruction that was treated successfully with endovascular treatment. The literature on noninflammatory IAAs combined with urinary obstruction was reviewed using the online databases PubMed, Embase, and the Cochrane Library. RESULTS: Thirty nine articles were found in the literature. OSR accounted for 80% of the treatment options in previous publications, while only 20% of patients accepted endovascular treatment. Hydronephrosis remission was observed in all patients treated by OSR, while the endovascular treatment group lacked long-term follow-up. In our retrospective study containing 7 cases, no complications, reintervention, or mortality was observed, all stents were patent, and all symptoms were relieved after the operation. Diameter reduction of the aneurysms was observed in 6/7 cases and the average aneurysm reduction was -24.1 ± 15.8 mm (P = 0.0097). Urinary obstruction faded in 5/7 cases where aneurysm shrinking was found simultaneously. CONCLUSIONS: This case series and literature review on noninflammatory IAAs combined with urinary obstruction suggests that the diameter of the responsible aneurysm may be reduced after endovascular treatment, which may also alleviate symptoms of urinary obstruction caused by noninflammatory IAAs.

Impaired pedal arch affects the treatment effect in patients with single tibial artery revascularization demonstrated by intraoperative perfusion.

Background: The treatment strategy for patients with multiple infrapopliteal artery occlusions remains controversial. In this study, we investigated how anatomic factors affect the treatment effect of infrapopliteal artery intervention and identified suitable intervention strategies for patients with multiple infrapopliteal artery occlusions. Methods: This was a prospective, single-center, observational cohort study. For each patient, the intrainterventional blood volume improvement of the dorsum and plantar surface was measured and classified into the direct perfused region (DR) or indirect perfused region (IR) on the basis of whether the supplying artery was revascularized. Digital subtraction angiography was performed to analyze how pedal arch patency affects blood communication between DR and IR. Results: A total of 38 patients treated with infrapopliteal intervention at the Department of Vascular Surgery of Peking Union Medical College Hospital from November 2016 to November 2020 were considered for inclusion in this study. Finally, 26 patients were included in the analysis. In patients with type III pedal arch, blood volume improvements for DR and IR were 70.50 (17.50, 191.75) and 11.25 (-10.25, 50.25) mL/1,000 mL, respectively (P=0.018). No significant difference was found between DR and IR in patients with type I pedal arch (P=0.208) and type II pedal arch (P=0.110). Conclusions: Impaired pedal arch has an adverse impact on foot collateral circulation. Patients with these conditions are more suitable for direct revascularization than other patients. Trial Registration: ClinicalTrials.gov identifier: NCT03248323.

Analysis and prediction of intersection traffic violations using automated enforcement system data.

The automated enforcement system (AES) is an effective way of supplementing traditional traffic enforcement, and the traffic violation data from AES can also be effectively used for safety research. In this study, traffic violation data were used to analyze the influencing factors associated with traffic violations and to predict the probability of violations at intersections. The potential factors influencing violations include 24 independent factors related to time, space, traffic and weather. Results from a logistic model showed that the midday period, weekends, residential districts, collector roads, congested traffic conditions, high traffic flow, lower wind speed and low temperature would increase the probability of traffic violations. The probability of violations was predicted by the random forest algorithm, which was proven to be the best traffic violation prediction model among logistic regression, Gaussian naive Bayes, and support vector machine. Moreover, the proximity weighted synthetic oversampling technique (ProWSyn) method was applied to reduce the impact of the imbalance ratio (IR) and improve the model's prediction performance. The receiver operating characteristics (ROC) curves and Precision-Recall (PR) curves illustrated that the random forest algorithm using oversampling data had the best classifier prediction performance than undersampling data. The area under curve (AUC) and out-of-bag (OOB) error with IR = 1 reached 0.914 and 0.0787, which showed the better performance of the random forest algorithm using ProWSyn in dealing with imbalanced traffic violation data.

One-year outcomes of drug-coated balloon treatment for long femoropopliteal lesions: a multicentre cohort and real-world study.

