Eye Movement Characteristics Research Articles

Predicting Chinese reading proficiency based on eye movement features and machine learning

Predicting Chinese reading proficiency based on eye movement features and machine learning

Advancements in driver fatigue detection: A comprehensive analysis of eye movement and facial feature approaches

This article goes into sophisticated tiredness detection methods that make use of computer vision algorithms, including the Eye Movement and Facial Feature approaches. Eye movement analysis, which emphasizes blink frequency, Scanning speed, and gaze length, excels in tasks requiring sustained visual attention, which is critical in high-attention vocations. Facial Feature detection, on the other hand, monitors changes in expressions, muscle activity, and emotions, providing flexibility across a wide range of settings. Eye movement enables accurate indications, real-time responsiveness, and task-specific precision, which is especially important in high-attention occupations. Facial Feature is ideal for a variety of scenarios since it gives complete assessments, flexibility, contextual analysis, and non-intrusiveness. Case examples show how eye-tracking and convolutional neural networks may be used to improve accuracy. This study helps to weariness management by detailing the advantages and disadvantages of Eye Movement and Facial Feature methods. Given the pervasiveness of fatigue, knowing these sensing systems is critical for maintaining safety, productivity, and general well-being in a variety of professional and everyday life scenarios.

PRECYSE: Predicting Cybersickness using Transformer for Multimodal Time-Series Sensor Data

Cybersickness, a factor that hinders user immersion in VR, has been the subject of ongoing attempts to predict it using AI. Previous studies have used CNN and LSTM for prediction models and used attention mechanisms and XAI for data analysis, yet none explored a transformer that can better reflect the spatial and temporal characteristics of the data, beneficial for enhancing prediction and feature importance analysis. In this paper, we propose cybersickness prediction models using multimodal time-series sensor data (i.e., eye movement, head movement, and physiological signals) based on a transformer algorithm, considering sensor data pre-processing and multimodal data fusion methods. We constructed the MSCVR dataset consisting of normalized sensor data, spectrogram formatted sensor data, and cybersickness levels collected from 45 participants through a user study. We proposed two methods for embedding multimodal time-series sensor data into the transformer: modality-specific spatial and temporal transformer encoders for normalized sensor data (MS-STTN) and modality-specific spatial-temporal transformer encoder for spectrogram (MS-STTS). MS-STTN yielded the highest performance in the ablation study and the comparison of the existing models. Furthermore, by analyzing the importance of data features, we determined their relevance to cybersickness over time, especially the salience of eye movement features. Our results and insights derived from multimodal time-series sensor data and the transformer model provide a comprehensive understanding of cybersickness and its association with sensor data. Our MSCVR dataset and code are publicly available: https://github.com/dayoung-jeong/PRECYSE.git.

Evaluation of mental load using EEG and eye movement characteristics

Mental load is a major cause of human-induced accidents. In this study, an explosive impact sensitivity experiment was used to induce mental load. A combination of subjective questionnaires and objective prospective time-distance tests were used to judge whether subjects experienced mental load. Four indicators, namely, β, γ, mean pupil diameter, and fixation time were selected by statistical analysis and PCA for the construction of a mental load assessment model. The study found that the occipital lobe was the most sensitive to mental load, especially β and γ bands. Lastly, it was found that subjects showed different degrees of mental load for the same mental load induction task. The results of the study are applicable to the evaluation and monitoring of the mental characteristics of workers and provide a scientific basis for adjusting the mental load of workers over time to reduce the rate of accidents and enhance production efficiency.

AI-assisted evaluation of problem-solving performance using eye movement and handwriting

With the ability to predict learning behaviors, artificial intelligence (AI) is increasingly involved in assessing the performance of problem solving. This study explored the potential of AI to predict whether mathematics problems could be solved based on eye movements and handwriting in a digital problem-solving environment. Sixty-one students participated in the experiment. The goal is to examine whether types of eye movement features (AOI-based and fixation-based features), levels of information while solving problems (separated and integrated steps), and handwriting could impact the performance of AI. The results indicated that fixation-based features outperform AOI-based features. Furthermore, information in separated steps could provide higher accuracy than that in integrated steps. Inconsistent patterns between human and AI-based assessment of solvers’ answers are discussed.

