An explainable self-attention deep neural network for detecting mild cognitive impairment using multi-inputbdigital drawing tasks

Mild cognitive impairment (MCI) is an early stage of age-inappropriate cognitive decline, which could develop into dementia – an untreatable neurodegenerative disorder. An early detection of MCI is a crucial step for timely prevention and intervention. To tackle this problem, recent studies have developed deep learning models to detect MCI and various types of dementia using data obtained from the classic clock-drawing test (CDT), a popular neuropsychological screening tool that can be easily and rapidly implemented for assessing cognitive impairments in an aging population. While these models succeed at distinguishing severe forms of dementia, it is still difficult to predict the early stage of the disease using the CDT data alone. Also, the state-of-the-art deep learning techniques still face the black-box challenges, making it questionable to implement them in the clinical setting. Here, we propose a novel deep learning modeling framework that incorporates data from multiple drawing tasks including the CDT, cube-copying, and trail-making tasks obtained from a digital platform. Using self-attention and soft-label methods, our model achieves much higher classification performance at detecting MCI compared to those of a well-established convolutional neural network model. Moreover, our model can highlight features of the MCI data that considerably deviate from those of the healthy aging population, offering accurate predictions for detecting MCI along with visual explanation that aids the interpretation of the deep learning model.

Computerised analysis of writing and drawing by essential tremor phenotype

We investigated whether computerised analysis of writing and drawing could discriminate essential tremor (ET) phenotypes according to the 2018 Consensus Statement on the Classification of Tremors. The Consensus scheme emphasises soft additional findings, mainly motor, that do not suffice to diagnose another tremor syndrome. Ten men and nine women were classified by blinded assessors according to Consensus Axis 1 definitions of ET and ET plus. Blinded scoring of tremor severity and alternating limb movement was also conducted. Twenty healthy participants acted as controls. Four writing and three drawing tasks were performed on a Wacom Intuos Pro Large digital tablet with a pressure-sensor mounted ink pen. Sixty-seven computerised measurements were obtained, comprising static (dimensional and temporal), kinematic and pen pressure features. The mean age of ET participants was 67.2±13.0 years and mean tremor duration was 21.7±19.0 years. Six were classified as ET, five had one plus feature and eight had two plus features. The computerised analysis could predict the presence and number of ET plus features. Measures of acceleration and variation of pen pressure performed strongly to separate ET phenotypes (p<0.05). Plus features were associated with higher scores on the Fahn-Tolosa-Marin Tremor Rating Scale (p=0.001) and it appeared that ET groups were mainly being separated according to severity of tremor and by compensatory manoeuvres used by participants with more severe tremor. There were, in addition, a small number of negative kinematic correlations suggesting some slowness with ET plus. Abnormal repetitive limb movement was also correlated with tremor severity (R=0.57) by clinical grading. Critics of the Consensus Statement have drawn attention to weaknesses of the ET plus concept in relation to duration and severity of ET. This classification of ET may be too biased towards tremor severity to assist in distinguishing underlying biological differences by clinical measurement.

The Visuospatial and Sensorimotor Functions of Posterior Parietal Cortex in Drawing Tasks: A Review

Drawing is a comprehensive skill that primarily involves visuospatial processing, eye-hand coordination, and other higher-order cognitive functions. Various drawing tasks are widely used to assess brain function. The neuropsychological basis of drawing is extremely sophisticated. Previous work has addressed the critical role of the posterior parietal cortex (PPC) in drawing, but the specific functions of the PPC in drawing remain unclear. Functional magnetic resonance imaging and electrophysiological studies found that drawing activates the PPC. Lesion-symptom mapping studies have shown an association between PPC injury and drawing deficits in patients with global and focal cerebral pathology. These findings depicted a core framework of the fronto-parietal network in drawing tasks. Here, we review neuroimaging and electrophysiological studies applying drawing paradigms and discuss the specific functions of the PPC in visuospatial and sensorimotor aspects. Ultimately, we proposed a hypothetical model based on the dorsal stream. It demonstrates the organization of a PPC-centered network for drawing and provides systematic insights into drawing for future neuropsychological research.

