Adversarial example defense based on image reconstruction

The rapid development of deep neural networks (DNN) has promoted the widespread application of image recognition, natural language processing, and autonomous driving. However, DNN is vulnerable to adversarial examples, such as an input sample with imperceptible perturbation which can easily invalidate the DNN and even deliberately modify the classification results. Therefore, this article proposes a preprocessing defense framework based on image compression reconstruction to achieve adversarial example defense. Firstly, the defense framework performs pixel depth compression on the input image based on the sensitivity of the adversarial example to eliminate adversarial perturbations. Secondly, we use the super-resolution image reconstruction network to restore the image quality and then map the adversarial example to the clean image. Therefore, there is no need to modify the network structure of the classifier model, and it can be easily combined with other defense methods. Finally, we evaluate the algorithm with MNIST, Fashion-MNIST, and CIFAR-10 datasets; the experimental results show that our approach outperforms current techniques in the task of defending against adversarial example attacks.

Efficacy of a Simplified Feedback Trainer for High-Quality Chest Compression Training: A Randomized Controlled Simulation Study

Background: The most recent international guidelines recommended support training of chest compression (CC) using feedback devices. This study aimed to compare the training efficacy of a simplified feedback trainer with the traditional cardiopulmonary resuscitation (CPR) simulator in CPR training.Methods: A total of 60 soldiers were randomly allocated into three groups equally, trained with a simplified external cardiac massage (ECM) trainer named Soul SheathTM (SS) (SS group), a Resusci Anne manikin (RA group), or traditional simulation training with instructor feedback, respectively. After 7 days of training, the CPR skills were tested blindly in a 2-min assessment session. The primary outcome was the proportion of effective CC, and the secondary outcome included CC rate, depth, compression position, and extent of the release.Results: The percentage of effective CC achieved in the SS group was comparable with the RA group (77.0 ± 15.52 vs. 77.5 ± 10.73%, p = 0.922), and significantly higher than that in the control group (77.0 ± 15.52 vs. 66.8 ± 16.87%, p = 0.037). Both the SS and RA groups showed better CC performance than the control group in terms of CC rate (SS group vs. control group, P = 0.032 and RA group vs. control group, P = 0.026), the proportion of shallow CC (SS group vs. control group, P = 0.011 and RA group vs. control group, P = 0.017). No difference between the SS group and RA group was found in all the CC parameters.Conclusions: The simplified ECM trainer (SS) provides a similar efficacy to the traditional manikin simulator with feedback in CC training to improve the quality of CPR skills.

Development controller and feedback of Chiang Mai automated external defibrillation trainer and manikin for basic life support training

Objectives Basic Life Support (BLS) should be taught to everyone; however, there is a lack of Automated External Defibrillation (AED) trainers in Thailand due to the cost of the imported equipment. An AED trainer, manikin, and manikin controller system was locally developed which can display training results to help reduce the cost. Methods This is a descriptive study of a locally developed controller and feedback system, the Chiang Mai AED trainer and manikin. The controller was examined by 5 BLS instructors using a 10-point scale questionnaire consisting of 3 questions to evaluate the efficacy of the device. Correlation among BLS instructors was also calculated. Results The controller and feedback system for the Chiang Mai AED trainer and manikin worked appropriately as intended. The instructor set up either a ‘shock advised’ or a ‘no shock advised’ scenario for 2 two-minute chest compressions. The AED trainer responded promptly when the electrode pads were attached to the manikin. The controller was also able to display compression depth, compression speed, chest recoils, rescue breaths, and time to first defibrillation. Evaluation of the controller using a 10-point scale resulted in a median of 27 (interquartile range 26-29) out of 30. Intraclass correlation was 0.97 (95% confidence interval 0.90-0.99, p < 0.001). Conclusions The controller for the Chiang Mai AED trainer and manikin is effective for use in BLS training. Chiang Mai Medical Journal 2021;60(1):87-98. doi 10.12982/CMUMEDJ.2021.08

Comparing Basic Life Support Serious Gaming Scores With Hands-on Training Platform Performance Scores: Pilot Simulation Study for Basic Life Support Training

