Spanish Non-small Cell Lung Cancer Patients - A Survival Analysis

• Autor(a) principal: Matos, Filipa Mendes de
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/133293
• topic: Health; Machine Learning; Non-small cell lung cancer; Kaplan-Meier; Cox's Multivariate; Proportional Hazard Model; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

Health is an invaluable asset that prevails in any society throughout human history. As a priceless good, humans are willing to make all sacrifices to ensure this precious good. The techniques and technologies used over time, their massification and constant evolution, end up raising a highly competitive and increasingly sophisticated industry in a constant search directed to the population's most pressing needs. The digitalization of clinical processes and the immensity data acquired is currently an enormous knowledge capital. Cancer is the second leading cause of death worldwide, and its increasing number is a concern for world health organisations that are trying to reverse this trend. In this context, there is an urgent need for projects and research such as this dissertation that aims to achieve early diagnosis, predict and prevent disease repercussions, improve quality of life, and increase survivorship rates in cancer patients. This work focuses explicitly on lung cancer patients. Lung cancer is one of the most typical cancers worldwide, being the most common global cause of cancer death in men and the third most common in women. Non-small cell lung cancer (NSCLC) accounts for approximately 80% of all lung malignancies. In addition, the incidence of lung cancer has been gradually increasing over the last 50 years, becoming a worldwide public health issue. The dataset analysed is specific for non-small cell lung cancer patients, and all its attributes were reviewed from a clinical perspective. After understanding the dataset's content and the pre-processing data phase followed a descriptive analysis of each attribute, and the use of the Kaplan-Meier method. Finally, this work proposes the use of Cox's Multivariate Proportional Hazard Model. Additionally, this dissertation reviews the applications domain, including the Healthcare industry structure and Information Systems and the technical knowledge necessary to implement Machine Learning algorithms. This work was supported by Holos S.A. and engaged with the CLARIFY project (European Union Horizon 2020- under grant agreement nº 875160).