BACKGROUND: Drug-coated balloons (DCBs) have shown superiority in the endovascular treatment of short femoropopliteal artery disease. Few studies have focused on outcomes in long lesions. This study aimed to evaluate the safety and effectiveness of Orchid® DCBs in long lesions over 1 year of follow-up. METHODS: This study is a multicentre cohort and real-world study. The patients had lesions longer than or equal to 150 mm of the femoropopliteal artery and were revascularized with DCBs. The primary endpoints were primary patency, freedom from clinically driven target lesion revascularization (TLR) at 12 months and major adverse events (all-cause death and major target limb amputation). The secondary endpoints were the changes in Rutherford classification and the ankle brachial index (ABI). RESULTS: One hundred fifteen lesions in 109 patients (mean age 67 ± 11 years, male proportion 71.6%) were included in this study. The mean lesion length was 252.3 ± 55.4 mm, and 78.3% of the lesions were chronic total occlusion (CTO). Primary patency by Kaplan-Meier estimation was 98.1% at 6 months and 82.1% at 12 months. The rate of freedom from TLR by Kaplan-Meier estimation was 88.4% through 12 months. There were no procedure- or device-related deaths through 12 months. The rate of all-cause death was 2.8%. Cox regression analysis suggested that renal failure and critical limb ischaemia (CLI) were statistically significant predictors of the primary patency endpoint. CONCLUSION: In our real-world study, DCBs were safe and effective when used in long femoropopliteal lesions, and the primary patency rate at 12 months by Kaplan-Meier estimation was 82.1%.

Meso-level hotspot identification for suburban arterials.

Accurate identification of crash hotspots forms the foundation of roadway safety improvement. The Highway Safety Manual micro-level approach uses individual intersections and road segments as analysis units, and correspondingly identifies some isolated road entities as hotspots. However, because traffic police and administrative agencies routinely conduct safety improvement based on multiple continuous segments and intersections, the identification of hotspots at the micro-level is inefficient for field application. To better meet this need, this study proposes a new meso-level approach to identify hotspots, specifically on suburban arterials. Meso-level analysis units of three different configurations (201, 150, and 100 units) were obtained by combining a set number of intersections and their adjacent segments according to crash distribution and homogeneity. Their influence areas were determined according to the proportion of urbanized land in areas perpendicularly adjacent to the arterials. Three Bayesian Poisson-lognormal conditional autoregressive models (PLN-CAR) considering spatial correlations were developed for each unit configuration, using the full Bayesian (FB) method to ameliorate random fluctuation in crash counts. Potential for safety improvement (PSI) values were calculated based on the modeling results and were used to identify hotspots. Two measures, i.e., the concentrated degree of hotspots (CDH) and the hotspot identification accuracy (HIA), were proposed to make a quantitative and comparative evaluation. Results showed that 1) arterials with more parallel roads suffer lower crash risk, and 2) considering both the hotspot distribution and the identification accuracy, the 150 meso-level unit configuration was the best. The proposed meso-level hotspot identification method promises to be adaptive to safety improvement practices on suburban arterials.

Developing safety performance functions for freeways at different aggregation levels using multi-state microscopic traffic detector data.

Safety Performance Functions (SPFs) have been widely used by researchers and practitioners to conduct roadway safety evaluation. Traditional SPFs are usually developed by using annual average daily traffic (AADT) along with geometric characteristics. However, the high level of aggregation may lead to a failure to capture the temporal variation in traffic characteristics (e.g., traffic volume and speed) and crash frequencies. In this study, SPFs at different aggregation levels were developed based on microscopic traffic detector data from California, Florida, and Virginia. More specifically, five aggregation levels were considered: (1) annual average weekday hourly traffic (AAWDHT), (2) annual average weekend hourly traffic (AAWEHT), (3) annual average weekday peak/off-peak traffic (AAWDPT), (4) annual average day of the week traffic (AADOWT), and (5) annual average daily traffic (AADT). Model estimation results showed that the segment length and volume, as exposure variables, are significant across all the aggregation levels. Average speed is significant with a negative coefficient, and the standard deviation of speed was found to be positively associated with the crash frequency. It is noteworthy that the operation of the high occupancy vehicle (HOV) lanes was found to have a positive effect on crash frequency across all the aggregation levels. The model results also showed that the AAWDPT and AADOWT models consistently performed better (the improvements range from 3.14%-16.20%) than the AADT-based SPF, which implies that the differences between the day of the week and peak/off-peak periods should be considered in the development of crash prediction models. The model transferability results indicated that the SPFs between Florida and Virginia are transferrable, while the models between California and the other two states are not transferrable.