An Emotion Recognition Method Based on Eye Movement and Audiovisual Features in MOOC Learning Environment

In recent years, more and more people have begun to use massive online open course (MOOC) platforms for distance learning. However, due to the space–time isolation between teachers and students, the negative emotional state of students in MOOC learning cannot be identified timely. Therefore, students cannot receive immediate feedback about their emotional states. In order to identify and classify learners’ emotions in video learning scenarios, we propose a multimodal emotion recognition method based on eye movement signals, audio signals, and video images. In this method, two novel features are proposed: feature of coordinate difference of eyemovement (FCDE) and pixel change rate sequence (PCRS). FCDE is extracted by combining eye movement coordinate trajectory and video optical flow trajectory, which can represent the learner’s attention degree. PCRS is extracted from the video image, which can represent the speed of image switching. A feature extraction network based on convolutional neural network (CNN) (FE-CNN) is designed to extract the deep features of the three modals. The extracted deep features are inputted into the emotion classification CNN (EC-CNN) to classify the emotions, including interest, happiness, confusion, and boredom. In single modal identification, the recognition accuracies corresponding to the three modals are 64.32%, 74.67%, and 71.88%. The three modals are fused by feature-level fusion, decision-level fusion, and model-level fusion methods, and the evaluation experiment results show that the method of decision-level fusion achieved the highest score of 81.90% of emotion recognition. Finally, the effectiveness of FCDE, FE-CNN, and EC-CNN modules is verified by ablation experiments.

Prevalence and Clinical Characteristics of Sleeping Paralysis: A Systematic Review and Meta-Analysis.

Sleep paralysis (SP) is a mixed state of consciousness and sleep, combining features of rapid eye movement (REM) sleep with those of wakefulness. The exact cause of SP is unknown, and its prevalence varies among the studies. We aim to identify SP's global prevalence, the affected population's characteristics, and the SP's clinical picture. We searched three databases (PubMed, Scopus, and Web of Science (WoS)) using a unique search strategy to identify eligible studies. All observational studies identifying the prevalence or frequency of sleeping paralysis were included. No exclusions are made based on country, race, or questionnaire. The analysis was performed using the latest version of R software (R Core Team, Vienna, Austria). The analysis included 76 studies from 25 countries with 167,133 participants. The global prevalence of SP was 30% (95% CI (22%, 39%)). There were similar frequencies of isolated SP and SP (33%, 95% CI (26%, 42%), I2 = 97%, P <0.01; 31%, 95% CI (21%, 43%), I2 = 100%, P = 0, respectively). A subgroup analysis showed that the majority of those who experienced SP were psychiatric patients (35%, 95% CI (20%, 55%), I2 = 96%, P <0.01). The prevalence among non-psychiatric patients was among students (34%, 95% CI (23%, 47%), I2 = 100%, P = 0). Auditory and visual hallucinations were reported in 24.25% of patients. Around 4% had only visual hallucinations. Meta-regression showed no association between the frequency of SP and sex. Publication bias was detected among the included studies through visual inspection of funnel plot asymmetry. Our findings revealed that 30% of the population suffered from SP, especially psychiatric patients and students.The majority of SP cases lacked associated hallucinations, while a noteworthy proportion experienced combined visual and auditory hallucinations.

Eye Movement Characteristics for Predicting a Transition to Psychosis: Longitudinal Changes and Implications.

Substantive inquiry into the predictive power of eye movement (EM) features for clinical high-risk (CHR) conversion and their longitudinal trajectories is currently sparse. This study aimed to investigate the efficiency of machine learning predictive models relying on EM indices and examine the longitudinal alterations of these indices across the temporal continuum. EM assessments (fixation stability, free-viewing, and smooth pursuit tasks) were performed on 140 CHR and 98 healthy control participants at baseline, followed by a 1-year longitudinal observational study. We adopted Cox regression analysis and constructed random forest prediction models. We also employed linear mixed-effects models (LMMs) to analyze longitudinal changes of indices while stratifying by group and time. Of the 123 CHR participants who underwent a 1-year clinical follow-up, 25 progressed to full-blown psychosis, while 98 remained non-converters. Compared with the non-converters, the converters exhibited prolonged fixation durations, decreased saccade amplitudes during the free-viewing task; larger saccades, and reduced velocity gain during the smooth pursuit task. Furthermore, based on 4 baseline EM measures, a random forest model classified converters and non-converters with an accuracy of 0.776 (95% CI: 0.633, 0.882). Finally, LMMs demonstrated no significant longitudinal alterations in the aforementioned indices among converters after 1 year. Aberrant EMs may precede psychosis onset and remain stable after 1 year, and applying eye-tracking technology combined with a modeling approach could potentially aid in predicting CHRs evolution into overt psychosis.

