Dr. Dallas H Snider

INTRODUCTION: Despite significant advancements in aerospace engineering and safety protocols over the last decade, U.S. Naval mishap rates have remained essentially unchanged. This paper explores how researchers may leverage current artificial intelligence (AI) technologies to enhance aviation safety.
METHODS: A critical review was performed identifying aviation research protocols which have incorporated machine learning (ML) to enhance the accuracy of detecting common aviation hazards leading to cognitive decrements. The review proposes a three-step methodology for creating protocols to identify cognitive decrements in aviators: 1) sensor selection; 2) preprocessing techniques; and 3) ML algorithm development. Natural language processing was utilized to assist with the development of aviation-related denoising and ML algorithm tables.
RESULTS: Several psychophysiological biosensors, enhanced by ML modeling, show promise in identifying cognitive deficits secondary to fatigue, hypoxia, and spatial disorientation. The most cited biosensors integrated with ML models include electroencephalographic, electrocardiographic, and eye-tracking devices. The application of preprocessing techniques to biosensor data is a critical methodological step prior to applying ML algorithms for data training and classification. ML algorithms utilized were categorized into supervised, unsupervised, and semi-supervised types, often used in combination for more accurate predictions.
DISCUSSION: Current literature suggests that AI, when used in conjunction with various psychophysiological sensors, can predict and potentially mitigate common aeromedical hazards such as fatigue, spatial disorientation, and hypoxia in simulated settings. The miniaturization of preprocessing and ML algorithmic hardware is the next phase of transitioning AI to operational environments for real-time continuous monitoring.

Actigraphy is a quantitative means of measuring activity data that has proven viable in post-surgery recovery analysis for arthroplasties in lower extremities, but scant literature has been published on the utilization actigraphy to evaluate shoulder motion and function before and after shoulder arthroplasty. The purpose of this prospective cohort study is to identify if actigraphy can serve as a valid means for objective evaluation of shoulder function and motion before and after shoulder arthroplasty. Secondarily, the data collected by the actigraphy can be analyzed with standard patient-reported outcomes to report correlations between the subjective and objective methods used in this study.
Sixty-four subjects wore an actigraphy device for one day at pre-op, six, twelve and twenty-four weeks. In addition, subjects completed three patient-reported outcome surveys at each time-point. Student t-tests were used to compare percent activity preoperatively with 24-weeks and to compare PROs preoperatively with 24-week results; categorical variables were compared with one-way ANOVAs.
All Patient reported outcome scores significantly improved following arthroplasty (p-value<0.001). The percent of physical activity was highly correlated with vector magnitude (p-value<0.001), but neither percent activity or the vector magnitude were correlated with any of the PROs: UCLA Pain p-value = 0.656, SANE p-value = 0.328, UCLA Function p-value = 0.532.
Actigraphy results from this study mirror findings in previous literature utilizing the technology in similar manners and demonstrate its potential for motion and function analysis before and after total shoulder arthroplasties. Despite both being suitable methods independently for the evaluation of shoulder function, there was no significant correlation between standard actigraphy measurements and PROs at 24-weeks. Future research to determine clinical utility and an overall broader scope for actigraphy monitoring could benefit from improved technology, such as increased battery life for prolonged durations of data collection during observation periods.

This paper presents an on-going research on machine learning (ML) systems for connected vehicle security. It proposes the application of various ML techniques that are applied to Basic Safety Message (BSM) test datasets, both on normal operation and anomalous behavior. The BSM test datasets conform with the SAE J2735 Standard on message sets that support vehicle-to-everything (V2X) communications systems. The purpose of the study is to determine the suitability of ML systems in identifying and classifying normal and anomalous BSM messages in a network of connected vehicles and the V2X systems.

This paper presents a subcomponent of an on-going research on connected vehicle security. It proposes techniques on the generation of synthetic Basic Safety Message (BSM) test datasets, both on normal operation and anomalous behavior. The synthetic test data conform with the SAE J2735 Standard on message sets that support vehicle-to-everything (V2X) communications systems. The purpose of such datasets is for the derivation of machine learning systems that can be deployed in a V2X operating environment.

