Examining Cognitive Factors for Alzheimer's Disease Progression Using Computational Intelligence.

Prognosis of Alzheimer's disease (AD) progression has been recognized as a challenging problem due to the massive numbers of cognitive, and pathological features recorded for patients and controls. While there have been many studies investigated the diagnosis of dementia using pathological characteristics, predicting the advancement of the disease using cognitive elements has not been heavily studied particularly using technologies like artificial intelligence and machine learning. This research aims at evaluating items of the Alzheimer's Disease Assessment Scale-Cognitive 13 (ADAS-Cog-13) test to determine key cognitive items that influence the progression of AD. A methodology that consists of machine learning and feature selection (FS) techniques was designed, implemented, and then tested against real data observations (cases and controls) of the Alzheimer's Disease Neuroimaging Initiative (ADNI) repository with a narrow scope on cognitive items of the ADAS-Cog-13 test. Results obtained by ten-fold cross validation and using dissimilar classification and FS techniques revealed that the decision tree algorithm produced classification models with the best performing results from the cognitive items. For ADAS-Cog-13 test, memory and learning features including word recall, delayed word recall and word recognition were the key items pinpointing to AD advancement. When these three cognitive items are processed excluding demographics by C4.5 algorithm the models derived showed 82.90% accuracy, 87.60% sensitivity and 78.20% specificity.

Dementia medical screening using mobile applications: A systematic review with a new mapping model.

Early detection is the key to successfully tackling dementia, a neurocognitive condition common among the elderly. Therefore, screening using technological platforms such as mobile applications (apps) may provide an important opportunity to speed up the diagnosis process and improve accessibility. Due to the lack of research into dementia diagnosis and screening tools based on mobile apps, this systematic review aims to identify the available mobile-based dementia and mild cognitive impairment (MCI) apps using specific inclusion and exclusion criteria. More importantly, we critically analyse these tools in terms of their comprehensiveness, validity, performance, and the use of artificial intelligence (AI) techniques. The research findings suggest diagnosticians in a clinical setting use dementia screening apps such as ALZ and CognitiveExams since they cover most of the domains for the diagnosis of neurocognitive disorders. Further, apps such as Cognity and ACE-Mobile have great potential as they use machine learning (ML) and AI techniques, thus improving the accuracy of the outcome and the efficiency of the screening process. Lastly, there was overlapping among the dementia screening apps in terms of activities and questions they contain therefore mapping these apps to the designated cognitive domains is a challenging task, which has been done in this research.

A new machine learning model based on induction of rules for autism detection.

Autism spectrum disorder is a developmental disorder that describes certain challenges associated with communication (verbal and non-verbal), social skills, and repetitive behaviors. Typically, autism spectrum disorder is diagnosed in a clinical environment by licensed specialists using procedures which can be lengthy and cost-ineffective. Therefore, scholars in the medical, psychology, and applied behavioral science fields have in recent decades developed screening methods such as the Autism Spectrum Quotient and Modified Checklist for Autism in Toddlers for diagnosing autism and other pervasive developmental disorders. The accuracy and efficiency of these screening methods rely primarily on the experience and knowledge of the user, as well as the items designed in the screening method. One promising direction to improve the accuracy and efficiency of autism spectrum disorder detection is to build classification systems using intelligent technologies such as machine learning. Machine learning offers advanced techniques that construct automated classifiers that can be exploited by users and clinicians to significantly improve sensitivity, specificity, accuracy, and efficiency in diagnostic discovery. This article proposes a new machine learning method called Rules-Machine Learning that not only detects autistic traits of cases and controls but also offers users knowledge bases (rules) that can be utilized by domain experts in understanding the reasons behind the classification. Empirical results on three data sets related to children, adolescents, and adults show that Rules-Machine Learning offers classifiers with higher predictive accuracy, sensitivity, harmonic mean, and specificity than those of other machine learning approaches such as Boosting, Bagging, decision trees, and rule induction.