Development of Laterality and Bimanual Interference of Fine Motor Movements in Childhood and Adolescence

Drawing and handwriting are fine motor skills acquired during childhood. We analyzed the development of laterality by comparing the performance of the dominant with the nondominant hand and the effect of bimanual interference in kinematic hand movement parameters (speed, automation, variability, and pressure). Healthy subjects (n = 187, 6–18 years) performed drawing tasks with both hands on a digitizing tablet followed by performance in the presence of an interfering task of the nondominant hand. Age correlated positively with speed, automation, and pressure, and negatively with variability for both hands. As task complexity increased, differences between both hands were less pronounced. Playing an instrument had a positive effect on the nondominant hand. Speed and automation showed a strong association with lateralization. Bimanual interference was associated with an increase of speed and variability. Maturation of hand laterality and the extent of bimanual interference in fine motor tasks are age-dependent processes.

Screening of Parkinson’s Disease Using Geometric Features Extracted from Spiral Drawings

Conventional means of Parkinson’s Disease (PD) screening rely on qualitative tests typically administered by trained neurologists. Tablet technologies that enable data collection during handwriting and drawing tasks may provide low-cost, portable, and instantaneous quantitative methods for high-throughput PD screening. However, past efforts to use data from tablet-based drawing processes to distinguish between PD and control populations have demonstrated only moderate classification ability. Focusing on digitized drawings of Archimedean spirals, the present study utilized data from the open-access ParkinsonHW dataset to improve existing PD drawing diagnostic pipelines. Random forest classifiers were constructed using previously documented features and highly-predictive, newly-proposed features that leverage the many unique mathematical characteristics of the Archimedean spiral. This approach yielded an AUC of 0.999 on the particular dataset we tested on, and more importantly identified interpretable features with good promise for generalization across diverse patient cohorts. It demonstrated the potency of mathematical relationships inherent to the drawing shape and the usefulness of sparse feature sets and simple models, which further enhance interpretability, in the face of limited sample size. The results of this study also inform suggestions for future drawing task design and data analytics (feature extraction, shape selection, task diversity, drawing templates, and data sharing).

Parkinson’s disease diagnosis using convolutional neural networks and figure-copying tasks

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that causes abnormal movements and an array of other symptoms. An accurate PD diagnosis can be a challenging task as the signs and symptoms, particularly at an early stage, can be similar to other medical conditions or the physiological changes of normal ageing. This work aims to contribute to the PD diagnosis process by using a convolutional neural network, a type of deep neural network architecture, to differentiate between healthy controls and PD patients. Our approach focuses on discovering deviations in patient’s movements with the use of drawing tasks. In addition, this work explores which of two drawing tasks, wire cube or spiral pentagon, are more effective in the discrimination process. With $$93.5\%$$ 93.5 % accuracy, our convolutional classifier, trained with images of the pentagon drawing task and augmentation techniques, can be used as an objective method to discriminate PD from healthy controls. Our compact model has the potential to be developed into an offline real-time automated single-task diagnostic tool, which can be easily deployed within a clinical setting.

Precise Extraction of Buildings from High-Resolution Remote-Sensing Images Based on Semantic Edges and Segmentation

Building extraction is a basic task in the field of remote sensing, and it has also been a popular research topic in the past decade. However, the shape of the semantic polygon generated by semantic segmentation is irregular and does not match the actual building boundary. The boundary of buildings generated by semantic edge detection has difficulty ensuring continuity and integrity. Due to the aforementioned problems, we cannot directly apply the results in many drawing tasks and engineering applications. In this paper, we propose a novel convolutional neural network (CNN) model based on multitask learning, Dense D-LinkNet (DDLNet), which adopts full-scale skip connections and edge guidance module to ensure the effective combination of low-level information and high-level information. DDLNet has good adaptability to both semantic segmentation tasks and edge detection tasks. Moreover, we propose a universal postprocessing method that integrates semantic edges and semantic polygons. It can solve the aforementioned problems and more accurately locate buildings, especially building boundaries. The experimental results show that DDLNet achieves great improvements compared with other edge detection and semantic segmentation networks. Our postprocessing method is effective and universal.

