Detection Of Changes In Behavior Research Articles

Future Burden of Cancer Attributable to Current Modifiable Behaviours: A Pooled Study of Seven Australian Cohorts

Background: The Population Attributable Fraction (PAF) quantifies the fraction of cancer cases attributable to specific exposures. PAF estimates for the future burden of cancer preventable through modifications to current exposure distributions are lacking. Previous PAF studies have also not compared population subgroup differences. Aim: To apply a novel PAF method and i) assess the future burden of cancer in Australia preventable through modifications to current behaviors, and ii) compare the distribution of the preventable cancer burden between population subgroups. Methods: We harmonized and pooled data from seven Australian cohort studies (N=367058) and linked them to national registries to identify cancers and deaths. We estimated the strength of the associations between behaviors and cancer incidence and death using a proportional hazards model, adjusting for age, sex, study and other risk factors. Exposure prevalence was estimated from contemporary national health surveys. We then combined these estimates to calculate PAFs and their 95% confidence intervals for both individual and joint behavior modifications using a novel method accounting for competing risk of death and risk factor interdependence. We also compared PAFs between population subgroups by calculating the 95% confidence interval of the difference in PAF estimates. Results: During the first 10 years of follow-up, there were 22078 deaths and 27483 incident cancers, including 2025 lung, 3471 colorectal, 640 premenopausal and 2632 postmenopausal breast cancers. The leading preventable cause for lung cancer is current smoking (PAF = 53.7%), for colorectal and postmenopausal breast cancer body fatness or BMI ≥ 25 kg/m2 (PAF = 11.1% and 10.9% respectively), and for premenopausal breast cancer regular alcohol intake (PAF = 12.3%). Three in five lung cancers, but only one in five colorectal and breast cancers, are jointly attributable to potentially modifiable exposures, which also included physical inactivity and inadequate fruit intake for lung, excessive alcohol intake and current smoking for colorectal, regular alcohol intake and current menopausal hormone therapy for 1 year or more for postmenopausal breast and current oral contraceptive use for 5 years or more for premenopausal breast cancer. The cancer burden attributable to modifiable factors is markedly higher in certain population subgroups, including men (lung, colorectal), people with risk factor clustering (lung, colorectal, breast), and individuals with low educational attainment (lung, breast). Conclusion: We provided up-to-date estimates of the future Australian cancer burden attributable to modifiable risk factors, and identified population subgroups that experience the highest preventable burden. Application of the novel PAF method can inform timely public health action to improve health and health equity, by identifying those with the most to gain from programs that support behavior change and early detection.

Long-term behavior change detection approach through objective technological observations toward better adaptation of services for elderly people

Behavior change is associated with important decrease of cognitive and physical capacities among elderly people. Therefore, a proactive detection of long-term behavior changes in early stages of their evolution is a keystone to improve elderly healthcare services. In fact, nowadays’ geriatric methods mainly rely on scales and questionnaires, and are inconvenient to investigate long-term changes on a daily basis. Therefore, our proposed approach for behavior change detection analyzes elderly people behavior over long periods via ambient technologies. In fact, employed technologies are unobtrusive, do not interfere with the natural behavior of elderly people and do not affect their privacy. Furthermore, our long-term behavior analysis is based on the identification of significant behavior change indicators (e.g., mobility, memory, nutrition and social life indicators significantly correlate with cognitive and physical diseases), and the application of efficient statistical techniques that differentiate long-term and short-term changes in analyzed behavior. In addition, our two-year deployment validates our objective technological observations through real correlations with medical observations of nursing-home team.

The Future Colorectal Cancer Burden Attributable to Modifiable Behaviors: A Pooled Cohort Study.