Organizations make substantial investments in implementing enterprise resource planning (ERP) systems to improve the efficiency and utilization of ERP systems. This study examined the factors influencing and moderating the use of ERP systems. The research variables' hypothetical relationships and moderation analysis were examined through factor analysis and partial least squares structural equation modeling. This study suggests that specific factors significantly influence and moderate the employees' system use. The research results could serve as a reference for vendors when planning the implementation of an ERP system.

Cloud computing is a technology that is expected to have a positive impact on the healthcare industry. However, this technology should be employed after a thorough investigation of relevant industry-specific requirements. This research aimed to uncover the relative importance and moderating effects of service quality, trust, and security on cloud systems usage in the healthcare industry. A web-based survey provided 291 usable responses from which the researchers examined the survey items' internal consistency, reliability, and discriminant and convergent validity. SmartPLS was utilised to investigate the hypothetical relationships using partial least squares (PLS) structural equation modelling (SEM). The study findings indicated that the research variables security, trust, and service quality (SVQ) were found to be the main predictors of cloud computing usage in healthcare. Cloud providers can use these results to deliver better service to the healthcare industry by focusing on the leading issues perceived by healthcare end-users.

Public acceptance of the HPV vaccine has not matched that of other common adolescent vaccines, and HPV vaccination rates remain below the Healthy People 2020 target of 80% compliance. The purpose of this study was to evaluate the capacity of nine pediatric clinics in a Federally Qualified Health Center organization to implement a systems-based intervention targeting office staff and providers using EHRs and a statewide immunization information system to increase HPV vaccination rates in girls and boys, ages 11 to 16 over a 16-month period. System changes included automated HPV prompts to staff, postcard reminders to parents when youths turned 11 or 12 years old, and monthly assessment of provider vaccination rates. During the intervention, 8960 patients (11–16 yo) were followed, with 48.8% girls (n=4370) and 51.2% boys (n=4590). For this study period, 80.5% of total patients received the first dose of the HPV vaccine and 47% received the second dose. For the first dose, 55.5% of 11 year old girls and 54.3% of 11 year old boys were vaccinated. For ages 12 to 16, first dose
vaccination rates ranged from the lowest rate of 84.5% for 14 yo girls up to the highest rate of 90.5% for 13 yo boys. Logistic regression showed age was highly significantly associated with first dose completion (OR 1.565, 95% CI 1.501, 1.631) while males did not have a significant association with first dose completion compared to females. The intervention increased overall counts of first and second HPV vaccination rates.

The capability of detecting symptoms of hypoxia (i.e., reduced oxygen) and other cognitive impairments in-flight with wearable sensors and machine learning based algorithms will benefit the aviation community by saving lives and preventing mishaps. In this study, knowledge discovery processes were implemented to build classification models to predict hypoxia from wearable, dry-EEG sensor data collected from 85 participants in a two-phase study. Over a 35-minute period and while wearing aviation flight masks which regulated their oxygen intake, participants would alternate between a 2-minute cognitive test on CogScreen Hypoxia Edition and a 3-minute simulated flying task on X-Plane 11, with the oxygen concentration reducing every 5 min following the simulated flight task. The decrease in oxygen each 5 min simulated an increase in altitude. Features extracted from the EEG waveforms were transformed using principal component analysis to reduce the dimensionality of the data. Naïve Bayes, decision tree, random forest, and neural network algorithms were utilized to classify the transformed brain wave data as either normal or hypoxic. The algorithms sensitivity ranged from 0.83 to 1.00 while the specificity ranged from 0.91 to 1.00. This study makes a step forward in developing a real-time, in-flight hypoxia detection system.

Enterprise resource planning (ERP) systems have been widely studied during the past decade, yet they often fail to deliver the intended benefits originally expected. One notable reason for their failures is the lack of understanding in users' requirements. This study was designed to understand the relative importance of system quality, information quality, service quality, and their influence on ERP users in the Middle East. The dependent variable individual impact was used to represent the ERP success at the individual level of analysis. The results from this study were compared to the results attained by Petter et al. in their 2008 analysis of North American ERP users. In addition, the moderating effect of users' characteristics on the individual impact variable was examined along an investigation of the items' reliability, internal consistency, convergent, and discriminant validity. Assessing the level of impact from users may help organizations assess the impacts of ERP users' performance and productivity and create training to improve attitudes toward ERP systems.