Infrapopliteal endovascular intervention and the angiosome concept: intraoperative real-time assessment of foot regions' blood volume guides and improves direct revascularization.

OBJECTIVE: There is no consensus for determining which vessel should be revascularized in patients with multiple diseased infrapopliteal arteries. The angiosome concept may guide a more efficient targeted direct revascularization. Therefore, we conducted a study to assess whether the regional evaluation of foot blood volume may guide direct revascularization (DR) and if it will lead to better perfusion improvement than indirect revascularization (IR). METHODS: We performed a prospective single-center observational cohort study in patients treated in the Department of Vascular Surgery of Peking Union Medical College Hospital from November 2016 to April 2019. Twenty-seven patients treated with endovascular intervention were included. The intraoperative parenchymal blood volume of different foot regions was obtained for each patient using C-arm CT before and after intervention. RESULTS: The intervention procedure significantly increased the overall blood volume (48.95 versus 81.97 ml/1000 ml, p = 0.002). Patients with direct revascularization had a 197% blood volume increase while patients with indirect revascularization had a 39% increase (p = 0.028). The preoperative blood volume was higher in patients with mild symptoms than in patients with severe symptoms (58.20 versus 30.45 ml/1000 ml, p = 0.039). However, in regard to postoperative blood volume, no significant difference was discovered between these two groups (75.05 versus 95.01 ml/1000 ml, p = 0.275). CONCLUSION: Based on quantitative measurements, we conclude that overall blood volume can rise significantly after the intervention. Revascularizing the supplying vessel of the ischemic area directly will result in better perfusion improvement than restoring blood supply through the collateral circulation. Preoperative blood volume is associated with preoperative symptoms. KEY POINTS: • Flat panel detector CT can obtain intraoperative perfusion status and guide treatment in endovascular intervention. • Revascularizing the supplying vessel of the ischemic area directly will result in better perfusion improvement than restoring the blood supply through the collateral circulation. • Patients with severer clinical manifestations have lower blood volumes.

Using bus critical driving events as surrogate safety measures for pedestrian and bicycle crashes based on GPS trajectory data.

Pedestrian and bicycle safety is a key component in traffic safety studies. Various studies were conducted to address pedestrian and bicycle safety issues for intersections, road segments, etc. However, only a few studies investigated pedestrian and bicycle safety for bus stops, which usually have a relatively larger volume of pedestrians and bicyclists. Moreover, traditional reactive safety approaches require a significant number of historical crashes, while pedestrian and bicycle crashes are usually rare events. Alternatively, surrogate safety measures could proactively evaluate traffic safety status when crash data are rare or unavailable. This paper utilized critical bus driving events extracted from GPS trajectory data as pedestrian and bicycle surrogate safety measures for bus stops. A city-wide trajectory data from Orlando, Florida was used, which contains around 300 buses, 6,700,000 GPS records, and 1300 bus stops. Three critical driving events were identified based on the buses' acceleration rates and stop time; hard acceleration, hard deceleration, and long stop. The relationships between critical driving events and crashes were examined using Spearman's rank correlation coefficient. All three events were positively correlated with pedestrian and bicycle crashes. Long stop event has the highest correlation coefficient, followed by hard acceleration and hard deceleration. A Bayesian negative binomial model incorporating spatial correlation (Bayesian NB-CAR) was built to estimate the pedestrian and bicycle crash frequency using the generated events. The results were consistent with the correlation estimation. For example, hard acceleration and long stop events were both positively related to pedestrian and bicycle crashes. Moreover, model evaluation results indicated that the proposed Bayesian NB-CAR outperformed the standard Bayesian negative binomial model with lower Watanabe-Akaike Information Criterion (WAIC) and Deviance Information Criteria (DIC) values. In conclusion, this paper suggests the use of critical bus driving events as surrogate safety measures for pedestrian and bicycle crashes, which could be implemented in a proactive traffic safety management system.

Crash data augmentation using variational autoencoder.