BackgroundPrevious estimates of the colorectal cancer (CRC) burden attributed to behaviors have not considered joint effects, competing risk, or population subgroup differences.MethodsWe pooled data from seven prospective Australian cohort studies (n = 367 058) and linked them to national registries to identify CRCs and deaths. We estimated the strength of the associations between behaviors and CRC risk using a parametric piecewise constant hazards model, adjusting for age, sex, study, and other behaviors. Exposure prevalence was estimated from contemporary National Health Surveys. We calculated population attributable fractions for CRC preventable by changes to current behaviors, accounting for competing risk of death and risk factor interdependence. Statistical tests were two-sided.ResultsDuring the first 10 years of follow-up, there were 3471 incident CRCs. Overweight or obesity explained 11.1%, ever smoking explained 10.7% (current smoking 3.9%), and drinking more than two compared with two or fewer alcoholic drinks per day explained 5.8% of the CRC burden. Jointly, these factors were responsible for 24.9% (95% confidence interval [CI] = 19.7% to 29.9%) of the burden, higher for men (36.7%) than women (13.2%, Pdifference < .001). The burden attributed to these factors was also higher for those born in Australia (28.7%) than elsewhere (16.8%, Pdifference = .047). We observed modification of the smoking-attributable burden by alcohol consumption and educational attainment, and modification of the obesity-attributable burden by age group and birthplace.ConclusionsWe produced up-to-date estimates of the future CRC burden attributed to modifiable behaviors. We revealed novel differences between men and women, and other high–CRC burden subgroups that could potentially benefit most from programs that support behavioral change and early detection.

Deficits in Burrowing Behaviors Are Associated With Mouse Models of Neuropathic but Not Inflammatory Pain or Migraine.

Burrowing, or the removal of material from an enclosed tube, is emerging as a prominent means of testing changes in a voluntary behavior in rodent models of various pain states. Here, we report no significant differences between male and female mice in terms of burrowing performance, in a substantially shorter time frame than previous reports. We found that the color of the burrow tube affects the variability of burrowing performance when tested in a lit room, suggesting that light aversion is at least a partial driver of this behavior. Spared nerve injury (SNI; as a model of neuropathy) impairs burrowing performance and correlates with enhanced mechanical sensitivity as assessed by von Frey filaments, as well as being pharmacologically reversed by an analgesic, gabapentin. Loss of the SNI-induced burrowing deficit was observed with daily testing post-surgery, but not when the testing interval was increased to 5 days, suggesting a confounding effect of daily repeat testing in this paradigm. Intraplantar complete Freund’s adjuvant (as a model of inflammatory pain) and systemic nitroglycerin (as a model of migraine-like symptoms) administration did not induce any burrowing deficit, indicating that assessment of burrowing behavior may not be universally suitable for the detection of behavioral changes across all rodent pain models.

Unsupervised behavior change detection using passive sensor systems in the homes of older adults

Unsupervised behavior change detection using passive sensor systems in the homes of older adults

Class Distribution Curve Based Discretization With Application to Wearable Sensors and Medical Monitoring

Understanding diseases and human activities, and constructing highly accurate classifiers are two important tasks in bio-medicine, healthcare, and wearable sensor technology. Being able to mine high-quality patterns is useful here, as such patterns can help improve understanding and build accurate classifiers. However, most pattern mining algorithms only operate on discrete data; applying them often requires a binning step to discretize continuous attributes. This article presents a new discretization technique, called Class Distribution Curve based Binning (CDC Binning); the main idea is to use a so-called class distribution curve, which measures the class purity in sliding windows over an attribute's range, to construct binning intervals. Experiments show that (1) CDC Binning outperforms existing binning methods in discovering high-quality patterns, especially when the class distribution curve is complicated (e.g. when the two classes are two fairly similar human activities), and (2) it can outperform other binning methods by 10% in classification accuracy when using discovered patterns as features. CDC Binning is particularly useful for applications where the classes/activities to be distinguished are similar to each other. This is especially important in wearable sensor technology where detection of behavioral or activity changes in a person (e.g. fall detection) could indicate a significant medical event.