Investigation of eye movement characteristics during free throws at varying intensities among basketball players and its correlation with free throw percentage.

Free throws serve as a crucial scoring mechanism in basketball games. As the level of basketball and the intensity of competition continue to improve, the frequency of free throw attempts gradually rises. However, with heightened game intensity, maintaining a consistent free throw success rate becomes increasingly challenging. The aim of this study was to investigate the eye movement characteristics exhibited by basketball players during free throws of varying intensities, as well as explore the relationship between these eye movement characteristics and free throw percentage. Twenty elite female basketball players were recruited to perform free throws at varying exercise intensities (low, moderate, and high) while wearing an eye tracker device. Eye tracking data was collected using the Tobii Glasses 3 eye tracker. Additionally, the Polar team pro, a heart rate monitoring system manufactured in Finland, was utilized to monitor participants' heart rates during different exercise intensities. The average number of fixations on the hoop and net during free throws of varying intensities exhibited statistically significant differences (P < 0.05). The fixation durations of the hoop, backboard, and net exhibited statistically significant differences (P < 0.05). Notably, the hoop accounted for the highest proportion of the overall average fixation duration. high-intensity free throws necessitate lengthier processing durations compared to low-intensity and medium-intensity free throws. Furthermore, for free throws of moderate intensity, there was a significant negative correlation between the number of fixations on the hoop and free throw percentage (P < 0.05). Conversely, for high intensity free throws, there was a significant positive correlation between the fixation duration on the hoop and free throw percentage (P < 0.05). Under the three kinds of sports intensity, the players mainly focus on the hoop position, the moderate intensity free throw has the best stability, and the information search strategy and information processing efficiency are the highest. In high intensity free throws, the longer the fixation duration at the hoop, the higher the free throw percentage; The higher the number of fixations at the hoop, the lower the percentage of free throws.

An interpretable English reading proficiency detection model in an online learning environment: A study based on eye movement

Aiming at the low English reading proficiency of ESL (English as second language) students in online learning environments, this study proposed an eye-movement-based machine learning monitoring model to detect English reading proficiency in real time. Eye-movement data from 43 students while completing online English reading tasks were recorded and 31 eye-movement features were extracted from the taxonomy of fixation, saccade, movement direction and gaze velocity. During the model training phase, LightGBM achieved an accuracy of 96.51 % in detection. An interpretable model, SHAP (SHapley Additive exPlanation), was used to explain the main effects of eye-movement features in detection, where high gaze velocity, absolute saccade direction, and average saccade duration were found to be strong indicators of English reading proficiency. Furthermore, SHAP analysis allows the identification of individual factors contributing to differences in English reading proficiency. This study demonstrated the effectiveness of the combination of eye-movement data with machine learning methods to identify students with low English reading proficiency in online reading and provided insights into the intrinsic correlation between eye movements and reading.

Анализ характеристик движения глаз оператора для определения степени его утомления

The article presents a method for searching characteristics of eye movements that correlate with fatigue. There are many characteristics of eye movements to determine fatigue. All these characteristics are calculated from such basic concepts of gaze movement as fixation and saccade. Characteristics can also be divided by the types of physical events on which they are based. It is possible to distinguish such characteristics as speed, time, quantity, size, percentage, frequency and ratio characteristics. To search for correlations between eye movement characteristics and fatigue, a dataset of eye movements and the results of the VAS-F fatigue questionnaire were analyzed in 6 subjects. The data set consists of operator parameters such as eye movements, scene camera image and gaze direction. To determine the level of fatigue, the participant completed the VAS-F questionnaire. This questionnaire consists of 18 questions about the degree of fatigue or enrgetic of a person. Each record from the data set corresponds to a questionnaire result. 60 characteristics of eye movements and the corresponding VAS-F test values were analyzed and the correlation between them was calculated. The characteristics of eye movements were then sorted in descending order of the obtained correlation values. For further analysis, the first 20 characteristics with the highest correlation were selected from each participant. A search was then made for characteristics that were found in two-thirds or more of the participants among the first 20 characteristics. As a result, 10 characteristics of eye movements were found that correlated with VAS-F test scores for each participant.

