Type 1 Diabetes (T1D) is an autoimmune disease that requires exogenous insulin via Multiple Daily Injections (MDIs) or subcutaneous pumps to maintain targeted glucose levels. Despite the advances in Continuous Glucose Monitoring (CGM), controlling glucose levels remains challenging. Large Language Models (LLMs) have produced impressive results in text processing, but their performance with other data modalities remains unexplored. The aim of this study is three-fold. First, to evaluate the effectiveness of LLM-based models for glucose forecasting. Second, to compare the performance of different models for predicting glucose in T1D individuals treated with MDIs and pumps. Lastly, to create a personalized approach based on patient-specific training and adaptive model selection. CGM data from the T1DEXI study were used for forecasting glucose levels. Different predictive models were evaluated using the mean absolute error (MAE) and the root mean squared error and considering the Prediction Horizons (PHs) of 60, 90, and 120 min. For short-term PHs (60 and 90 min), the personalized approach achieved the best results, with an average MAE of 15.7 and 20.2 for MDIs, and a MAE of 15.2 and 17.2 for pumps. For long-term PH (120 min), TIDE obtained an MAE of 19.8 for MDIs, whereas Patch-TST obtained a MAE of 18.5. LLM-based models provided similar MAE values to state-of-the-art models but presented a reduced variability. The proposed personalized approach obtained the best results for short-term periods. Our work contributes to developing personalized glucose prediction models for enhancing glycemic control, reducing diabetes-related complications.

Machine learning (ML) models have been extensively used for tabular data classification but recent works have been developed to transform tabular data into images, aiming to leverage the predictive performance of convolutional neural networks (CNNs). However, most of these approaches fail to convert data with a low number of samples and mixed-type features. This study aims: to evaluate the performance of the tabular-to-image method named low mixed-image generator for tabular data (LM-IGTD); and to assess the effectiveness of transfer learning and fine-tuning for improving predictions on tabular data. We employed two public tabular datasets with patients diagnosed with cardiovascular diseases (CVDs): Framingham and Steno. First, both datasets were transformed into images using LM-IGTD. Then, Framingham, which contains a larger set of samples than Steno, is used to train CNN-based models. Finally, we performed transfer learning and fine-tuning using the pre-trained CNN on the Steno dataset to predict CVD risk. The CNN-based model with transfer learning achieved the highest AUCORC in Steno (0.855), outperforming ML models such as decision trees, K-nearest neighbors, least absolute shrinkage and selection operator (LASSO) support vector machine and TabPFN. This approach improved accuracy by 2% over the best-performing traditional model, TabPFN. To the best of our knowledge, this is the first study that evaluates the effectiveness of applying transfer learning and fine-tuning to tabular data using tabular-to-image approaches. Through the use of CNNs' predictive capabilities, our work also advances the diagnosis of CVD by providing a framework for early clinical intervention and decision-making support.

Noncommunicable diseases (NCDs) are the leading cause of morbidity and mortality worldwide. Cardiovascular diseases (CVDs) and diabetes are the most prevalent NCDs, causing 1.9 and 1.5 million deaths yearly. Individuals diagnosed with type 1 diabetes (T1D) are at high risk of developing CVDs. Machine learning (ML) models have provided outstanding results in different domains, including healthcare, allowing to obtain models with high predictive performance. The aim of this study was to develop an interpretable data-driven approach to predict the 10-year CVD risk for T1D older individuals, aiming to provide both reasonable predictive performance and the identification of risk factors associated with CVDs. Data from T1D individuals at the Steno Diabetes Center Copenhagen were used. Different ML-based models were considered, including KNN, decision tree, random forest, and multilayer perceptron (MLP). To enhance the predictive performance of ML models, the conditional tabular generative adversarial network (CTGAN) was used to create synthetic data and increase the size of the training data. Several filter and wrapper feature selection (FS) techniques were considered for identifying the most relevant features involved in CVD risk and enhancing the performance of the ML-based models used. To gain interpretability on predictive models, we used the post-hoc methods: SHAP and accumulated local effects. The experimental results showed a great performance of FS and ML-based models for predicting CVD risk. In particular, the MLP obtained the best results, with a mean absolute error of 0.0088 and mean relative absolute error of 0.0817. Regarding risk factors, age, Hba1c, and albuminuria were identified as crucial in CVD risk prediction, which is in line with recent clinical evidence. Our study contributes to identifying CVD risk and associated risk factors in a data-driven manner, helping to make early interventions and adequate treatments to prevent CVDs.

Background: Antidepressants are essential in managing depression, including treatment-resistant cases. Public perceptions of these medications, shaped by social media platforms like X (formerly Twitter), can influence treatment adherence and outcomes. This study explores public attitudes toward antidepressants through sentiment and topic modeling analysis of tweets in English and Spanish from 2007 to 2022. Methods: Tweets mentioning antidepressants approved for depression were collected. The analysis focused on selective serotonin reuptake inhibitors (SSRIs) and glutamatergic drugs. Sentiment analysis and topic modeling were conducted to identify trends, concerns, and emotions in discussions across both languages. Results: A total of 1,448,674 tweets were analyzed (1,013,128 in English and 435,546 in Spanish). SSRIs were the most mentioned antidepressants (27.9% in English, 58.91% in Spanish). Pricing and availability were key concerns in English tweets, while Spanish tweets highlighted availability, efficacy, and sexual side effects. Glutamatergic drugs, especially esketamine, gained attention (15.61% in English, 25.23% in Spanish), evoking emotions such as fear, sadness, and anger. Temporal analysis showed significant increases in discussions, with peaks in 2012 and 2021 for SSRIs in Spanish, and exponential growth from 2018 to 2021 for glutamatergic drugs. Emotional tones varied across languages, reflecting cultural differences. Conclusions: Social media platforms like X provide valuable insights into public perceptions of antidepressants, highlighting cultural variations in attitudes. Understanding these perceptions can help clinicians address concerns and misconceptions, fostering informed treatment decisions. The limitations of social media data call for careful interpretation, emphasizing the need for continued research to improve pharmacovigilance and public health strategies.

Type 1 Diabetes (T1D) is an autoimmune condition that results in an insulin deficiency. People with T1D require the administration of exogenous insulin to maintain target glucose levels. However, insulin therapy can cause hypoglycemic episodes, which occur when blood glucose levels fall below 70 mg/dL. Nocturnal Hypoglycemia (NH) occurs while the individual is asleep and can lead to different clinical complications. Developing predictive approaches to predict NH before sleep could reduce these episodes and mitigate acute complications. While numerous models exist for Time Series Classification (TSC), their use for NH prediction remains limited. This study evaluates 14 different TSC models for NH prediction, assessing their performance by evaluating classification metrics, computational time, and environmental impact (measured by energy consumption and CO₂ emissions). The approaches include distance-based, convolutional-based, deep learning, dictionary-based, feature-based, shapelet-based, and interval-based methods. We employed glucose data from 52 individuals with T1D. Experimental results showed that interval-based and feature-based approaches achieved the best predictive performance, obtaining the highest Area Under the Curve Operator (AUCROC) of 0.703. Additionally, both demonstrated low environmental impact due to their short computational time. However, substantial differences in environmental impact were observed depending on the approach. Distance-based methods and deep learning approaches exhibited the highest environmental impact. This paper provides key insights into the effectiveness of TSC models for NH prediction, highlighting the trade-off between model performance and environmental impact.