Short communication: Automated detection of behavioral changes from respiratory disease in pre-weaned calves

Group housing of calves can pose a challenge in identifying respiratory disease; therefore, it is necessary to develop tools that can identify these disease events. In this experiment, pre-weaned calves (n = 30) were housed in groups with an automatic calf feeder and were fitted with an accelerometer. Step activity, lying behaviors, and feeding behaviors were recorded to determine the effect of respiratory disease. All calves were health scored twice daily, and calves with respiratory scores ≥5 were diagnosed with respiratory disease (n = 10). Each diseased calf was match paired with a healthy control based on the date of disease diagnosis, breed, and age. Control calves were determined to be healthy if they had respiratory scores ≤4, as well as fecal, navel, and joint scores of 0 or 1. Diseased calves were less active before, on the day of, and after respiratory disease diagnosis. Furthermore, diseased calves had reduced lying frequencies starting 2 d before diagnosis, as well as after diagnosis. Last, diseased calves consumed less milk on the day of diagnosis when compared with healthy controls. Step activity, lying bouts, and milk intake may prove to be a useful tool in identifying respiratory disease under practical farming, but this requires further research.

Behavioural equivalents as predictors of psychiatric disorder in people with intellectual disability

IntroductionThe last decades’ considerable advances of the psychiatric assessment in People with Intellectual Disability (PwID) do not include the ability to identify Behavioural Equivalents (BE) of psychiatric symptoms and their relationship with different psychiatric syndromes. Recent reports have found BE to be differentially associated with one or more specific Psychiatric Disorders (PD).AimsThe present study was aimed at evaluating the correlations between BE and DSM-5 syndromic groups of symptoms, in a wide multicentric sample of PwID.MethodsAn observational cross-sectional analysis was performed for a sample of 843 adults with ID, randomly or consecutively recruited among those living in residential facilities of the National Healthcare System, or in private institutes of care, or those attending psychiatric outpatient clinics. The total sample was administered with the SPAID-G (Psychiatric Instrument for Intellectual Disabled Adult - General version), which is a checklist for the detection of significant behavioural changes from the baseline. The items of the checklist represents BE of the symptoms of the main psychiatric disorders included in the DSM-5.ResultsMany significant correlations were found, some of the most relevant were for mood disorders. Psychomotor agitation, aggressivity, disorganised behaviour and distractibility were most pronounced in bipolar patients; for depressed patients, irritability and weight loss had higher correlations than in other diagnostic groups.ConclusionsAlthough not diagnostically specific, some BE seem to be more strongly related with specific PD. This line of research could improve the definition of the specific expression and clustering of psychiatric symptoms in PwID.

Using Smart Homes to Detect and Analyze Health Events

Smart homes offer an unprecedented opportunity to unobtrusively monitor human behavior in everyday environments and to determine whether relationships exist between behavior and health changes. Behavior change detection (BCD) can be used to identify changes that accompany health events, which can potentially save lives.

Dose-dependent effects of alcohol administration on behavioral profiles in the MCSF test.

The acute effects of alcohol administration are age-, dose-, time- and task-dependent. Although generally considered to be a sedative drug, alcohol has both stimulatory and depressant effects on behavior, depending on dose and time. Alcohol-induced motor activating effects are consistently shown in mice but rarely demonstrated in adult, outbred rats using conventional behavioral tests. The aim of the present experiment was to study acute alcohol-induced effects on behavioral profiles in a more complex environment using the novel multivariate concentric square field™ (MCSF) test, designed for assessing different behaviors in the same trial including locomotor activity. Adult male Wistar rats (Sca:WI) were administered one intraperitoneal (i.p.) injection of alcohol (0.0 g/kg, 0.5 g/kg, 1.0 g/kg, or 1.5 g/kg) 5 min prior to the 30-min MCSF test. The two highest doses induced marked motor-suppressing effects. A significant interaction between group and time was found in general activity when comparing rats exposed to alcohol at 0.0 g/kg and 0.5 g/kg. In contrast to the 0.0 g/kg dose that increased the activity over time, animals administered the low dose (0.5 g/kg) demonstrated an initial high activity followed by a decline over time. No indications for acute alcohol-induced anxiolytic-like effects were found. The multivariate setting in the MCSF test appears to be sensitive for detecting motor-activating effects of low doses of alcohol as well as reduced locomotion at doses lower than in other behavioral tasks. The detection of subtle changes in behavior across time and dose is important for understanding alcohol-induced effects. This approach may be useful in evaluating alcohol doses that correspond to different degrees of intoxication in humans.