In this paper, we present a data augmentation technique to reproduce crash data. The dataset comprising crash and non-crash events are extremely imbalanced. For instance, the dataset used in this paper consists of only 625 crash events for over 6.5 million non-crash events. Thus, learning algorithms tend to perform poorly on these datasets. We have used variational autoencoder to encode all the events into a latent space. After training, the model could successfully separate crash and non-crash events. To generate data, we sampled from the latent space containing crash data. The generated data was compared with the real data from different statistical aspects. t-Test, Levene-test and Kolmogrove Smirnov test showed that the generated data was statistically similar to the real data. It was also compared to some of the minority oversampling techniques like SMOTE and ADASYN as well as the GAN framework for generating data. Crash prediction models based on Logistic Regression (LR), Support Vector Machine (SVM) and Artificial Neural Network (ANN) were used to compare the generated data from the different oversampling techniques. Overall, variational autoencoder (VAE) showed excellent results compared to the other data augmentation methods. Specificity is improved by 8% and 4% for VAE-LR and VAE-SVM respectively when compared to SMOTE while the sensitivity is improved by 6% and 5% when compared to ADASYN. Moreover, VAE generated data also helps to overcome the overfitting problem in SMOTE and ADASYN since there is flexibility in choosing the decision boundary.

Multi-Objective reinforcement learning approach for improving safety at intersections with adaptive traffic signal control.

Adaptive traffic signal control (ATSC) systems improve traffic efficiency, but their impacts on traffic safety vary among different implementations. To improve the traffic safety pro-actively, this study proposes a safety-oriented ATSC algorithm to optimize traffic efficiency and safety simultaneously. A multi-objective deep reinforcement learning framework is utilized as the backend algorithm. The proposed algorithm was trained and evaluated on a simulated isolated intersection built based on real-world traffic data. A real-time crash prediction model was calibrated to provide the safety measure. The performance of the algorithm was evaluated by the real-world signal timing provided by the local jurisdiction. The results showed that the algorithm improves both traffic efficiency and safety compared with the benchmark. A control policy analysis of the proposed ATSC revealed that the abstracted control rules could help the traditional signal controllers to improve traffic safety, which might be beneficial if the infrastructure is not ready to adopt ATSCs. A hybrid controller is also proposed to provide further traffic safety improvement if necessary. To the best of the authors' knowledge, the proposed algorithm is the first successful attempt in developing adaptive traffic signal system optimizing traffic safety.

In-depth approach for identifying crash causation patterns and its implications for pedestrian crash prevention.

INTRODUCTION: A pedestrian crash occurs due to a series of contributing factors taking effect in an antecedent-consequent order. One specific type of antecedent-consequent order is called a crash causation pattern. Understanding crash causation patterns is important for clarifying the complicated growth of a pedestrian crash, which ultimately helps recommend corresponding countermeasures. However, previous studies lack an in-depth investigation of pedestrian crash cases, and are insufficient to propose a representative picture of causation patterns. METHOD: In this study, pedestrian crash causation patterns were discerned by using the Driving Reliability and Error Analysis Method (DREAM). One hundred and forty-two pedestrian crashes were investigated, and five pedestrian pre-crash scenarios were extracted. Then, the crash causation patterns in each pre-crash scenario were analyzed; and finally, six distinct patterns were identified. Accordingly, 17 typical situations corresponding to these causation patterns were specified as well. RESULTS: Among these patterns, the pattern related to distracted driving and the pattern related to an unexpected change of pedestrian trajectory contributed to a large portion of the total crashes (i.e., 27% and 24%, respectively). Other patterns also played an important role in inducing a pedestrian crash; these patterns include the pattern related to an obstructed line of sight caused by outside objects (9%), the pattern that involves reduced visibility (13%), and the pattern related to an improper estimation of the gap distance between the vehicle and the pedestrian (10%). The results further demonstrated the inter-heterogeneity of a crash causation pattern, as well as the intra-heterogeneity of pattern features between different pedestrian pre-crash scenarios. Conclusions and practical applications: Essentially, a crash causation pattern might involve different contributing factors by nature or dependent on specific scenarios. Finally, this study proposed suggestions for roadway facility design, roadway safety education and pedestrian crash prevention system development.

Analyzing traffic violation behavior at urban intersections: A spatio-temporal kernel density estimation approach using automated enforcement system data.

