Tumor heterogeneity in PET-CT images
Cancer is a leading cause of morbidity and mortality [1]. The most frequent cancers worldwide are non–small cell lung carcinoma (NSCLC) and breast cancer [2], being their management a challenging task [3]. Tumor diagnosis is usually made through biopsy [4]. However, medical imaging also plays an imp...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Fecha de publicación: | 2021 |
| País: | España |
| Institución: | Universidad Complutense de Madrid (UCM) |
| Repositorio: | Docta Complutense |
| Idioma: | inglés |
| OAI Identifier: | oai:docta.ucm.es:20.500.14352/11595 |
| Acceso en línea: | https://hdl.handle.net/20.500.14352/11595 |
| Access Level: | acceso abierto |
| Palabra clave: | 616-073.7(043.2) PET-CT images Positron Emission Tomography Computed Tomography Tumors Cancer Tomography Imágenes PET-CT Tomografía por emisión de positrones Tomografía computarizada Tumores Cáncer Tomografía Diagnóstico por imagen y medicina nuclear Oncología 3204.01 Medicina Nuclear 3201.01 Oncología |
| Sumario: | Cancer is a leading cause of morbidity and mortality [1]. The most frequent cancers worldwide are non–small cell lung carcinoma (NSCLC) and breast cancer [2], being their management a challenging task [3]. Tumor diagnosis is usually made through biopsy [4]. However, medical imaging also plays an important role in diagnosis, staging, response to treatment, and recurrence assessment [5]. Tumor heterogeneity is recognized to be involved in cancer treatment failure, with worse clinical outcomes for highly heterogeneous tumors [6,7]. This leads to the existence of tumor sub-regions with different biological behavior (some more aggressive and treatment-resistant than others) [8-10]. Which are characterized by a different pattern of vascularization, vessel permeability, metabolism, cell proliferation, cell death, and other features, that can be measured by modern medical imaging techniques, including positron emission tomography/computed tomography (PET/CT) [10-12]. Thus, the assessment of tumor heterogeneity through medical images could allow the prediction of therapy response and long-term outcomes of patients with cancer [13]. PET/CT has become essential in oncology [14,15] and is usually evaluated through semiquantitative metabolic parameters, such as maximum/mean standard uptake value (SUVmax, SUVmean) or metabolic tumor volume (MTV), which are valuables as prognostic image-based biomarkers in several tumors [16-17], but these do not assess tumor heterogeneity. Likewise, fluorodeoxyglucose (18F-FDG) PET/CT is important to differentiate malignant from benign solitary pulmonary nodules (SPN), reducing so the number of patients who undergo unnecessary surgical biopsies. Several publications have shown that some quantitative image features, extracted from medical images, are suitable for diagnosis, tumor staging, the prognosis of treatment response, and long-term evolution of cancer patients [18-20]. The process of extracting and relating image features with clinical or biological variables is called “Radiomics” [9,20-24]. Radiomic parameters, such as textural features have been related directly to tumor heterogeneity [25]. This thesis investigated the relationships of the tumor heterogeneity, assessed by 18F-FDG-PET/CT texture analysis, with metabolic parameters and pathologic staging in patients with NSCLC, and explored the diagnostic performance of different metabolic, morphologic, and clinical criteria for classifying (malignant or not) of solitary pulmonary nodules (SPN). Furthermore, 18F-FDG-PET/CT radiomic features of patients with recurrent/metastatic breast cancer were used for constructing predictive models of response to the chemotherapy, based on an optimal combination of several feature selection and machine learning (ML) methods... |
|---|