How do Eye Movements and Visual Attention in Letter-Finding Tasks Differ in Children with Different Levels of Reading Skills?

Children with different levels of reading proficiency demonstrate different information processing speed, reading accuracy and cognitive strategies. The study presents the results of analyses of eye movement features when performing letter search tasks in lexical and sub-lexical environments. The task of searching for a particular grapheme by the respondents is approximated to real-life conditions, since in the course of language learning students regularly face the tasks of searching for certain orthograms in educational texts or self-checking tasks (i.e., errors in their own texts), which is especially important for children with dysorphographia, regulatory dysgraphia, and other learning difficulties. Forty-nine children aged 9 to 10 years participated in the study. Reading skills were assessed using the Standardized Assessment of Reading Skills (SARS). Nonverbal intelligence was assessed by using Raven’s coloured progressive matrices. Participants performed a letter search task in text and letter list, eye movements were recorded using the EyeLink 1000 Plus eye tracker. Rank correlation analyses showed that the level of reading skill development was related to the efficiency of finding specific items in a given context. In addition, the duration and number of gaze fixations on the searched units in the letter list were not always related to the number of correct answers, indicating a complex process of visual attention during the search task. In addition, it was observed that foveal vision was not always necessary for performing search tasks in children with normally developed reading skill levels. The study also investigated the concept of functional visual field and how the surroundings of the items being searched affect the attentional mechanisms involved in visual search. Overall, this study provides evidence on the relationship between reading skills, attention and visual information processing, which contributes to understanding reading strategies in children with dyslexia and creating new methods for screening children with dyslexia in the future.

A comparative analysis for eye movement characteristics between professional and non-professional players in FIFA eSports game

Eye-tracking technology has been widely applied to various kinds of traditional sports research, such as football, basketball and table tennis. eSports is an emerging digital sport, which is currently enjoying rapid development. The previous study primarily focuses on the practical applications of eye movement in eSports, such as game design, visual attention, and performance analysis. However, there is a scientific gap for the specific differences between expert and novice players, the potential of eye-tracking as a biological marker for eSports performance, and the transfer of insights from real sports to eSports through eye-tracking studies. This study aims to uncover disparities in eye movement characteristics between professional players (N = 14, 24.1 ± 2.5 years) and non-professional players (N = 14, 25.2 ± 2.3 years) by employing the expert-novice paradigm. We extracted several key eye movement features, including average fixation counts, average fixation duration, first fixation duration, eye movement trajectory, and visits to areas of interest. The results indicate that professional players exhibit higher fixation counts, shorter first fixation duration, reduced average fixation duration, and employ a more effective visual search strategy. Professional players’ fixation trajectories are characterized by clarity and focus, with a greater number of fixation points concentrated on key areas of interest. In contrast, non-professional players demonstrate relatively random fixation trajectories, lacking distinct focal points. This study not only advances our understanding of eye-tracking in eSports but also provides technical guidance and practical significance for further research in this burgeoning field.

Eye Movement on Predicting Serve Return of College Female Tennis Players of Different Levels in Henan Province

Objective: To quantitatively evaluate the influence of sports level on the behavioral indexes and eye movement indexes of female tennis players by independent sample T test analysis, and to provide strategic basis for visual search of athletes. Methods: In this study, 20 female tennis players from colleges and universities in Henan Province were randomly selected as professional group, and 20 female tennis students from colleges and universities in Henan Province were randomly selected as amateur group. All the tested players had normal vision (5.0 after correction), healthy body, and were right-handed racket holding players. The experiment was conducted to compare the two groups of subjects' prediction accuracy rate, reaction time, total number of effective gaze, gaze track and interest area and other eye movement characteristics when watching videos. Results: The professional group had shorter reaction time, and the correct rate was obviously higher than the amateur group. The total effective fixation point of the professional group was more than that of the amateur group, and the professional group could focus more information. The professional group had a mature grasp of cognitive processing strategies, and their visual search times were targeted; Athletes in the professional group can allocate more attention to the key areas of interest that affect the serve landing point, and highly focus on the interest areas directly related to the direction of the serve, so that they can search information more accurately and improve the accuracy of the prediction.

