RT info:eu-repo/semantics/doctoralThesis
T1 La inteligencia artificial como herramienta para la detección de patrones de inmunodeficiencia en el lupus eritematoso sistémico
A1 Usategui Martín, Iciar
A2 Universidad de Valladolid. Escuela de Doctorado
K1 Lupus eritematoso sistémico
K1 Systemic lupus erythematosus
K1 Lupus eritematosos sistémico
K1 Immunodeficiency
K1 Inmunodeficiencia
K1 Artificial intelligence
K1 Inteligencia artificial
K1 Machine learning
K1 Aprendizaje automático
K1 32 Ciencias Médicas
AB There is a link between autoimmune diseases such as SLE and primary immunodeficiency, with common genetic bases and shared clinical manifestations. Secondary immunodeficiency due to treatments is also a common problem. In addition, infections remain one of the main causes of early mortality in patients with SLE.Machine learning (ML), a branch of artificial intelligence, is capable of processing large amounts of data and identifying patterns, which is why it has been applied in different medical areas. This study proposes an ML system to help diagnose patients with SLE and immunodeficiency traits.The sample was taken from the group of patients diagnosed with SLE and being monitored by the Systemic Autoimmune Diseases Unit of the Internal Medicine Service of the University Clinical Hospital of Valladolid. After analyzing its characteristics, it was concluded that it was a solid sample for the study, in which the traits of immunodeficiency and its infectious complications were a reality.The main predictors of immunodeficiency were: weight loss, mucous ulcers, anti-MS antibodies and concurrent dose of hydroxychloroquine, with previous exposure to rituximab or corticosteroids being of lower weight.The EXtreme Gradient Boosting (XGB) method was selected and implemented for its high performance and accuracy, which outperformed the following model by almost 5%, k- Nearest Neighbor (KNN). It should be noted that the proposed system also showed the area under the curve highest (AUC 90%).The present study shows how ML is a useful instrument to recognize and predict immunodeficiency traits in patients with SLE, highlighting the XGB method, so that it could be incorporated into clinical practice as a tool to support diagnosis and ultimately improve morbidity and quality of life of patients
YR 2024
FD 2024
LK https://uvadoc.uva.es/handle/10324/66922
UL https://uvadoc.uva.es/handle/10324/66922
LA spa
NO Escuela de Doctorado
DS UVaDOC
RD 18-nov-2024