Publicação

Developing A Novel Dashboard To Detect And Monitor Lung Cancer Cachexia

Ver documento

Detalhes bibliográficos
Resumo:Cancer cachexia worsens outcomes in cancer patients [1], causing skeletal muscle and fat loss. Up to 60% of lung cancer patients develop cachexia [1]. Its detection and management are challenging due to unclear mechanisms of origin [2]. Emerging artificial intelligence (AI) models aim to segment automatically and analyse imaging data from cancer treatment, though not yet applied in clinical practice. A user-centred dashboard was proposed to facilitate diagnosing and monitoring lung cancer cachexia, showing automated analyses of non-small-cell lung cancer (NSCLC) patient-specific data. A design sprint was conducted to understand how to develop a dashboard tailored to clinicians’ needs. Based on these insights, the developed dashboard combined clinical information and analyses from previously developed AI models [11] for segmentation using cone beam computational tomography (CBCT) images of NSCLC patients. The dashboard was tested on 31 stage III NSCLC patients, and the deeper analysis of three patients is presented. Two clinicians evaluated the developed dashboard in usefulness and usability. The design sprint proved to be efficient in accelerating the development of novel ideas with early end-user feedback, overcoming barriers to innovation implementation in healthcare [8]. The user-centred dashboard was highly appreciated and was expected to aid cachexia diagnosis and monitoring. The analyses offer clinicians valuable insights into patient-specific cachexia progression considering changes in body weight, muscle and adipose tissue waste during cancer treatment. The automated analysis has shown to be beneficial for aiding clinical decision-making. Therefore, the developed dashboard contributes to improved monitoring of cachexia in NSCLC patients. It promotes earlier detection and understanding of cachexia progression throughout cancer treatment. Consequently, it will potentially improve patient care and patient outcomes.
Autores principais:Manso, Sara Isabel Silva
Assunto:caquexia oncológica dashboard pensamento do design análise de imagens automática cancro do pulmão de células não pequenas Teses de mestrado - 2025
Ano:2025
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso restrito
Instituição associada:Universidade de Lisboa
Idioma:inglês
Origem:Repositório da Universidade de Lisboa

Registos relacionados

Registos relacionados