Vertical dynamic visual acuity is significantly lower than horizontal dynamic visual acuity

Dynamic visual acuity (DVA) is crucial for the perception of moving objects. While traditional DVA assessment tools predominantly focus on horizontal movements, the evaluation of vertical DVA remains unstandardized. Consequently, the disparities between vertical and horizontal DVAs are yet to be thoroughly investigated. Therefore, we designed a system capable of conducting multidirectional DVA tests and eye movement measurements. During the experiments, the participants identified the gap direction of the Landolt-C ring moving either horizontally or vertically. The speed of movement decelerated from its maximum as a high-speed infrared camera captured the pupil movements of the left eye at 500 fps. We conducted tests on 15 healthy university students (aged 22.7±1.2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$22.7\\pm 1.2$$\\end{document} years) and measured vertical and horizontal DVAs five times each. DVA was deduced from the Landolt-C ring speed with accurate gap direction responses, and eye movement was assessed based on the total gaze movement distance. The results revealed superior DVA and eye movement in the horizontal direction compared with the vertical direction (p<0.001\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$p < 0.001$$\\end{document}). This highlights the anisotropic characteristics of DVA and eye movement. The proposed system has the potential for multidirectional dynamic vision evaluation and training in clinical scenarios.

Differential eye movement features between Alzheimer's disease patients with and without depressive symptoms.

Accurately diagnosing depressive symptoms in Alzheimer's disease (AD) patients is often challenging. Eye movement parameters have been demonstrated as biomarkers for assessing cognition and psychological conditions. To investigate the differences in eye movement between AD patients with and without depressive symptoms. Eye movement data of 65 AD patients were compared between the depressed AD (D-AD) and non-depressed AD (nD-AD) groups. Logistic regression analysis was employed to identify diagnostic biomarkers and the ROC curve was plotted. The correlation between eye movement and HAMD-17 scores was assessed by partial correlation analysis. The D-AD patients showed longer saccade latency and faster average/peak saccade velocities in the overlap prosaccade test, longer average reaction time and faster average saccade velocity in the gap prosaccade test, longer start-up durations, slower pursuit velocity, more offsets, and larger total offset degrees in the smooth pursuit test, and poorer fixation stability in both the central and lateral fixation tests compared to nD-AD patients. The start-up duration in the smooth pursuit test and the number of offsets in the central fixation test were identified as the diagnostic eye movement parameters for D-AD patients with the area under the ROC curves of 0.8011. Partial correlation analysis revealed that the start-up duration and pursuit velocity in the smooth pursuit test and the total offset degrees in the lateral fixation test were correlated with HAMD-17 scores in D-AD patients. Eye movement differences may help to differentiate D-AD patients from nD-AD patients in a non-invasive and cost-effective manner.

Visual search strategies of performance monitoring used in action anticipation of basketball players.

Numerous studies have found that expert players anticipate better than novices. If more accurate prediction represents performance monitoring of experts, what are the advantages of elite basketball players in identifying and processing available cues? There is still a lack of sufficient evidence. This study examined the visual search in basketball players and explored the performance monitoring of action anticipation, adopting an expert-novice paradigm and eye-movement technology. Forty basketball players were recruited in this study: 20 in the expert group and 20 in the novice group. Participants were asked to predict the outcome of videotaped basketball throws and their accuracy and eye-movement characteristics were record. The accuracy of the expert was significantly higher than that of the novice. The experts were able to instantly search and identify important cues in anticipation, and the gaze area of the experts was concentrated on the area of interest of the body. Additionally, the expert group showed long, repetitive, and rapid visual search of vital information, and improved their performance of the task. The experts could monitor the performance of prediction by grabbing vital shooting information (such as the body of a player). The results suggest the athletes and coaches that if they want to improve the ability of prediction, it may be useful to shift their focus of attention from ball trajectory to body action.

