Repository logo

Browsing by Issue Date, starting with "2024-04-10"

Filter results by year or month
Now showing 1 - 2 of 2
  • Exploring Lifestyle Factors and Treatment Adherence among Older Adults with Hypertension Attending a Mobile Health Unit (MHU) in a Rural Area of Central Portugal
    Publication . Pinto, Cátia; Margarida Correia Balula Chaves, Cláudia; Duarte, João; Raposo, António; Zandonadi, Renata Puppin; Monteiro, Sara; Teixeira-Lemos, Edite
    This cross-sectional and analytical study aimed to characterize a sample of hypertensive older adults attending a Mobile Health Unit (MHU) in a rural area of central Portugal according to their lifestyle and to analyze the impact of lifestyles on treatment adherence. The sample comprised 235 Portuguese hypertense patients, mainly females (63.8%) with a mean age of 75 years (±8.14 years) and low level of education. The data collection was carried out through a questionnaire consisting of sociodemographic questions, dietary variables, an Alcohol Dependence Questionnaire, an International Physical Activity Questionnaire (Short Version), a Nutrition Health Determination Questionnaire, a Self-Care with Hypertension Scale, and an Adherence to Treatments Measurement Scale. Only 34.5% of the hypertensive patients have controlled blood pressure values (28.2% men and 38% women). However, more than half (56.2%) of the hypertensive patients are classified as adherent to therapeutic measures. The hypertensive individuals, who present higher levels of adherence to the treatment, do not present alcohol dependence, are frequent consumers of aromatic herbs, sporadically consume salt, present good nutritional health, and practice moderate physical activity. The predictor variables for treatment adherence are the self-care dimensions general dietary (p = 0.001), specific dietary (p = 0.034), physical activity (p = 0.031), and antihypertensive medication intake (p < 0.001). Hypertensive patients with healthier lifestyles present better levels of treatment adherence. Therefore, promoting physical activity and healthy dietary practices is necessary to improve treatment adherence and increase antihypertensive treatment’s effectiveness.
    2024-04-10Journal article Open access Show more
  • Detection of fake images generated by deep learning
    Publication . Monteiro, Stéphane Mesquita; Lacerda, Ana Cristina Wanzeller Guedes de; Caldeira, Filipe Manuel Simões
    During the last few years, the amount of audiovisual content produced is continually increasing with technology development. Along with this growth comes the availability of the same information through numerous devices that any individual holds, including smartphones, laptops, tablets, and smart TVs, in an entirely free and open manner. These type of content are considered an authenticity element since they represent a reality record. For example, in court, photos frequently determine the jury's course of action since what is available is a recorded picture that validates a narrative and usually does not leave room for doubts. However, with the advancement of Deep Learning (DL) algorithms, a new and dangerous trend known as Deepfakes begins to emerge. For example, a deepfake can be a video or an image of a person on which their face or body is totally or partially modified to appear to be someone else. This technique is often used for manipulation, blackmailing, and spreading false information. After recognizing such a dangerous problem, this study aims to uncover patterns that deepfakes show to identify authenticity as accurately as possible, using machine learning and deep learning algorithms. To get the highest level of accuracy, these algorithms were trained on datasets that included both real and phony photos. The outcomes demonstrate that deepfakes can be accurately identified and that the optimal model may be selected based on the specific requirements of the application.
    2024-04-10Master thesis Open access Show more