Search

Knowing how students engage with their materials helps educators design methodologies to enhance the learning process. To help teachers in determining students’ average learning style profiles and analyzing potential connections between learning styles and academic performance, new freely accessible software, integrated with Moodle, has been developed in this work. A comprehensive case study involving one thousand students across various subjects and degrees from several branches of knowledge in different years has been conducted using this tool. The analysis reveals low participation in the survey, with only 6 out of 19 courses having a participation rate that surpasses 50%), where respondents generally achieved higher grades compared to non-participants (on a grade scale from 0 to 10, participants scored on average one point higher than those who did not complete the questionnaire), suggesting that they were more motivated. Higher-level courses have higher participation rates, but overall participation has declined over the years. Despite slight differences in learning styles between courses, the general pattern remains similar. The correlation between learning styles and grades, analyzed through linear regression, shows that this model performs better when both learning styles and grade marks are distributed by score rather than by level.

Medical image segmentation is crucial for diagnostics and treatment planning, yet traditional methods often struggle with the variability of real-world clinical data. Deep learning models, like the Segment Anything Model (SAM), have been proposed as a powerful tool that helps to delimit regions using a prompt. This work proposes a methodology to improve the quality of the segmentation by integrating test-time augmentation (TTA) with the SAM for medical applications (MedSAM) by using random circular shifts, addressing challenges such as misalignments and imaging variability. The method generates several input variations during inference that are combined after, improving robustness and segmentation accuracy without requiring retraining. Evaluated across diverse computed tomography (CT) datasets, including Medical Segmentation Decathlon (MSD), KiTS, and COVID-19-20, the proposed method demonstrated consistent improvements in Dice Similarity Coefficient (DSC) and Normalized Surface Dice (NSD) metrics. The highest performances were 93.6% DSC and 97% NSD. Notably, it achieved superior boundary precision and surface alignment in complex regions like the pancreas and colon, outperforming baseline models, including MedSAM and DeepLabv3+. The approach is computationally feasible, leveraging a balance of augmentation intensity and segmentation accuracy.

Invasive Coronary Angiography (ICA) images are considered the gold standard for assessing the state of the coronary arteries. Deep learning classification methods are widely used and well-developed in different areas where medical imaging evaluation has an essential impact due to the development of computer-aided diagnosis systems that can support physicians in their clinical procedures. In this paper, a new performance analysis of deep learning methods for binary ICA classification with different lesion degrees is reported. To reach this goal, an annotated dataset of ICA images containing the ground truth, the location of lesions, and seven possible severity degrees ranging between 0% and 100% was employed. The ICA images were divided into “lesion” or “non-lesion” patches. We aim to study how binary classification performance is affected by the different lesion degrees considered in the positive class. Therefore, five Convolutional Neural Network architectures – DenseNet-201, MobileNet-V2, NasNet-Mobile, ResNet-18, and ResNet-50 – were trained with different input images where different lesion degree ranges were gradually incorporated until considering the seven lesion degrees. Besides, four types of experiments with and without data augmentation were designed, whose F-measure and Area Under Curve (AUC) were computed. Reported results achieved an F-measure and AUC of 92.7% and 98.1%, respectively. However, lesion classification is highly affected by the degree of the lesion intended to be classified, with 15% less accuracy when < 99% lesion patches are present.

The prognostic impact of non-obstructive coronary artery disease (CAD) remains controversial. Therefore, the objective of this study is to assess the long-term prognostic significance of non-obstructive CAD using machine learning models. We designed a multicenter retrospective, longitudinal, and observational study that included 3265 patients classified into three groups: 1426 patients with lesions < 20%, 643 patients with non-obstructive CAD (lesions 20–50%), and 1196 patients with obstructive CAD (lesions > 70%). A composite cardiovascular event (acute myocardial infarction, stroke, hospitalization due to heart failure, or cardiovascular-related death) was assessed after a mean follow-up of 43 months. To achieve this, various machine learning models were constructed. The model with the highest accuracy was selected to perform a Shapley Additive Explanations (SHAP) analysis, revealing the contribution of different variables in predicting an event. The SHAP analysis suggested that the percentage of coronary lesion was the most significant predictor of cardiovascular events. None of the models demonstrated adequate capability in predicting the event, showing only a good predictive ability for the absence of an endpoint. In conclusions, this study demonstrates how machine learning techniques may facilitate the development of high-performing models for predicting long-term cardiovascular events in patients undergoing coronary angiography.