Unsupervised discovery of activities of daily living characterized by their periodicity and variability

Habits characterize the activities of elderly people. Monitoring their habits and their ability to carry out the activities of daily living is a great challenge in order to improve aging at home. In particular, the detection of changes in regular behavior may help to detect emerging disorders. The emergence of smart homes and sensor networks allows the non-intrusive collection of data describing the activities in the home. The collected data is indeed an objective source to mine periodic patterns representing the habits of a particular individual. Extended Episode Discovery (xED) algorithm is described and discussed. This algorithm searches for regular patterns, highlighting the periodicity and variability of each discovered pattern. This approach allows a high adaptability to different users and lifestyles. Experiments on six real-life datasets illustrate the interest of xED.

The automatic detection of dairy cow feeding and standing behaviours in free-stall barns by a computer vision-based system

Changes in cow behaviour may occur in relation to health disorders. In literature the suitability of using behavioural changes to provide an early indication of disease is studied. The possibility of achieving a real-time analysis of a number of specific changes in behaviours, such as lying, feeding, and standing, is crucial for disease prevention. Cow feeding and standing behaviour detectors were modelled and validated by defining a methodology based on the Viola–Jones algorithm and using a multi-camera video-recording system to obtain panoramic top-view images of an area of the barn. Assessment of the detection results was carried out by comparison with the results generated by visual recognition. The ability of the system to detect cow behaviours was shown by the high values of its sensitivity achieved for the behaviours of feeding and standing which were about 87% and 86%, respectively. Branching factor values for the two behaviours showed that one false positive was detected for every 13 and 6 well-detected cows, respectively. On the basis of these research outcomes, the proposed system is suitable for computing cow behavioural indices and the real-time detection of behavioural changes.

Unsupervised density-based behavior change detection in data streams

The ability to detect changes in the data distribution is an important issue in Data Stream mining. Detecting changes in data distribution allows the adaptation of a previously learned model to accommodate the most recent data and, therefore, improve its prediction capability. This paper proposes a framework for non-supervised automatic change detection in Data Streams called M-DBScan. This framework is composed of a density-based clustering step followed by a novelty detection procedure based on entropy level measures. This work uses two different types of entropy measures, where one considers the spatial distribution of data while the other models temporal relations between observations in the stream. The performance of the method is assessed in a set of experiments comparing M-DBScan with a proximity-based approach. Experimental results provide important insight on how to design change detection mechanisms for streams.

Growing pigs’ drinking behaviour: number of visits, duration, water intake and diurnal variation