Automated scoring for a Tablet-based Rey Figure copy task differentiates constructional, organisational, and motor abilities

Accuracy in copying a figure is one of the most sensitive measures of visuo-constructional ability. However, drawing tasks also involve other cognitive and motor abilities, which may influence the final graphic produced. Nevertheless, these aspects are not taken into account in conventional scoring methodologies. In this study, we have implemented a novel Tablet-based assessment, acquiring data and information for the entire execution of the Rey Complex Figure copy task (T-RCF). This system extracts 12 indices capturing various dimensions of drawing abilities. We have also analysed the structure of relationships between these indices and provided insights into the constructs that they capture. 102 healthy adults completed the T-RCF. A subgroup of 35 participants also completed a paper-and-pencil drawing battery from which constructional, procedural, and motor measures were obtained. Principal component analysis of the T-RCF indices was performed, identifying spatial, procedural and kinematic components as distinct dimensions of drawing execution. Accordingly, a composite score for each dimension was determined. Correlational analyses provided indications of their validity by showing that spatial, procedural, and kinematic scores were associated with constructional, organisational and motor measures of drawing, respectively. Importantly, final copy accuracy was found to be associated with all of these aspects of drawing. In conclusion, copying complex figures entails an interplay of multiple functions. T-RCF provides a unique opportunity to analyse the entire drawing process and to extract scores for three critical dimensions of drawing execution.

Navigating in Virtual Environments: Does a Map or a Map-Based Description Presented Beforehand Help?

Background. One of the aims of research in spatial cognition is to examine the factors capable of optimizing environment learning from navigation, which can be examined using a virtual environment (VE). Different learning conditions can play an important part. Aim. This study examined the benefits of presenting configured information (layout with elements arranged in it) using a map or verbal description before a learner navigates in a new environment. Method. Ninety participants were assigned to three learning groups of 30 individuals (15 males and 15 females). Before participants navigated in a VE, one group was shown a map of the environment (“map before navigation”), a second group read a map-like description of the environment (“description before navigation”), and a third group started navigating without any prior input (“only navigation”). Participants then learned a path in a VE (presented as if they were driving a car). Their recall was subsequently tested using three types of task: (i) route retracing; (ii) pointing; (iii) path drawing. Several measures were administered to assess participants’ individual visuospatial and verbal factors. Results. There were no differences between the three groups in route retracing. The “map before navigation” group performed better than the “only navigation” group in both the pointing and the path drawing tasks, however, and also outperformed the “description before navigation” group in the path drawing task. Some relations emerged between participants’ individual difference factors and their recall performance. Conclusions. In learning from navigation, seeing a map beforehand benefits learning accuracy. Recall performance is also supported, at least in part, by individual visuospatial and verbal factors.

Studying the coherence of studentsʼ portrayed representations of the atomic structure - Connections with conceptions and misconceptions

The present study investigated the association of students’ fundamental ideas and misconceptions about ontological features of atom identity and behavior with the formation of their portrayed representations of the atomic structure. Participants (n = 421) were secondary education students in the eighth, tenth, and twelfth grades. Students’ portrayed representations of the atomic structure were accessed through drawing tasks, while their understanding of the ontological features of atom was measured through a specially designed questionnaire. Latent Class Analysis (LCA), a psychometric method, was applied to the elementary features of the portrayed representations to classify them and test the potential coherence of their representations regarding atomic structure. The LCA revealed three latent classes, which showed a relative coherence in three of the anticipated models, “Particle model,” “Nuclear model,” and “Bohrʼs model.” Moreover, students’ conceptions and misconception about the ontological features of atom were used as covariates in the LCA and their effects on the above-mentioned class-memberships were estimated. Results indicated a significant effect of students’ conceptions of the atomic ontological features on their portrayed representations of the atomic structure. Implications for theory and practice are discussed.