Introduction: Migraine with aura (MwA) is a neurological condition manifested in moderate to severe headaches associated with transient visual and somatosensory symptoms, as well as higher cortical dysfunctions. Considering that about 5% of the world’s population suffers from this condition and manifestation could be abundant and characterized by various symptoms, it is of great importance to focus on finding new and advanced techniques for the detection of different phenotypes, which in turn, can allow better diagnosis, classification, and biomarker validation, resulting in tailored treatments of MwA patients.Methods: This research aimed to test different machine learning techniques to distinguish healthy people from those suffering from MwA, as well as people with simple MwA and those experiencing complex MwA. Magnetic resonance imaging (MRI) post-processed data (cortical thickness, cortical surface area, cortical volume, cortical mean Gaussian curvature, and cortical folding index) was collected from 78 subjects [46 MwA patients (22 simple MwA and 24 complex MwA) and 32 healthy controls] with 340 different features used for the algorithm training.Results: The results show that an algorithm based on post-processed MRI data yields a high classification accuracy (97%) of MwA patients and precise distinction between simple MwA and complex MwA with an accuracy of 98%. Additionally, the sets of features relevant to the classification were identified. The feature importance ranking indicates the thickness of the left temporal pole, right lingual gyrus, and left pars opercularis as the most prominent markers for MwA classification, while the thickness of left pericalcarine gyrus and left pars opercularis are proposed as the two most important features for the simple and complex MwA classification.Discussion: This method shows significant potential in the validation of MwA diagnosis and subtype classification, which can tackle and challenge the current treatments of MwA.
Computer software that is able to identify the plant type given the image can be very beneficial tool in the area of botany, horticulture and agriculture. Besides improving the research possibilities in the mentioned fields, it can be used as entertaining learning tool or it can be applied to other similar domains. Flower classification can be quite challenging task, since the majority of flowers have highly similar main features. This paper proposes using convolutional neural networks for flower classification. The first step in this research was preparing the dataset for network training. Numerous of network models were implemented during this research, but the main focus is on LeNet and AlexNet models. AlexNet model with Sigmoid Uniform function for allocating initial weights provided the best classification results in this research. Network training, as well as the data preprocessing was conducted in deeplearning4j Java library.
This paper explores the application of artificial intelligence (AI) in education to improve teaching processes. Through the analysis of relevant literature, various aspects of the application of AI technologies in the educational sector are explored. It has been found that AI can provide a personalized academic experience and more efficient information retrieval; however, AI answers are perceived as often lacking in accuracy and correctness and should not be unquestioningly trusted. The research instrument was an online survey given to 105 university students with several types of questions to discover whether our students were familiar with the concept of AI tools. Through the analysis of the available literature and the student’s response, it is concluded that a viable option to improve the learning and teaching process is a balance between traditional teaching methods and the application of AI technologies, where AI could support teachers through dynamic and interactive lessons and provide additional resources tailored to individual student needs. At the same time, the need for careful development and use of AI technologies concerning ethical principles and continuous expert supervision is strongly emphasized, underlining the crucial role of educators, researchers, and policymakers in ensuring responsible AI use in education.
Online learning has become increasingly prevalent in all education levels during recent years. While in highly developed regions transition from traditional to online learning happens without significant difficulties, in underdeveloped and developing countries introducing students to online learning is typically followed by complications and frustration. Many researchers conducted studies to solve the issue of conforming to online learning and provide equal opportunities to all students regardless of their demographical characteristics and environmental factors. Introducing artificial intelligence tools to this problem can provide valuable insight into patterns and predictors in online education. This study proposes a machine learning model for predicting the low-level student adaptability to online learning. This model can indicate students who might have difficulties adapting to online learning with 94% accuracy based on their demographical and environmental characteristics. The model is developed using locally weighted learning with a C4.5 decision tree classifier. This paper contributes to understanding the problems underlying online learning adaptability and offers an accurate tool for detecting students prone to online learning issues, which can help persons of authority provide dependable and rapid aid.
Digital resilience refers to the adaptive capacity for overcoming the challenges of the digital landscape. It encompasses the ability to effectively respond to and recover from adverse online experiences, demonstrating mental agility and digital competencies. We developed a program „Digital Resilience and Networking“ under the Erasmus+ project „Enhancing digital and psychological resilience through peer networking in the online environment in times of crisis“. This paper describes the layout and implementation of the program: the piloting phase (implemented in hybrid mode in the English language at three partner universities from Serbia, Poland, and Italy), the evaluation, and its implementation in Serbian language (national training, held fully online). We prepared a questionnaire to evaluate the effects of the course and cross-analyzed the training session evaluation (international pilot – 35 feedback records, and national training – 32 records). The evaluation showed high ratings for both phases, revealing significant differences between them. Furthermore, a strong correlation was found between participants’ satisfaction with the program’s implementation and their perceived skill acquisition. These results offer valuable insights into the efficacy of the program in enhancing digital resilience skills and emphasize the significance of well-implemented courses for optimal learning outcomes.