Individual drinking patterns are a potential tool for disease monitoring in pigs. However, to date, individual pig drinking behaviour has not been described, and effects of external factors have not been examined. The aim of this study was to perform detailed quantification of drinking behaviour of growing pigs and to examine effects of period of day and effects of competition for access to the drinking nipple on the drinking behaviour, amount of water used and water wastage. In all, 52 cross-bred castrated male pigs (live weight 20.5±1.7 kg; mean±s.d.) maintained as either 3 (N3) or 10 (N10) pigs per pen and water nipple (four groups/treatment) were used. All pigs were fitted with a transponder ear tag. A radio frequency identification reader recorded and time stamped visits at the nipple. In each pen, water flow was logged every second. The drinking behaviour was recorded for 4 consecutive days and analysed using a linear mixed model. Overall, the pigs spent 594 s at the nipple during 24 h distributed among 44 visits. During this period, 5 l of water were used, of which >30% was wasted. Social competition did not affect the drinking behaviour over 24 h, except for the proportion of interrupted visits where pigs, kept with recommended nipple availability (N10), showed an increased proportion of interrupted drinking bouts compared with pigs kept at very low level of competition (N3) (0.18±0.01 v. 0.11±0.01; P<0.01). However, splitting data into 8-h periods (P1, P2, P3) starting from 0600 h revealed differences between treatments, showing that in N3, water use per visit was lower in P1 than P2 and P3 (110±10 v. 126±7 and 132±7 ml; P<0.05), whereas in N10, the water used per visit was higher during P3 than during the other periods (P1: 107±14 ml, P2: 112±10 ml v. P3: 151±10 ml; P<0.001). A similar pattern was found for visit duration. In N3, fewer nipple visits were observed in P2 than P1 (15.6±1.2 v. 22.0±1.2; P<0.001), whereas no difference was found between P1 and P2 in N10. The results demonstrate that growing pigs at the two levels of competition maintained a comparable level of 24 h water intake by changing behavioural variables involved in drinking. This dynamic characteristic of drinking behaviour means that if individual drinking patterns are to be used as disease monitoring tools, it is important to consider effects of external factors and include data on period level to allow rapid detection of behavioural changes.

Process mining for individualized behavior modeling using wireless tracking in nursing homes.

The analysis of human behavior patterns is increasingly used for several research fields. The individualized modeling of behavior using classical techniques requires too much time and resources to be effective. A possible solution would be the use of pattern recognition techniques to automatically infer models to allow experts to understand individual behavior. However, traditional pattern recognition algorithms infer models that are not readily understood by human experts. This limits the capacity to benefit from the inferred models. Process mining technologies can infer models as workflows, specifically designed to be understood by experts, enabling them to detect specific behavior patterns in users. In this paper, the eMotiva process mining algorithms are presented. These algorithms filter, infer and visualize workflows. The workflows are inferred from the samples produced by an indoor location system that stores the location of a resident in a nursing home. The visualization tool is able to compare and highlight behavior patterns in order to facilitate expert understanding of human behavior. This tool was tested with nine real users that were monitored for a 25-week period. The results achieved suggest that the behavior of users is continuously evolving and changing and that this change can be measured, allowing for behavioral change detection.

Detecting the Effect of Alzheimer's Disease on Everyday Motion Behavior

Early detection of behavioral changes in Alzheimer's disease (AD) would help the design and implementation of specific interventions. The target of our investigation was to establish a correlation between diagnosis and unconstrained motion behavior in subjects without major clinical behavior impairments. We studied everyday motion behavior in 23 dyads with one partner suffering from AD dementia and one cognitively healthy partner in the subjects' home, employing ankle-mounted three-axes accelerometric sensors. We determined frequency features obtained from the signal envelopes computed by an envelope detector for the carrier band 0.5 Hz to 5 Hz. Based on these features, we employed quadratic discriminant analysis for building models discriminating between AD patients and healthy controls. After leave-one-out cross-validation, the classification accuracy of motion features reached 91% and was superior to the classification accuracy based on the Cohen-Mansfield Agitation Inventory (CMAI). Motion features were significantly correlated with MMSE and CMAI scores. Our findings suggest that changes of everyday behavior are detectable in accelerometric behavior protocols even in the absence of major clinical behavioral impairments in AD.

Detection of nonlinearity effects in structural integrity monitoring methods for offshore jacket-type structures based on principal component analysis

The detection of changes in the dynamic behavior of structures is an important issue in structural safety assessment. The development of detection methods assumes greater significance in the case of offshore platforms because the inherent problems are compounded by the harsh environment. Here, we describe an instrumented physical model for the structural health monitoring of an offshore jacket-type structure and the results of tests in several different damage scenarios. In a comparative investigation of two different methods, we discuss the difficulties of implementing damage detection techniques for complex structures, such as offshore platforms. The combined algorithm of a fuzzy logic system and a model updating method are briefly discussed, and a method based on stochastic autoregressive moving average with exogenous input is adopted for the structure. The consideration of uncertainties and the effects of nonlinearity were major objectives. So, the methods were also investigated based on the test scenarios consisting of the physical model with a geometric nonlinearity. The principal component analysis method was utilized for the detection of nonlinearity in the recorded data. The results show that the developed methods are suitable for damage classification, but the quality of the acquired signals must be considered an important factor influencing successful classification. The development of these methods may be extremely useful, as such technologies could be applied for offshore platforms in service, enabling damage detection with fewer false alarms.