The COVID-19 pandemic has caused a challenge for our society due to the post-acute sequelae of the disease. Persistent symptoms and long-term multiorgan complications, known as post-acute COVID-19 syndrome, can occur beyond 4 weeks from the onset of the COVID-19 infection. Postural orthostatic tachycardia syndrome (POTS) is considered a variety of dysautonomia, which is characterized by chronic symptoms that occur with standing and a sustained increase in heart rate, without orthostatic hypotension. POTS can lead to debilitating symptoms, significant disability, and impaired quality of life. In this narrative review, the etiopathogenic basis, epidemiology, clinical manifestations, diagnosis, treatment, prognosis, and socioeconomic impact of POTS, as well as other related dysautonomic disorders, after COVID-19 infection and SARS-CoV-2 postvaccination, were discussed. After a search conducted in March 2023, a total of 89 relevant articles were selected from the PubMed, Google Scholar, and Web of Science databases. The review highlights the importance of recognizing and managing POTS after COVID-19 infection and vaccination, and the approach to autonomic disorders should be known by all specialists in different medical areas. The diagnosis of POTS requires a comprehensive clinical assessment, including a detailed medical history, physical examination, orthostatic vital signs, and autonomic function tests. The treatment of POTS after COVID-19 infection or vaccination is mainly focused on lifestyle modifications, such as increased fluid and salt intake, exercise, and graduated compression stockings. Pharmacotherapy, such as beta-blockers, fludrocortisone, midodrine, and ivabradine, may also be used in selected cases. Further research is needed to understand the underlying mechanisms, risk factors, and optimal treatment strategies for this complication.

This paper investigates the distribution characteristics of Fourier, discrete cosine, and discrete sine transform coefficients in T1 MRI images. This paper reveals their adherence to Benford’s law, characterized by a logarithmic distribution of first digits. The impact of Rician noise on the first digit distribution is examined, which causes deviations from the ideal distribution. A novel methodology is proposed for noise level estimation, employing metrics such as the Bhattacharyya distance, Kullback–Leibler divergence, total variation distance, Hellinger distance, and Jensen–Shannon divergence. Supervised learning techniques utilize these metrics as regressors. Evaluations on MRI scans from several datasets coming from a wide range of different acquisition devices of 1.5 T and 3 T, comprising hundreds of patients, validate the adherence of noiseless T1 MRI frequency domain coefficients to Benford’s law. Through rigorous experimentation, our methodology has demonstrated competitiveness with established noise estimation techniques, even surpassing them in numerous conducted experiments. This research empirically supports the application of Benford’s law in transforms, offering a reliable approach for noise estimation in denoising algorithms and advancing image quality assessment.

Skin lesions are caused due to multiple factors, like allergies, infections, exposition to the sun, etc. These skin diseases have become a challenge in medical diagnosis due to visual similarities, where image classification is an essential task to achieve an adequate diagnostic of different lesions. Melanoma is one of the best-known types of skin lesions due to the vast majority of skin cancer deaths. In this work, we propose an ensemble of improved convolutional neural networks combined with a test-time regularly spaced shifting technique for skin lesion classification. The shifting technique builds several versions of the test input image, which are shifted by displacement vectors that lie on a regular lattice in the plane of possible shifts. These shifted versions of the test image are subsequently passed on to each of the classifiers of an ensemble. Finally, all the outputs from the classifiers are combined to yield the final result. Experiment results show a significant improvement on the well-known HAM10000 dataset in terms of accuracy and F-score. In particular, it is demonstrated that our combination of ensembles with test-time regularly spaced shifting yields better performance than any of the two methods when applied alone.