Background Serious games enrich simulation-based health care trainings and improve knowledge, skills, and self-confidence of learners while entertaining them. Objective A platform which can combine performance data from a basic life support (BLS) serious game app and hands-on data based on the same scoring system is not available in the market. The aim of this study was to create such a platform and investigate whether performance evaluation of BLS trainings would be more objective compared to conventional Objective Structured Clinical Examination (OSCE) examinations if these evaluations were carried out with the platform which combines OSCE scoring criteria with sensor data retrieved from the simulator’s sensors. Methods Participants were 25 volunteers (11 men [44.0%] and 14 [56.0] women) among Acıbadem Mehmet Ali Aydınlar University students without prior knowledge of the BLS protocol. A serious game module has been created for teaching learners the European Resuscitation Council Basic Life Support 2015 protocol. A second module called the hands-on module was designed for educators. This module includes a checklist used for BLS OSCE examinations and can retrieve sensor data such as compression depth, compression frequency, and ventilation volume from the manikin (CPR Lilly; 3B Scientific GmbH) via Bluetooth. Data retrieved from the sensors of the manikin enable educators to evaluate learners in a more objective way. Performance data retrieved from the serious gaming module have been combined with the results of the hands-on module. Data acquired from the hands-on module have also been compared with the results of conventional OSCE scores of the participants, which were obtained by watching the videos of the same trainings. Results Participants were considered successful in the game if they scored 80/100 or above. Overall, participants scored 80 or above in an average of 1.4 (SD 0.65) trials. The average BLS serious game score was 88.3/100 (SD 5.17) and hands-on average score was 70.7/100 (SD 17.3), whereas the OSCE average score was 84.4/100 (SD 12.9). There was no statistically significant correlation between success on trials (score ≥80/100), serious game, hands-on training app, and OSCE scores (Spearman rho test, P>.05). The mean BLS serious game score of the participants was 88.3/100 (SD 5.17), whereas their mean hands-on training app score was 70.7/100 (SD 17.3) and OSCE score was 84.4/100 (SD 12.9). Conclusions Although scoring criteria for OSCE and hands-on training app were identical, OSCE scores were 17% higher than hands-on training app scores. After analyzing the difference of scores between hands-on training app and OSCE, it has been revealed that these differences originate from scoring parameters such as compression depth, compression frequency, and ventilation volume. These data suggest that evaluation of BLS trainings would be more objective if these evaluations were carried out with the modality, which combines visual OSCE scoring criteria with sensor data retrieved from the simulator’s sensors. Trial Registration ClinicalTrials.gov NCT04533893; https://clinicaltrials.gov/ct2/show/NCT04533893

PP25 Does unguided cardio-pulmonary-resuscitation in copenhagen achieve high quality recommendations?

BackgroundSurvival from out-of-hospital cardiac arrest (OHCA) is associated with the quality of cardio-pulmonary-resuscitation (CPR). The European Resuscitation Council (ERC) and American Heart Association (AHA) define high quality CPR as compression depth of 5–6 centimeters, compression rate of 100–120 compressions/minute, full recoil (>400 milliseconds) after each compression and a hands-on time (compression fraction) of at least 60% (ERC) or 80% (AHA). The aim of this study was to investigate if unguided CPR performed by Copenhagen Emergency Medical Services (EMS) met these recommendations.MethodFrom October throughout December 2018, OHCA data were collected from ambulances within the Capital Region of Denmark using Zoll X-series defibrillator (without CPR feedback dashboard or metronome). Only cases where EMS performed CPR were included. Data was uploaded to a central database and extracted to EXCEL for descriptive statistics and preliminary results.ResultsEMS CPR was performed in 330 cases of which 252 were available for analysis. Mean (SD) compression depth was 5.6±1.7 centimeters, compression rate was 110±9.8 compressions/minute, release velocity was 410±125.1 milliseconds, compression quality (correct compression depth + correct compression rate) was 13.8% and compression fraction was 69.7%.ConclusionThe quality of EMS-delivered CPR, unguided by feedback or metronome, was within recommendations for compression depth, compression rate and release velocity. CPR fraction was between ERC and AHA guidelines. Compression quality, which is not included in ERC/AHA recommendations, did not reach the manufactures recommended >60%. Further work is ongoing to evaluate the effect of adding real-time feedback to guide EMS CPR.

