Machine-learning-based head impact subtyping based on the spectral densities of the measurable head kinematics

Background:Traumatic brain injury can be caused by head impacts,but many brain injury risk estimation models are not equally accurate across the variety of impacts that patients may undergo,and the characteristics of different types of impacts are not well studied.We investigated the spectral characteristics of different head impact types with kinematics classification.Methods:Data were analyzed from 3262 head impacts from lab reconstruction,American football,mixed martial arts,and publicly available car crash data.A random forest classifier with spectral densities of linear acceleration and angular velocity was built to classify head impact types(e.g.,football,car crash,mixed martial arts).To test the classifier robustness,another 271 lab-reconstructed impacts were obtained from 5 other instrumented mouthguards.Finally,with the classifier,type-specific,nearest-neighbor regression models were built for brain strain.Results:The classifier reached a median accuracy of 96% over 1000 random partitions of training and test sets.The most important features in the classification included both low-and high-frequency features,both linear acceleration features and angular velocity features.Different head impact types had different distributions of spectral densities in low-and high-frequency ranges(e.g.,the spectral densities of mixed martial arts impacts were higher in the high-frequency range than in the low-frequency range).The type-specific regression showed a generally higher R2value than baseline models without classification.Conclusion:The machine-learning-based classifier enables a better understanding of the impact kinematics spectral density in different sports,and it can be applied to evaluate the quality of impact-simulation systems and on-field data augmentation.

Standardized Evaluation of Camera-based Driver State Monitoring Systems

Driver state sensing technologies, such as vehicular systems, start to be widely considered by automotive manufacturers. To reduce the cost and minimize the intrusiveness towards driving, the majority of these systems rely on the in-cabin camera(s) and other optical sensors. With their great capabilities in detecting and intervening of driver distraction and inattention,these technologies may become key components in future vehicle safety and control systems. However, to the best of our knowledge,currently, there is no common standard available to objectively compare the performance of these technologies. Thus, it is imperative to develop one standardized process for evaluation purposes.In this paper, we propose one systematic and standardized evaluation process after successfully addressing three difficulties:1) defining and selecting the important influential individual and environmental factors, 2) countering the effects of individual differences and randomness in driver behaviors, and 3) building a reliable in-vehicle driver head motion tracking tool to collect ground-truth motion data. We have collected data on a large scale on a commercial driver state-sensing platform. For each subject, 30 to 40 minutes of head motion data was collected and included variables, such as lighting conditions, head/face features,and camera locations. The collected data was analyzed based on a proposed performance measure. The results show that the developed process can efficiently evaluate an individual camerabased driver state sensing product, which builds a common base for comparing the performance of different systems.