Remote health monitoring plays a significant role in research areas related to medicine, neurology, rehabilitation, and robotic systems. These applications include Human Activity Recognition (HAR) using wearable sensors, signal processing, mathematical methods, and machine learning to improve the accuracy of remote health monitoring systems. To improve the detection and accuracy of human activity recognition, we create a novel method to reduce the complexities of extracting features using the HuGaDB dataset. Our model extracts power spectra; due to the high dimensionality of features, sliding windows techniques are used to determine frequency bandwidth automatically, where an improved QRS algorithm selects the first dominant spectrum amplitude. In addition, the bandwidth algorithm has been used to reduce the dimensionality of data, remove redundant dimensions, and improve feature extraction. In this work, we have considered widely used machine learning classifiers. Our proposed method was evaluated using the accelerometer angles spectrum installed in six parts of the body and then reducing the bandwidth to know the evolution. Our approach attains an accuracy rate of 95.1% in the HuGaDB dataset with 70% of bandwidth, outperforming others in the human activity recognition accuracy.

Donepezil (DP), a cognitive-enhancing drug targeting the cholinergic system, combined with massed sentence repetition training augmented and speeded up recovery of speech production deficits in patients with chronic conduction aphasia and extensive left hemisphere infarctions (Berthier et al., 2014). Nevertheless, a still unsettled question is whether such improvements correlate with restorative structural changes in gray matter and white matter pathways mediating speech production. In the present study, we used pharmacological magnetic resonance imaging to study treatment-induced brain changes in gray matter and white matter tracts in a right-handed male with chronic conduction aphasia and a right subcortical lesion (crossed aphasia). A single-patient, open-label multiple-baseline design incorporating two different treatments and two post-treatment evaluations was used. The patient received an initial dose of DP (5 mg/day) which was maintained during 4 weeks and then titrated up to 10 mg/day and administered alone (without aphasia therapy) during 8 weeks (Endpoint 1). Thereafter, the drug was combined with an audiovisual repetition-imitation therapy (Look-Listen-Repeat, LLR) during 3 months (Endpoint 2). Language evaluations, diffusion weighted imaging (DWI), and voxel-based morphometry (VBM) were performed at baseline and at both endpoints in JAM and once in 21 healthy control males. Treatment with DP alone and combined with LLR therapy induced marked improvement in aphasia and communication deficits as well as in selected measures of connected speech production, and phrase repetition. The obtained gains in speech production remained well-above baseline scores even 4 months after ending combined therapy. Longitudinal DWI showed structural plasticity in the right frontal aslant tract and direct segment of the arcuate fasciculus with both interventions. VBM revealed no structural changes in other white matter tracts nor in cortical areas linked by these tracts. In conclusion, cholinergic potentiation alone and combined with a model-based aphasia therapy improved language deficits by promoting structural plastic changes in right white matter tracts.

Foreign accent syndrome (FAS) is a speech disorder that is defined by the emergence of a peculiar manner of articulation and intonation which is perceived as foreign. In most cases of acquired FAS (AFAS) the new accent is secondary to small focal lesions involving components of the bilaterally distributed neural network for speech production. In the past few years FAS has also been described in different psychiatric conditions (conversion disorder, bipolar disorder, schizophrenia) as well as in developmental disorders (specific language impairment, apraxia of speech). In the present study, two adult males, one with atypical phonetic production and the other one with cluttering, reported having developmental FAS (DFAS) since their adolescence. Perceptual analysis by naïve judges could not confirm the presence of foreign accent, possibly due to the mildness of the speech disorder. However, detailed linguistic analysis provided evidence of prosodic and segmental errors previously reported in AFAS cases. Cognitive testing showed reduced communication in activities of daily living and mild deficits related to psychiatric disorders. Psychiatric evaluation revealed long-lasting internalizing disorders (neuroticism, anxiety, obsessive-compulsive disorder, social phobia, depression, alexithymia, hopelessness, and apathy) in both subjects. Diffusion tensor imaging (DTI) data from each subject with DFAS were compared with data from a group of 21 age- and gender-matched healthy control subjects. Diffusion parameters (MD, AD, and RD) in predefined regions of interest showed changes of white matter microstructure in regions previously related with AFAS and psychiatric disorders. In conclusion, the present findings militate against the possibility that these two subjects have FAS of psychogenic origin. Rather, our findings provide evidence that mild DFAS occurring in the context of subtle, yet persistent, developmental speech disorders may be associated with structural brain anomalies. We suggest that the simultaneous involvement of speech and emotion regulation networks might result from disrupted neural organization during development, or compensatory or maladaptive plasticity. Future studies are required to examine whether the interplay between biological trait-like diathesis (shyness, neuroticism) and the stressful experience of living with mild DFAS lead to the development of internalizing psychiatric disorders.