Comparing Basic Life Support Serious Gaming Scores With Hands-on Training Platform Performance Scores: Pilot Simulation Study for Basic Life Support Training (Preprint)

BACKGROUND Serious games enrich simulation-based health care trainings and improve knowledge, skills, and self-confidence of learners while entertaining them. OBJECTIVE A platform which can combine performance data from a basic life support (BLS) serious game app and hands-on data based on the same scoring system is not available in the market. The aim of this study was to create such a platform and investigate whether performance evaluation of BLS trainings would be more objective compared to conventional Objective Structured Clinical Examination (OSCE) examinations if these evaluations were carried out with the platform which combines OSCE scoring criteria with sensor data retrieved from the simulator’s sensors. METHODS Participants were 25 volunteers (11 men [44.0%] and 14 [56.0] women) among Acıbadem Mehmet Ali Aydınlar University students without prior knowledge of the BLS protocol. A serious game module has been created for teaching learners the European Resuscitation Council Basic Life Support 2015 protocol. A second module called the hands-on module was designed for educators. This module includes a checklist used for BLS OSCE examinations and can retrieve sensor data such as compression depth, compression frequency, and ventilation volume from the manikin (CPR Lilly; 3B Scientific GmbH) via Bluetooth. Data retrieved from the sensors of the manikin enable educators to evaluate learners in a more objective way. Performance data retrieved from the serious gaming module have been combined with the results of the hands-on module. Data acquired from the hands-on module have also been compared with the results of conventional OSCE scores of the participants, which were obtained by watching the videos of the same trainings. RESULTS Participants were considered successful in the game if they scored 80/100 or above. Overall, participants scored 80 or above in an average of 1.4 (SD 0.65) trials. The average BLS serious game score was 88.3/100 (SD 5.17) and hands-on average score was 70.7/100 (SD 17.3), whereas the OSCE average score was 84.4/100 (SD 12.9). There was no statistically significant correlation between success on trials (score ≥80/100), serious game, hands-on training app, and OSCE scores (Spearman rho test, <i>P</i>>.05). The mean BLS serious game score of the participants was 88.3/100 (SD 5.17), whereas their mean hands-on training app score was 70.7/100 (SD 17.3) and OSCE score was 84.4/100 (SD 12.9). CONCLUSIONS Although scoring criteria for OSCE and hands-on training app were identical, OSCE scores were 17% higher than hands-on training app scores. After analyzing the difference of scores between hands-on training app and OSCE, it has been revealed that these differences originate from scoring parameters such as compression depth, compression frequency, and ventilation volume. These data suggest that evaluation of BLS trainings would be more objective if these evaluations were carried out with the modality, which combines visual OSCE scoring criteria with sensor data retrieved from the simulator’s sensors. CLINICALTRIAL ClinicalTrials.gov NCT04533893; https://clinicaltrials.gov/ct2/show/NCT04533893

OPTIMIZATION OF DOUBLY REINFORCED BEAM DESIGN USING SIMULATED ANNEALING

The optimization of the doubly reinforced concrete beam was investigated in this paper using the simulated annealing. Materials costs are considered as the objective function. The variables are the width, depth, compression steel, tension steel and cost. The constraints are the ultimate moment of resistance, compression/tension-steel ratio, minimum and maximum area of reinforcements. At the concrete compressive strength of 25 MPa, it is demonstrated that simulated annealing method can be used to optimize the design of concrete beams.

A Complexity Reduction Scheme for Depth Coding in 3D-HEVC

3D-high efficiency video coding (3D-HEVC) is the next-generation compression standard for multiview system applications, which has recently been approved by MPEG and VCEG as an extension of HEVC. To improve the compression efficiency of depth map, several compression tools have been developed for a better representation depth edges. These supplementary coding tools together with existing prediction modes can achieve high compression efficiency, but require a very high complexity that restricts the encoders from ongoing application. In this paper, we introduce a fast scheme to reduce complexity of depth coding in inter and intramode prediction procedure. A simulation analysis is performed to study intra and intermode distribution correlations in the depth compression information. Based on that correlation, we exploit two complexity reduction strategies, including early SKIP and adaptive intra prediction selection. Experimental results demonstrate that our scheme can achieve a complexity reduction up to 63.0%, without any noticeable loss of compression efficiency.