A Mobile-Based Screening System for Data Analyses of Early Dementia Traits Detection.

Existing early detection methods that deal with the pre-diagnosis of dementia have been criticised as not being comprehensive as they do not measure certain cognitive functioning domains besides being inaccessible. A more realistic approach is to develop a comprehensive outcome that includes cognitive functioning of dementia, as this will offer a robust and unbiased outcome for an individual. In this research, a mobile screening application for dementia traits called DementiaTest is proposed, which adopts the gold standard assessment criteria of Diagnostic and Statistical Manual of Mental Disorders (DSM-V). DementiaTest is implemented and tested on the Android and IOS stores. More importantly, it collects data from cases and controls using an easy, interactive, and accessible platform. It provides patients and their family with quick pre-diagnostic reports using certain cognitive functioning indicators; these can be utilized by general practitioners (GPs) for referrals for further assessment in cases of positive outcomes. The data gathered using the new application can be analysed using Artificial Intelligence methods to evaluate the performance of the screening to pinpoint early signs of the dementia.

Early Autism Screening: A Comprehensive Review.

Autistic spectrum disorder (ASD) refers to a neurodevelopmental condition associated with verbal and nonverbal communication, social interactions, and behavioural complications that is becoming increasingly common in many parts of the globe. Identifying individuals on the spectrum has remained a lengthy process for the past few decades due to the fact that some individuals diagnosed with ASD exhibit exceptional skills in areas such as mathematics, arts, and music among others. To improve the accuracy and reliability of autism diagnoses, many scholars have developed pre-diagnosis screening methods to help identify autistic behaviours at an early stage, speed up the clinical diagnosis referral process, and improve the understanding of ASD for the different stakeholders involved, such as parents, caregivers, teachers, and family members. However, the functionality and reliability of those screening tools vary according to different research studies and some have remained questionable. This study evaluates and critically analyses 37 different ASD screening tools in order to identify possible areas that need to be addressed through further development and innovation. More importantly, different criteria associated with existing screening tools, such as accessibility, the fulfilment of Diagnostic and Statistical Manual of Mental Disorders (DSM-5) specifications, comprehensibility among the target audience, performance (specifically sensitivity, specificity, and accuracy), web and mobile availability, and popularity have been investigated.

A machine learning autism classification based on logistic regression analysis.

Autistic Spectrum Disorder (ASD) is a neurodevelopmental condition associated with significant healthcare costs; early diagnosis could substantially reduce these. The economic impact of autism reveals an urgent need for the development of easily implemented and effective screening methods. Therefore, time-efficient ASD screening is imperative to help health professionals and to inform individuals whether they should pursue formal clinical diagnosis. Presently, very limited autism datasets associated with screening are available and most of them are genetic in nature. We propose new machine learning framework related to autism screening of adults and adolescents that contain vital features and perform predictive analysis using logistic regression to reveal important information related to autism screening. We also perform an in-depth feature analysis on the two datasets using information gain (IG) and Chi square testing (CHI) to determine the influential features that can be utilized in screening for autism. Results obtained reveal that machine learning technology was able to generate classification systems that have acceptable performance in terms of sensitivity, specificity and accuracy among others.

On the Relation Between Marr's Levels: A Response to Blokpoel (2017).

Blokpoel reminds us of the importance of consistency of function across Marr's levels, but we argue that the approach to ensuring consistency that he advocates-a strict relation through exact implementation of the higher level function at the lower level-is unnecessarily restrictive. We show that it forces overcomplication of the computational level (by requiring it to incorporate concerns from lower levels) and results in the sacrifice of the distinct responsibilities associated with each level. We propose an alternative, no less rigorous, potential characterization of the relation between levels.

Effectiveness of Immersive Virtual Reality in Surgical Training-A Randomized Control Trial.

