Revista Mexicana de Ingenieria Biomedica

Monitoring and Prediction of Infectious Diseases through Social Network Analysis

2025-08-14T19:54:15+00:00

This article presents a comprehensive approach for monitoring and predicting infectious diseases through social media analysis, focusing on the COVID-19 pandemic. The objective is to identify real-time contagion reports and estimate disease trends as a complement to traditional epidemiological surveillance. We developed a system integrating natural language processing techniques, employing the BERT model to classify contagion statements on X, and using the Gompertz function to forecast short-term case growth. The methodology also incorporates analysis of georeferenced posts, predictions via rolling windows, and spatial representation of risk areas through heat maps. Results indicate a significant correlation between X mentions and official health reports, suggesting temporal synchronicity between both data sources. Important limitations are acknowledged, such as the urban bias in X user samples and the underrepresentation of rural populations. Finally, it is concluded that social media represent a potentially valuable resource as a complementary source for generating timely epidemiological alerts, thereby strengthening public health decision-making.

WAMDS2: Early detection of wet AMD using Swin Transformer V2

2025-07-22T22:00:24+00:00

Age-related macular degeneration (AMD) is a progressive eye disease that primarily affects individuals over 50 years old. Among the AMD variants, wet is the most severe, as it represents the advanced stage of dry AMD and can cause severe vision loss if not detected in time. This study focuses on the development of WAMDS2, a web module designed to identify characteristics associated with wet AMD, facilitating early and accurate detection. To achieve this, a literature review was conducted on AMD and advanced techniques in computer vision and deep learning. The proposed model integrates Swin Transformer V2, a vision transformer implemented in PyTorch, to analyze fundus images and classify the different stages of the disease. The system’s performance was evaluated using metrics such as accuracy, recall, and F1-Score. An accuracy of 84.76% was achieved on the test set, suggesting its feasibility in clinical settings. The obtained results highlight the potential of WAMDS2 in ophthalmology and computer vision, demonstrating its capability to enhance automated diagnosis and patient care.

Modification of screen-printed electrodes with Prussian Blue (PB) and enzyme for diabetes control in sweat: Preliminary feasibility study

2025-03-01T00:12:05+00:00

The prevalence of diabetes in Mexico has increased considerably over the past 20 years, thus increasing the need for the development of non-invasive and reliable glucose monitoring platforms. The present research focus on the development and evaluation of a novel non-invasive glucose monitor for the quantification of glucose levels in sweat. This platform is based on the use of screen-printed electrodes modified with a Prussian Blue (PB) electrochemical mediator and the glucose oxidase (GO_x) enzyme to provide glucose specificity. Additionally, we present the design of a portable electronic device for generating the input signal and processing the analyte output signals. The system was evaluated in a 0.1 M phosphate buffered saline (PBS) solution supplemented with different physiologically relevant glucose concentrations. The results demonstrated that the PB mediator plays a key role in the detection of glucose, since the control electrode modified with the enzyme but lacking the PB mediator did not show trend, while the electrode modified with PB and enzyme showed a good correlation coefficient with analytical sensitivity of 0.0008 mA μM^-1. This finding could revolutionize the field of glucose monitoring, facilitating more accessible and less invasive health technologies.