Cargando...
Fecha
2025-09
Editor/a
Director/a
Tutor/a
Coordinador/a
Prologuista
Revisor/a
Ilustrador/a
Derechos de acceso
info:eu-repo/semantics/openAccess
Título de la revista
ISSN de la revista
Título del volumen
Editorial
Citas
0 citas en 
0 citas en 
Resumen
La población mundial, en concreto aquella que habita en países desarrollados, está envejeciendo a un ritmo cada vez más acelerado. Asociadas precisamente a este envejecimiento generalizado de la sociedad, aparecen con mayor prevalencia enfermedades como el cáncer o la sarcopenia. La detección temprana de la sarcopenia en pacientes es muy importante para mejorar la salud de una población cada vez más envejecida. Además, dada la sinergia que tiene la sarcopenia con el cáncer (aumenta la toxicidad del mismo), este problema cobra aún más relevancia, sobre todo en poblaciones más envejecidas. En los últimos años, el uso de aprendizaje automático y aprendizaje profundo, en concreto, en contextos asociados a imagen médica, ha visto un aumento en la literatura científica. La sarcopenia es una enfermedad que, en sí misma, es muy difícil de diagnosticar. Es por esto que el tiempo que se tarda en diagnosticarla es alto. Por tanto, el uso de modelos dedicados a la segmentación de masa muscular (redes neuronales convolucionales, por ejemplo) en pacientes sanos versus pacientes enfermos puede ayudar a acelerar el proceso de detección. Este trabajo busca dar mayor visibilidad a esta problemática: mediante el uso de la herramienta del meta-análisis, se navegará a través de la literatura asociada a este problema y se desvelará cuál es el rendimiento general de los modelos de aprendizaje automático, de forma que se pueda avanzar hacia una mayor integración de los mismos en los sistemas de decisión de los sistemas sanitarios públicos. Los resultados obtenidos indican un conjunto de datos bastante heterogéneo, con mucho sesgo de publicación. Sin embargo, el análisis de subgrupos presenta menos heterogeneidad. En general, los resultados obtenidos ofrecen resultados bastante positivos en lo que se refiere a la segmentación de músculo esquelético en estos contextos, lo que puede derivar en que se pueda implementar a medio plazo en los sistemas de decisión sanitarios.
The world’s population, particularly those living in developed countries, is ageing at an increasingly rapid rate. Associated precisely with this general age-trend of society, diseases such as cancer and sarcopenia are becoming more prevalent. Early detection of sarcopenia in patients is essential for improving the health of an increasingly older population. Furthermore, given the synergy between sarcopenia and cancer (it increases the toxicity of the latter), this problem becomes even more relevant, especially in older populations. In recent years, the application of machine learning and deep learning, particularly in medical imaging contexts, has seen a significant increase in scientific literature. Sarcopenia is a disease that is very difficult to diagnose on its own. This is why it takes a considerable amount of time to diagnose. Therefore, the use of models dedicated to muscle mass segmentation (convolutional neural networks, for example) in healthy patients versus sick patients can help detect it earlier. This work seeks to give greater visibility to this problem: using the meta-analysis tool, we will navigate through the literature associated with this problem and reveal the overall performance of machine learning models, so that we can move towards greater integration of these models into the decision-making systems of public health systems. The results obtained indicate a fairly heterogeneous data set, with a high degree of publication bias. However, the subgroup analysis shows less heterogeneity. In general, the results obtained are quite positive in terms of skeletal muscle segmentation in these contexts, which may lead to its implementation in healthcare decision-making systems in the medium term.
The world’s population, particularly those living in developed countries, is ageing at an increasingly rapid rate. Associated precisely with this general age-trend of society, diseases such as cancer and sarcopenia are becoming more prevalent. Early detection of sarcopenia in patients is essential for improving the health of an increasingly older population. Furthermore, given the synergy between sarcopenia and cancer (it increases the toxicity of the latter), this problem becomes even more relevant, especially in older populations. In recent years, the application of machine learning and deep learning, particularly in medical imaging contexts, has seen a significant increase in scientific literature. Sarcopenia is a disease that is very difficult to diagnose on its own. This is why it takes a considerable amount of time to diagnose. Therefore, the use of models dedicated to muscle mass segmentation (convolutional neural networks, for example) in healthy patients versus sick patients can help detect it earlier. This work seeks to give greater visibility to this problem: using the meta-analysis tool, we will navigate through the literature associated with this problem and reveal the overall performance of machine learning models, so that we can move towards greater integration of these models into the decision-making systems of public health systems. The results obtained indicate a fairly heterogeneous data set, with a high degree of publication bias. However, the subgroup analysis shows less heterogeneity. In general, the results obtained are quite positive in terms of skeletal muscle segmentation in these contexts, which may lead to its implementation in healthcare decision-making systems in the medium term.
Descripción
Categorías UNESCO
Palabras clave
meta-análisis, aprendizaje profundo, sarcopenia, cáncer, segmentación, musculatura esquelética, revisión sistemática, heterogeneidad, meta-analysis, deep learning, sarcopenia, cancer, segmentation, skeletal muscle, systematic review, heterogeneity
Citación
Chamizo González, Aitor. Trabajo Fin de Máster: "Aprendizaje profundo para la segmentación de músculo esquelético y detección de sarcopenia en pacientes oncológicos: Revisión sistemática y meta-análisis". Universidad Nacional de Educación a Distancia (UNED), 2025
Centro
E.T.S. de Ingeniería Informática