Online behavior change detection in computer games

Player Modelling has been receiving much attention from the game community in the recent years. The ability to build accurate models of player behavior can be useful in many aspects of a game. One important aspect is the tracking of a player’s behavior along time, informing every time a change is perceived. This way, the game Artificial Intelligence can adapt itself to better respond to this new behavior. In order to build models of player behavior, researchers frequently resort to Machine Learning techniques. Such methods work on previously recorded game metrics representing player’s interactions with the game environment. However, if the player changes styles over time, the constructed models get out of date. In order to address this drawback, this work proposes the use of and incremental learning technique to track a player’s behavior during his/her interaction with the game environment. Our approach attempts to automatically detect the moments in time when the player changes behavior. We apply a change detection technique from the area of Data Stream Mining that is based on incremental clustering and novelty detection. We also propose three modifications to the original technique, in order to formalize change detection, improve detection rate and reduce detection delay. Simulations were performed considering data produced by the Unreal Tournament game, showing the applicability of the method to online tracking of a player’s behavior and informing whenever behavior changes occur.

Response to ethanol induced ataxia between C57BL/6J and 129X1/SvJ mouse strains using a treadmill based assay

More sensitive assays of mouse motor ataxia may provide a better understanding of the pathological profile. Treadmill gait analysis using ventral imaging allows for unhindered access to the ambulating mouse. In contrast to genetic mutations or exogenous brain injury, ethanol (EtOH) allows for the detection of dose dependent changes in motor behavior, which can be used to assess an assay's detection sensitivity. EtOH induced ataxia was assessed in C57BL/6J (B6) and 129X1/SvJ (129) mice using the DigiGait imaging system. Gait was analyzed across EtOH dosage (1.75, 2.25 and 2.75g/kg) in each strain using a linear mixed effects model. Overall, 129 mice displayed greater susceptibility to EtOH ataxia than their B6 counterparts. In both strains, hind paws exhibited greater sensitivity to EtOH dosage than fore paws. Across most variables analyzed, only a modest EtOH-induced change in motor behavior was observed in each strain with the 1.75g/kg EtOH doses failing to elicit significant change. These data indicate the ability to detect motor differences between strains, yet only moderate ability to detect change across EtOH dosage using the automated treadmill. Rotarod assays, however, were able to detect motor impairment at lower doses of EtOH. The significant, but opposite changes in paw placement with increasing EtOH doses highlight strain-specific differences in biophysical adaptations in response to acute EtOH intoxication.

Application of Nonlinear Model Updating for a Reinforced Concrete Shear Wall

The detection of changes in vibrational behavior has long been applied as a potential means of damage detection. However, existing damage diagnosis techniques are primarily limited to linear models in which the stiffness of some of the elements is reduced to represent damage. Although these methods appear to be adequate for locating and quantifying the present damage in the structure, they are not sufficient for determining future performance of the structure, which can only be determined using nonlinear models. This paper focuses on the challenging task of updating nonlinear models for reinforced concrete civil engineering structures. A nonlinear hysteretic material model is applied to accurately portray the fundamental hysteretic behavior of concrete structures. A systematic methodology to perform damage detection and, more importantly, update the nonlinear model for prediction is proposed. This new method is designed to use low-level ambient vibration data to detect changes in the modal parameters. With the acquired modal information, the damage parameters that control the nonlinear material model are updated. The final updated nonlinear model may be utilized not only to evaluate the structure’s current damage state but also to predict its future behavior. A concrete shear wall is analyzed numerically to demonstrate the proposed method.