Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

In vivo molecular imaging of epithelial pre-cancers based on dynamic optical scattering modeling

Papoutsoglou Georgios

Full record


URI: http://purl.tuc.gr/dl/dias/5508297E-5C28-4CCB-ABEA-177CA49AB588
Year 2014
Type of Item Doctoral Dissertation
License
Details
Bibliographic Citation George Papoutsoglou, "In vivo molecular imaging of epithelial pre-cancers based on dynamic optical scattering modeling", PhD Thesis, School of Electronic and Computer Engineering, Technical University of Crete, Greece, Chania, 2014 https://doi.org/10.26233/heallink.tuc.17651
Appears in Collections

Summary

We present a novel biophotonic method and imaging modality for estimating and mapping neoplasia-specific functional and structural parameters of the cervical precancerous epithelium. Estimations were based on experimental data obtained from dynamic contrast-enhanced optical imaging of cervix, in vivo. The dynamic characteristics of the measured optical signal are governed by the epithelial transport effects of the biomarker. A compartmental, pharmacokinetic, model of the cervical neoplastic epithelium has been developed, which predicts the dynamic optical effects in all possible parameter value combinations. Nine biological parameters, both structural and func-tional, have been identified to be potentially correlated with the neoplasia growth and to be mani-fested to the measured data in a convoluted manner. We have performed Global Sensitivity Analy-sis for the purpose of identifying the subset of the input parameters that are the key determinants of the model’s output. We have for the first time shown that it is possible to estimate, from in vivo measured dynamic optical data, the following neoplasia related parameters: number of neoplastic layers, extracellular space dimensions, functionality of tight junctions and extracellular pH. Global optimization techniques showed that the estimations of our method are of adequate accuracy and precision. Particularly, the Differential Evolution algorithm converged to the set of the four, most identifiable, parameters with an error of roughly 1%. We show that the estimated, in two millions of pixels, values of the four parameters are quite consistent with information provided in the literature. Our results are unique in the sense that for the first time functional and microstructural parameter maps can be estimated and displayed together, thus maximizing the diagnostic information. The quantity and the quality of this information are unattainable by other invasive and non invasive methods. The findings of this thesis suggest strongly that our method can improve our understand-ing of the neoplasia development mechanisms and of tumor growth and metastasis physiology. Corollary, it may become a valuable diagnostic tool that will also facilitate the development and evaluation of new cancer therapies.

Available Files

Services

Statistics