Your browser does not support JavaScript!

Home    Collections    Type of Work    Doctoral theses  

Doctoral theses

Search command : Author="Στεφανίδης"  And Author="Κωνσταντίνος"

Current Record: 14 of 2177

Back to Results Previous page
Next page
Add to Basket
[Add to Basket]
Identifier 000439891
Title Advanced magnetic resonance imaging techniques in soft tissue sarcoma studies – modelling of quantitative MRI parameters
Alternative Title Νεώτερες τεχνικές απεικόνισης μαγνητικού συντονισμού στη μελέτη όγκων μαλακών μορίων - Μοντελοποίηση δεικτών q-MRI
Author Μανίκης, Γεώργιος
Thesis advisor Καραντάνας, Απόστολος
Μαρής, Θωμάς
Μαριάς, Κωνσταντίνος
Abstract Soft tissue tumors comprise a broad spectrum of mesenchymal neoplasms, including over a hundred different subtypes. Preoperative diagnosis routinely established by core needle biopsy and subsequent histopathologic examination is essential for assessment of histological subtype and biological behavior (benign or malignant, malignancy grade) in order to determine the optimal treatment. In the case of soft tissue sarcomas (STSs), i.e. those with malignant behavior, wide excision of the tumor together with a rim of adjacent healthy tissue is the surgical treatment of choice to reduce the risk of local recurrence. To assess STSs aggressiveness, the French Federation of Cancer Centers Sarcoma Group (FNCLCC) grading system is widely used. However, preoperative radiotherapy or chemotherapy may be indicated since the performance of preoperative core needle or open biopsy is an invasive procedure associated with complications that may lead to morbidity, misdiagnosis and alteration to less optimal treatment. Therefore, reliable preoperative diagnosis by a non-invasive method like imaging would be of immense value. Preoperative imaging of STSs is optimally performed by magnetic resonance imaging (MRI), as it provides supreme soft tissue contrast and may direct core or open biopsies to be taken at those most representative sites in heterogeneous tumors. Specifically, diffusion weighted imaging (DWI) has the potential to reveal insights into the structural and functional tissue properties such as cellularity, neovascularity and tissue integrity. To capture, non-invasively, these properties, apparent diffusion coefficient (ADC) was the first diffusion-related imaging parameter to quantify tissue cellularity. Next, the intravoxel incoherent model (IVIM) model was proposed, introducing a bi-exponential representation of the signal attenuation. In addition, the non-compartmentalized models including the stretched-exponential and the diffusion kurtosis model were developed to quantify tumor microstructural heterogeneity and tissue complexity. Apart from the quantitative MRI analysis, radiomics created an unprecedented momentum in computational medical imaging over the last years by significantly advancing and empowering correlational and predictive quantitative studies in numerous clinical applications. An important element of this exciting field of research is multiscale texture analysis yielding the extraction of high-throughput quantitative features from complex patterns of the diagnostic images that can rarely be seen by the human eye, subsequently used as highly informative and non-invasive imaging features for precise diagnosis, therapy planning and disease monitoring. On one hand, a significant number of DWI studies, developed from single DWI models, has been conducted to characterize STSs microenvironment, differentiate soft tissue tumor grading and assess treatment response, assuming that single models can solely characterize the overall tissue microenvironment. However, this assumption is in contradiction to the heterogeneous nature of the tumor where recent reports claimed that single models fail to appropriately capture regional functional and anatomical tumor properties, concluding to incorrect diffusion parameter values and statistics with no prior examination of models' applicability. On the other hand, although radiomics has been extensively studied in many anatomical areas, to the best of our knowledge, few studies examined the role of radiomics in STSs grading. Additionally, several concerns exist regarding the plethora of radiomics features used in the literature especially regarding their performance consistency across studies and the lack of a robust and transparent framework for the validation of the radiomic results. Motivated by the aforementioned observations, conference paper I provides a comprehensive analysis framework for DWI quantification from multiple models where most of its functionalities were further used in the pre-processing part of the radiomic analysis pipeline. DWI model development and validation was performed according to the mathematical models and the statistical analysis framework reported in Book Chapters I and II. Paper III and IV introduce a statistical analysis framework in which suitability of several DWI models was examined across all tumor pixels and a classification map was generated reflecting DWI model preference on a pixel-by-pixel-basis. These publications have set the basis to develop hybrid diffusion parameters from multiple models to differentiate low from high STSs grading. The results, published in paper I, were validated by histopathological examination of the surgical specimens, yielding to novel parameters of high discriminatory power. A secondary goal was considered in this thesis in order to examine the application of radiomics and the use of highresolution T2-MRI in the differentiation of the STSs staging. This is outlined in paper II, following a thorough investigation published in conference Paper II to assess the generalization performance and the intra-observer agreement of radiomic models as well as the relative importance of radiomics features in predicting cancer therapy response. The goals of this thesis were set towards the use of DWI quantification and radiomics into the non-invasive STSs characterization and grading differentiation.
Language English, Greek
Subject Radiomics
Ραδιομική
Σαρκώματα μαλακών μορίων
Υβριδικοί παραμετρικοί χάρτες
Issue date 2021-07-29
Collection   School/Department--School of Medicine--Department of Medicine--Doctoral theses
  Type of Work--Doctoral theses
Permanent Link https://elocus.lib.uoc.gr//dlib/7/7/5/metadata-dlib-1621243832-591944-24633.tkl Bookmark and Share
Views 27

Digital Documents
No preview available

No permission to view document.
It won't be available until: 2022-07-30