PURPOSE: Surgical training methods are evolving with the technological advancements, including the application of virtual reality (VR) and augmented reality. However, 28 to 40% of novice residents are not confident in performing a major surgical procedure. VR surgery, an immersive VR (iVR) experience, was developed using Oculus Rift and Leap Motion devices (Leap Motion, Inc, San Francisco, CA) to address this challenge. Our iVR is a multisensory, holistic surgical training application that demonstrates a maxillofacial surgical technique, the Le Fort I osteotomy. The main objective of the present study was to evaluate the effect of using VR surgery on the self-confidence and knowledge of surgical residents. MATERIALS AND METHODS: A multisite, single-blind, parallel, randomized controlled trial (RCT) was performed. The participants were novice surgical residents with limited experience in performing the Le Fort I osteotomy. The primary outcome measures were the self-assessment scores of trainee confidence using a Likert scale and an objective assessment of the cognitive skills. Ninety-five residents from 7 dental schools were included in the RCT. The participants were randomly divided into a study group of 51 residents and a control group of 44. Participants in the study group used the VR surgery application on an Oculus Rift with Leap Motion device. The control group participants used similar content in a standard PowerPoint presentation on a laptop. Repeated measures multivariate analysis of variance was applied to the data to assess the overall effect of the intervention on the confidence of the residents. RESULTS: The study group participants showed significantly greater perceived self-confidence levels compared with those in the control group (P = .034; α = 0.05). Novices in the first year of their training showed the greatest improvement in their confidence compared with those in their second and third year. CONCLUSIONS: iVR experiences improve the knowledge and self-confidence of the surgical residents.

Expert interpretation of bar and line graphs: the role of graphicacy in reducing the effect of graph format.

The distinction between informational and computational equivalence of representations, first articulated by Larkin and Simon (1987) has been a fundamental principle in the analysis of diagrammatic reasoning which has been supported empirically on numerous occasions. We present an experiment that investigates this principle in relation to the performance of expert graph users of 2 × 2 "interaction" bar and line graphs. The study sought to determine whether expert interpretation is affected by graph format in the same way that novice interpretations are. The findings revealed that, unlike novices-and contrary to the assumptions of several graph comprehension models-experts' performance was the same for both graph formats, with their interpretation of bar graphs being no worse than that for line graphs. We discuss the implications of the study for guidelines for presenting such data and for models of expert graph comprehension.

Thirty Years After Marr's Vision: Levels of Analysis in Cognitive Science.

Thirty years after the publication of Marr's seminal book Vision (Marr, 1982) the papers in this topic consider the contemporary status of his influential conception of three distinct levels of analysis for information-processing systems, and in particular the role of the algorithmic and representational level with its cognitive-level concepts. This level has (either implicitly or explicitly) been downplayed or eliminated both by reductionist neuroscience approaches from below that seek to account for behavior from the implementation level and by Bayesian approaches from above that seek to account for behavior in purely computational-level terms.

Beyond single-level accounts: the role of cognitive architectures in cognitive scientific explanation.

We consider approaches to explanation within the cognitive sciences that begin with Marr's computational level (e.g., purely Bayesian accounts of cognitive phenomena) or Marr's implementational level (e.g., reductionist accounts of cognitive phenomena based only on neural-level evidence) and argue that each is subject to fundamental limitations which impair their ability to provide adequate explanations of cognitive phenomena. For this reason, it is argued, explanation cannot proceed at either level without tight coupling to the algorithmic and representation level. Even at this level, however, we argue that additional constraints relating to the decomposition of the cognitive system into a set of interacting subfunctions (i.e., a cognitive architecture) are required. Integrated cognitive architectures that permit abstract specification of the functions of components and that make contact with the neural level provide a powerful bridge for linking the algorithmic and representational level to both the computational level and the implementational level.

The effect of Gestalt laws of perceptual organization on the comprehension of three-variable bar and line graphs.

