RT info:eu-repo/semantics/doctoralThesis
T1 Contribución de las Tecnologías de la Información y las Comunicaciones para ayudar a la prevención del Comportamiento Suicida en Castilla y León: Un enfoque de Machine Learning y Salud Digital.
A1 Castillo Sánchez, Gema Anabel
A2 Universidad de Valladolid. Escuela de Doctorado
K1 Tecnología de la información
K1 Machine Learning
K1 e-health
K1 Salud digital
K1 Public health
K1 Salud pública
K1 5306.02 Innovación Tecnológica
AB Suicide is a public health problem. In 2020, 3,941 suicide deaths were recorded. In 2021, 4,003 suicides were reported, remaining the first External cause of death in Spain. It is estimated that 4.4% of deaths in Spain correspond to mental and behavioral disorders.The World Health Organization (WHO) encourages all countries to develop prevention strategies. However, no single factor is sufficient to explain a death by suicide, given the complexity of suicidal behaviors (SB). Therefore, some research indicates that suicide prevention will not be achieved with a single approach or strategy either.This research considers the strategies of the region, the available resources and the choice of some Information and Communication Technologies (ICT). This thesis aims to help prevent CS using ICT in the following aspects: (1) Determine in public hospitals in the CyL region, the mental disorders that most influence SB readmissions. (2) Evaluate a Digital Health Strategy (DHS) by patients with SB.Two components in the selection of attributes/variables were used to carry out the feature selection (FS) by ML, with the objective of identifying the mental disorders that influence the readmissions of patients with SB by hospitals in this region. The first component consisted of using classical statistics, which was evaluated by the Chi-Square (X2) distribution technique. The second component involves the use of ML techniques with three different approaches such as: entropy, probability and the linear relationship of the variables. Then, the evaluation was carried out by experts.The FS with ML has allowed the identification of mental disorders, which were included in the design of the proposed DHS. This technological proposal is based on a software ecosystem, to provide patient follow-up, help prevent SB and reduce the hospital burden in CyL considering the characteristics of the region.The main conclusions are: (1) ML techniques help to identify the mental disorders that most influence the readmission of patients with SB, by hospitals in CyL, with 95% significance. (2) According to ML techniques, the most influential mental disorders in hospital readmission in CyL are Adjustment Disorder, Alcohol Abuse, Depressive Syndrome, Personality Disorder and Dysthymic Disorder. (3) The Mental Disorders that most influence readmissions of patients with SB were included in the design of the DHS, to help reduce the care or hospital burden. (4) The proposed ESD was designed by different mobile, WEB and communication technological structures, which have been integrated to offer care and follow-up services to patients with SB. (5) The results of the evaluation of the patients on the APP, improvements can be made according to the issues identified in the FG. Therefore, the improvement in User Interface (UI), User Experience (UX), and Acceptability can be approached with realistic objectives. This design does not replace face-to-face monitoring by the mental health professional but is presented as an innovative complement to care services. DHS to prevent SB is promising and more research should be done in this field.Using these approaches from ICT helps to prevent SB and contributes at a technical, technological and social level in decision-making on management strategies for these patients.
YR 2023
FD 2023
LK https://uvadoc.uva.es/handle/10324/66285
UL https://uvadoc.uva.es/handle/10324/66285
LA spa
NO Escuela de Doctorado
DS UVaDOC
RD 21-dic-2024