Spatiotemporal CNN with Pyramid Bottleneck Blocks: Application to eye blinking detection.

Eye blink detection is a challenging problem that many researchers are working on because it has the potential to solve many facial analysis tasks, such as face anti-spoofing, driver drowsiness detection, and some health disorders. There have been few attempts to detect blinking in the wild scenario, while most of the work has been done under controlled conditions. Moreover, current learning approaches are designed to process sequences that contain only a single blink ignoring the case of the presence of multiple eye blinks. In this work, we propose a fast framework for eye blink detection and eye blink verification that can effectively extract multiple blinks from image sequences considering several challenges such as lighting changes, variety of poses, and change in appearance. The proposed framework employs fast landmarks detector to extract multiple facial key points including the ones that identify the eye regions. Then, an SVD-based method is proposed to extract the potential eye blinks in a moving time window that is updated with new images every second. Finally, the detected blink candidates are verified using a 2D Pyramidal Bottleneck Block Network (PBBN). We also propose an alternative approach that uses a sequence of frames instead of an image as input and employs a continuous 3D PBBN that follows most of the state-of-the-art approaches schemes. Experimental results show the better performance of the proposed approach compared to the state-of-the-art approaches.

The rate of success of the conservative management of liver trauma in a developing country.

BACKGROUND: The conservative treatment of liver trauma has made important progress over the last 10 years at the Trauma University Hospital in Tirana, Albania. The percentage of success was 58.7%. The aims of this study were to analyze the conservative treatment of liver trauma and to compare the results with those in the literature. METHODS: This study was conducted prospectively from January 2009 to December 2012. We analyzed 173 patients admitted to our hospital with liver trauma. Liver injuries were evaluated according to the American Association for the Surgery of Trauma and the World Society of Emergency Surgery classification, while the anatomic gravity of the associated injuries was defined using the Injury Severity Score system. The potential mortality was estimated with the Revised Trauma Score. RESULTS: Out of the 173 patients with liver trauma, 83.2% were male. The main cause of liver trauma was motor vehicle crashes (50.9%). Blunt trauma was the cause of liver injury in 129 cases (74.6%), and penetrating trauma occurred in 44 cases (25.4%). Initially, the decision was to manage 88 cases (50.9%) via the conservative approach. Of these, 73 cases (42.2%) were successfully treated with conservative treatment, while in 15 cases (17.2%), this approach failed. The success rate of conservative treatment by grade of injuries was as follows: grade I (38.4%), grade II (30.1%), grade III (28.8%), and grade IV (2.7%). The likelihood of the success of conservative treatment had a significant correlation with the grade of the liver injury (p < 0.00001), associated intra-abdominal injuries (p = 0.00051), and complications (z = 2.3169, p = 0.02051). The overall mortality rate of liver trauma was 13.2%. CONCLUSIONS: The likelihood of success in using conservative treatment had a significant correlation with the grade of liver injury and associated intra-abdominal injuries. The limited hospital resources and low level of consensus on conservative treatment had a negative impact on the level of success.

The influence of the risk factor on the abdominal complications in colon injury management.

INTRODUCTION: The management of colon injuries has distinctly evolved over the last three decades. However, trauma surgeons often find themselves in a dilemma, whether to perform a diversion or to perform a primary repair. The purpose of this study is to evaluate risk factors in colon injury management and their influence on abdominal complications. PATIENTS AND METHODS: This is a prospective study conducted at a national level I trauma center in Tirana, Albania from January 2009 to December 2012. The data with respect to demographics, physiological risk factors, intraoperative findings, and surgical procedures were collected. Colonic injury-related morbidity and mortality were analyzed. Multivariate logistic regression analysis was performed by assessing the influence of risk factors on abdominal complications. RESULTS: Of the 157 patients treated with colon injury, was performed a primary repair in 107 (68.15%) of the patients and a diversion in the remaining 50 (31.85%). The mean PATI was 18.6, while 37 (23.6%) of patients had PATI greater than 25. The complications and their frequencies according to the surgical technique used (primay repair vs diversion respectively) includes: wound infections (9.3% vs 50%), anastomotic leak (1.8% vs 8.7%), and intra-abdominal abscess (1.8% vs 6.5%). The multivariate analysis identified two independent risk factors for abdominal complications: transfusions of 4 units of blood within the first 24 hours (OR = 1.2 95% CI (1.03 - 1.57) p =0.02), and diversion (OR = 9.6, 95% CI 4.4 - 21.3, p<0.001). CONCLUSION: Blood transfusions of more than 4 units within the first 24 hours and diversion during the management of destructive colon injuries are both independent risk factors for abdominal complications. The socioeconomic impact and the need for a subsequent operation in colostomy patients are strong reasons to consider primary repair in the management of colon injuries.