The Automated Enforcement System (AES) has become the most important traffic enforcement system in China. In this study, a spatio-temporal kernel density estimation (STKDE) model, integrating spatio-temporal statistics and three-dimensional visualization techniques, was applied to reveal the spatial and temporal patterns of traffic violation behavior at urban intersections. The multivariate Gaussian kernel function was selected for space and time density estimation, as it has been shown to be a good arbitrary probability density function for continuous multivariate data. Because the STKDE model builds a space-time cube that adopts different colors of voxels to visualize the density of traffic violations, an optimal bandwidth selector that combines unconstrained pilot bandwidth matrices with a data-driven method was selected for achieving the best visualization result. The raw AES traffic violation data over 200 weekdays from 69 intersections in the city of Wujiang were empirically analyzed. The results show that the STKDE space-time cube made it easier to detect the spatio-temporal patterns of traffic violations than did the traditional hotspots map. An interesting finding was that traffic sign violations and traffic marking violations were primarily concentrated not in regular peak hours, but during the time period of 14:00-16:00, which indicates that these intersections were the most congested during this period. Primarily, the STKDE model identified seven patterns of spatio-temporal traffic violation hotspots and coldspots. These results are important because their prediction of temporal trends of traffic violations may help contribute toward the understanding and improvement of intersection safety problems.

Time-varying Analysis of Traffic Conflicts at the Upstream Approach of Toll Plaza.

This study investigates the traffic conflict risks at the upstream approach of toll plaza during the vehicles' diverging period from the time of arrival at the diverging area to that of entering the tollbooths. Based on the vehicle's trajectory data extracted from unmanned aerial vehicle (UAV) videos using an automated video analysis system, vehicles' collision risk is computed by extended time to collision (TTC). Then, two time-varying mixed logit models including time-varying random effects logistic regression (T-RELR) model and time-varying random parameters logistic regression (T-RPLR) are developed to examine the time varying effects of influencing factors on vehicle collision risk, and four models including the standard random effects logistic regression (S-RELR) model, standard random parameters logistic regression (S-RPLR) model, distance-varying random effects logistic regression (D-RELR) model and distance-varying random parameters logistic regression (D-RPLR) are developed for model performance comparison. The results indicate that the T-RPLR model has the highest prediction accuracy. Eight influencing factors including following vehicle's travel distance, following vehicle's initial lane, following vehicle's toll collection type, leading vehicle's toll collection type, distance between two vehicles' centroids, and following vehicle's speed, are found to have time-varying effects on collision risk. Meanwhile, the first six factors are found to exhibit heterogeneous effects over the travel time. Another important finding is that the vehicle that comes from the innermost lane has an increasing trend to be involved in traffic conflicts, whereas the collision risks of other vehicles decrease as the travel time increases. Moreover, vehicles with higher speed have a decreasing probability to be involved in crashes over the travel time. Interestingly, the results of D-RPLR model are similar with that of T-RPLR model. These findings provide helpful information for accurate assessment of collision risk, which is a key step toward improving safety performance of the toll plazas' diverging areas.

Predicting real-time traffic conflicts using deep learning.

Recently, technologies for predicting traffic conflicts in real-time have been gaining momentum due to their proactive nature of application and the growing implementation of ADAS technology in intelligent vehicles. In ADAS, machine learning classifiers are utilised to predict potential traffic conflicts by analysing data from in-vehicle sensors. In most cases, a condition is classified as a traffic conflict when a safety surrogate (e.g. time-to-collision, TTC) crosses a pre-defined threshold. This approach, however, largely ignores other factors that influence traffic conflicts such as speed variance, traffic density, speed and weather conditions. Considering all these factors in detecting traffic conflicts is rather complex as it requires an integration and mining of heterodox data, the unavailability of traffic conflicts and conflict prediction models capable of extracting meaningful and accurate information in a timely manner. In addition, the model has to effectively handle large imbalanced data. To overcome these limitations, this paper presents a centralised digital architecture and employs a Deep Learning methodology to predict traffic conflicts. Highly disaggregated traffic data and in-vehicle sensors data from an instrumented vehicle are collected from a section of the UK M1 motorway to build the model. Traffic conflicts are identified by a Regional-Convolution Neural Network (R-CNN) model which detects lane markings and tracks vehicles from images captured by a single front-facing camera. This data is then integrated with traffic variables and calculated safety surrogate measures (SSMs) via a centralised digital architecture to develop a series of Deep Neural Network (DNN) models to predict these traffic conflicts. The results indicate that TTC, as expected, varies by speed, weather and traffic density and the best DNN model provides an accuracy of 94% making it reliable to employ in ADAS technology as proactive safety management strategies. Furthermore, by exchanging this traffic conflict awareness data, connected vehicles (CVs) can mitigate the risk of traffic collisions.