Universality in eye movements and reading: A replication with increased power

Liversedge, Drieghe, Li, Yan, Bai and Hyönä (2016) reported an eye movement study that investigated reading in Chinese, Finnish and English (languages with markedly different orthographic characteristics). Analyses of the eye movement records showed robust differences in fine grained characteristics of eye movements between languages, however, overall sentence reading times did not differ. Liversedge et al. interpreted the entire set of results across languages as reflecting universal aspects of processing in reading. However, the study has been criticized as being statistically underpowered (Brysbaert, 2019) given that only 19–21 subjects were tested in each language. Also, given current best practice, the original statistical analyses can be considered to be somewhat weak (e.g., no inclusion of random slopes and no formal comparison of performance between the three languages). Finally, the original study did not include any formal statistical model to assess effects across all three languages simultaneously. To address these (and some other) concerns, we tested at least 80 new subjects in each language and conducted formal statistical modeling of our data across all three languages. To do this, we included an index that captured variability in visual complexity in each language. Unlike the original findings, the new analyses showed shorter total sentence reading times for Chinese relative to Finnish and English readers. The other main findings reported in the original study were consistent. We suggest that the faster reading times for Chinese subjects occurred due to cultural changes that have taken place in the decade or so that lapsed between when the original and current subjects were tested. We maintain our view that the results can be taken to reflect universality in aspects of reading and we evaluate the claims regarding a lack of statistical power that were levelled against the original article.

Attention-based automatic editing of virtual lectures for reduced production labor and effective learning experience

Recently there has been a surge in demand for online video-based learning, and the importance of high-quality educational videos is ever-growing. However, a uniform format of videos that neglects individual differences and the labor-intensive process of editing are major setbacks in producing effective educational videos. This study aims to resolve the issues by proposing an automatic lecture video editing pipeline based on each individual’s attention pattern. In this pipeline, the eye-tracking data are obtained while each individual watches virtual lectures, which later go through multiple filters to define the viewer’s locus of attention and to select the appropriate shot at each time point to create personalized videos. To assess the effectiveness of the proposed method, video characteristics, subjective evaluations of the learning experience, and objective eye-movement features were compared between differently edited videos (attention-based, randomly edited, professionally edited). The results showed that our method dramatically reduced the editing time, with similar video characteristics to those of professionally edited versions. Attention-based versions were also evaluated to be significantly better than randomly edited ones, and as effective as professionally edited ones. Eye-tracking results indicated that attention-based videos have the potential to decrease the cognitive load of learners. These results suggest that attention-based automatic editing can be a viable or even a better alternative to the human expert-dependent approach, and individually-tailored videos have the potential to heighten the learning experience and effect.

Evaluation of Traffic Signs Information Volume at Highway Tunnel Entrance Zone Based on Driver’s Visual Characteristics

Reasonable and appropriate traffic sign information volume (TSIV) is crucial to ensure road traffic safety, especially at the entrance zones of highway tunnels. This research compares how various levels of TSIV affect both visual characteristics and visual workload of drivers through real road driving tests. Forty participants were recruited to conduct a field driving experiment at six highway tunnels. The eye movement data of drivers were collected by an eye tracker and the effects of TSIV on drivers’ eye movement characteristics, visual stability, visual SampEn (sample entropy), and visual workload intensity were analyzed and evaluated. At the T3 level (48.31 bits) of TSIV, the drivers’ average fixation duration and average saccade duration were both at the lowest value, while the drivers’ average saccade amplitude reached the maximum, and the dispersion of the three eye movement indicators was the smallest. In addition, the drivers’ visual SampEn increased continuously when approaching the tunnel portal, and was the lowest at T3 level. With the increase of TSIV, drivers’ visual workload intensity decreased first and then increased, the minimum being at T3 level. The drivers’ visual behavior is more stable, visual coordination ability is better, and visual workload and psychological pressure are least under the T3 level, which is beneficial to ensure driving safety at the entrance zone of a highway tunnel. Inappropriate levels of TSIV at highway tunnel entrance zones will cause inevitable risks to driving safety.