Machine learning models for the prediction of acuity and variability of eye-positioning using features extracted from oculography

Disrupted voxel-mirrored homotopic connectivity in congenital nystagmus using resting-state fMRI

OBJECTIVES

Idiopathic congenital nystagmus (CN) is a rare eye disease that can cause early blindness (EB). CN deficits are observed most frequently with oculomotor dysfunction; however, it is still unclear what neuromechanics underly CN with EB. Based on that visual experience requires the functional integration of both hemispheres, we hypothesized that CN adolescents with EB might exhibit impaired interhemispheric synchrony. Our study aimed to investigate the interhemispheric functional connectivity alterations using voxel-mirrored homotopic connectivity (VMHC) and their relationships with clinical features in CN patients.

MATERIALS AND METHODS

This study included 21 patients with CN and EB, and 21 sighted controls (SC), who were matched for sex, age and educational level. The 3.0 T MRI scan and ocular examination were performed. The VMHC differences were examined between the two groups, and the relationships between mean VMHC values in altered brain regions and clinical variables in the CN group were evaluated by Pearson correlation analysis.

RESULTS

Compared with the SC group, the CN had increased VMHC values in the bilateral cerebellum posterior and anterior lobes/cerebellar tonsil/declive/pyramis/culmen/pons, middle frontal gyri (BA 10) and frontal eye field/superior frontal gyri (BA 6 and BA 8). No particular areas of the brain had lower VMHC values. Furthermore, no correlation with the duration of disease or blindness could be demonstrated in CN.

CONCLUSION

Our results suggest the existence of interhemispheric connectivity changes and provide further evidence for the neurological basis of CN with EB.

Automated caries detection with smartphone color photography using machine learning

Untreated caries is significant problem that affected billion people over the world. Therefore, the appropriate method and accuracy of caries detection in clinical decision-making in dental practices as well as in oral epidemiology or caries research, are required urgently. The aim of this study was to introduce a computational algorithm that can automate recognize carious lesions on tooth occlusal surfaces in smartphone images according to International Caries Detection and Assessment System (ICDAS). From a group of extracted teeth, 620 unrestored molars/premolars were photographed using smartphone. The obtained images were evaluated for caries diagnosis with the ICDAS II codes, and were labeled into three classes: "No Surface Change" (NSC); "Visually Non-Cavitated" (VNC); "Cavitated" (C). Then, a two steps detection scheme using Support Vector Machine (SVM) has been proposed: "C versus (VNC + NSC)" classification, and "VNC versus NSC" classification. The accuracy, sensitivity, and specificity of best model were 92.37%, 88.1%, and 96.6% for "C versus (VNC + NSC)," whereas they were 83.33%, 82.2%, and 66.7% for "VNC versus NSC." Although the proposed SVM system required further improvement and verification, with the data only imaged from the smartphone, it performed an auspicious potential for clinical diagnostics with reasonable accuracy and minimal cost.

Machine learning analysis: general features, requirements and cardiovascular applications

Artificial intelligence represents the science which will probably change the future of medicine by solving actually challenging issues. In this special article, the general features of machine learning are discussed. First, a background explanation regarding the division of artificial intelligence, machine learning and deep learning is given and a focus on the structure of machine learning subgroups is shown. The traditional process of a machine learning analysis is described, starting from the collection of data, across features engineering, modelling and till the validation and deployment phase. Due to the several applications of machine learning performed in literature in the last decades and the lack of some guidelines, the need of a standardization for reporting machine learning analysis results emerged. Some possible standards for reporting machine learning results are identified and discussed deeply; these are related to study population (number of subjects), repeatability of the analysis, validation, results, comparison with current practice. The way to the use of machine learning in clinical practice is open and the hope is that, with emerging technology and advanced digital and computational tools, available from hospitalization and subsequently after discharge, it will also be possible, with the help of increasingly powerful hardware, to build assistance strategies useful in clinical practice.

Toward Predicting Motion Sickness Using Virtual Reality and a Moving Platform Assessing Brain, Muscles, and Heart Signals

Motion sickness (MS) and postural control (PC) conditions are common complaints among those who passively travel. Many theories explaining a probable cause for MS have been proposed but the most prominent is the sensory conflict theory, stating that a mismatch between vestibular and visual signals causes MS. Few measurements have been made to understand and quantify the interplay between muscle activation, brain activity, and heart behavior during this condition. We introduce here a novel multimetric system called BioVRSea based on virtual reality (VR), a mechanical platform and several biomedical sensors to study the physiology associated with MS and seasickness. This study reports the results from 28 individuals: the subjects stand on the platform wearing VR goggles, a 64-channel EEG dry-electrode cap, two EMG sensors on the gastrocnemius muscles, and a sensor on the chest that captures the heart rate (HR). The virtual environment shows a boat surrounded by waves whose frequency and amplitude are synchronized with the platform movement. Three measurement protocols are performed by each subject, after each of which they answer the Motion Sickness Susceptibility Questionnaire. Nineteen parameters are extracted from the biomedical sensors (5 from EEG, 12 from EMG and, 2 from HR) and 13 from the questionnaire. Eight binary indexes are computed to quantify the symptoms combining all of them in the Motion Sickness Index (I _MS ). These parameters create the MS database composed of 83 measurements. All indexes undergo univariate statistical analysis, with EMG parameters being most significant, in contrast to EEG parameters. Machine learning (ML) gives good results in the classification of the binary indexes, finding random forest to be the best algorithm (accuracy of 74.7 for I _MS ). The feature importance analysis showed that muscle parameters are the most relevant, and for EEG analysis, beta wave results were the most important. The present work serves as the first step in identifying the key physiological factors that differentiate those who suffer from MS from those who do not using the novel BioVRSea system. Coupled with ML, BioVRSea is of value in the evaluation of PC disruptions, which are among the most disturbing and costly health conditions affecting humans.

Investigating the Adoption of Big Data Management in Healthcare in Jordan

Software developers and data scientists use and deal with big data to easily discover useful knowledge and find better solutions to improve healthcare services and patient safety. Big data analytics (BDA) is getting attention due to its role in decision-making across the healthcare field. Therefore, this article examines the adoption mechanism of big data analytics and management in healthcare organizations in Jordan. Additionally, it discusses health big data’s characteristics and the challenges, and limitations for health big data analytics and management in Jordan. This article proposes a conceptual framework that allows utilizing health big data. The proposed conceptual framework suggests a way to merge the existing health information system with the National Health Information Exchange (HIE), which might play a role in extracting insights from our massive datasets, increases the data availability and reduces waste in resources. When applying the framework, the collected data are processed to develop knowledge and support decision-making, which helps improve the health care quality for both the community and individuals by improving diagnosis, treatment, and other services.