OBJECTIVE: We report three experiments investigating the ability of undergraduate college students to comprehend 2 x 2 "interaction" graphs from two-way factorial research designs. BACKGROUND: Factorial research designs are an invaluable research tool widely used in all branches of the natural and social sciences, and the teaching of such designs lies at the core of many college curricula. Such data can be represented in bar or line graph form. Previous studies have shown, however, that people interpret these two graphical forms differently. METHOD: In Experiment 1, participants were required to interpret interaction data in either bar or line graphs while thinking aloud. Verbal protocol analysis revealed that line graph users were significantly more likely to misinterpret the data or fail to interpret the graph altogether. RESULTS: The patterns of errors line graph users made were interpreted as arising from the operation of Gestalt principles of perceptual organization, and this interpretation was used to develop two modified versions of the line graph, which were then tested in two further experiments. One of the modifications resulted in a significant improvement in performance. CONCLUSION: Results of the three experiments support the proposed explanation and demonstrate the effects (both positive and negative) of Gestalt principles of perceptual organization on graph comprehension. APPLICATION: We propose that our new design provides a more balanced representation of the data than the standard line graph for nonexpert users to comprehend the full range of relationships in two-way factorial research designs and may therefore be considered a more appropriate representation for use in educational and other nonexpert contexts.

The effect of emergent features on judgments of quantity in configural and separable displays.

Two experiments investigated effects of emergent features on perceptual judgments of comparative magnitude in three diagrammatic representations: kiviat charts, bar graphs, and line graphs. Experiment 1 required participants to compare individual values; whereas in Experiment 2 participants had to integrate several values to produce a global comparison. In Experiment 1, emergent features of the diagrams resulted in significant distortions of magnitude judgments, each related to a common geometric illusion. Emergent features are also widely believed to underlie the general superiority of configural displays, such as kiviat charts, for tasks requiring the integration of information. Experiment 2 tested the extent of this benefit using diagrams with a wide range of values. Contrary to the results of previous studies, the configural display produced the poorest performance compared to the more separable displays. Moreover, the pattern of responses suggests that kiviat users switched from an integration strategy to a sequential one depending on the shape of the diagram. The experiments demonstrate the powerful interaction between emergent visual properties and cognition and reveal limits to the benefits of configural displays for integration tasks.

Sorting preference in children with autism: the dominance of concrete features.

The current study investigates preference to sort objects on the basis of either concrete or abstract features in children with and without autism. Participants were asked to sort a set of books into two groups that could be differentiated according to concrete (color, size) or abstract criteria (category membership: sports/games). The results showed that those with autism, unlike controls, were significantly more likely to sort according to a concrete criterion. In a further phase of testing, those with autism still did not sort according to abstract criteria, even when this was the only basis for sorting systematically. The findings are interpreted as evidence for a preference in autism to process concrete over abstract features of stimuli.

Modeling the effect of task and graphical representation on response latency in a graph reading task.

We report an investigation into the processes involved in a common graph-reading task using two types of Cartesian graph. We describe an experiment and eye movement study, the results of which show that optimal scan paths assumed in the task analysis approximate the detailed sequences of saccades made by individuals. The research demonstrates the computational inequivalence of two sets of informationally equivalent graphs and illustrates how the computational advantages of a representation outweigh factors such as user unfamiliarity. We describe two models, using the ACT rational perceptual motor (ACT-R/PM) cognitive architecture, that replicate the pattern of observed response latencies and the complex scan paths revealed by the eye movement study. Finally, we outline three guidelines for designers of visual displays: Designers should (a) consider how different quantities are encoded within any chosen representational format, (b) consider the full range of alternative varieties of a given task, and (c) balance the cost of familiarization with the computational advantages of less familiar representations. Actual or potential applications of this research include informing the design and selection of appropriate visual displays and illustrating the practice and utility of task analysis, eye tracking, and cognitive modeling for understanding interactive tasks with external representations.