利用螢光蒙地卡羅模型建立螢光強度資訊表格定量雙層組織的螢光光學參數

Abstract

本研究主要是研究組織的螢光光譜，利用蒙地卡羅模型，開發一套反向螢光光譜擬合工具，用來定量組織的螢光光學參數，分析癌組織病變過程的螢光特性。 我們利用兩種方式，參數法和矩陣疊積法，建立兩種不同的螢光強度資訊表格，用於反向螢光光譜擬合工具中，避免蒙地卡羅模擬耗時的特性，增加我們擷取螢光光學參數時所需的時間成本。 透過表格格點大小分析，我們發現在表格建立的過程中，兩者方式於不同的參數組合數量下，在佔有記憶體空間和花費時間上各有優劣，參數法在參數組合數量少時較優，反之，矩陣疊積法在參數組合數量較多時，才會顯現他的優勢。 同時，我們配合本實驗室所建立的光譜量測系統，模擬臨床實驗上，反向螢光光譜擬合過程中，可能遇到的雜訊問題，包括螢光光譜測量上的誤差，和反向螢光光譜擬合過程中，用於尋找螢光強度資訊表格的漫反射光學參數誤差。我們發現，系統的量測誤差，對反向螢光光譜擬合工具，在定量螢光光學參數上，影響不大，但漫反射光學參數的誤差，卻有明顯的影響。 未來，我們將應用反向螢光光譜擬合工具於臨床實驗，配合光譜量測系統實際分析活體組織的螢光光譜，期待定量出來的螢光光學參數，能幫助癌病變的篩檢與診斷。

This research is to develop an inverse fluorescence spectra fitting tool, using the Monte Carlo model, to quantify the fluorescence optical parameters of tissue, hoping that we can acquire the tissue fluorescence information for analyzing and distinguishing the cancer tissues form normal ones. We use two different ways, MPG (Method of parameter grid) and MMC (Method of matrix convolution), to build the fluorescence information table, and apply them into our inverse fluorescence spectra fitting tool to reduce the time cost during the parameters extracted process, to make the inverse tool faster. We use different grid sizes of information table in our inverse fluorescence spectra fitting tool and analyze the correction of extracting parameters to get the ideas how big of information table we should build. When the size of information table is small, we find that MPG needs less memory and time than MMC does to build the information table. However, the result may reverse if the table size keeps increasing. We also analyze the fluorescence spectra noise depends on the portable spectroscopy we designed and analyze the diffusion optical parameters error which may have an effect on information table searching during parameters extracted process. We find that the former one does not have an obvious influence on the correction of extracting fluorescence optical parameters, while the later one does. In the future, we will use the inverse fluorescence spectra fitting tool to analyze the in vivo fluorescence spectra measured from our portable spectroscopy, and hoping that the optical parameters we get can help the tumor diagnosis.