| Summary: | Background and Objective: Fluorescence spectroscopy is a potential tool for brain tumour resection surgery. The main advantages of fluorescence spectroscopy are the objective fluorescence detection, possibility to quantify the signals and high sensitivity for instance to distinguish between healthy and neoplastic tissue. This distinction is a difficult task attributed to the diffuse growth pattern of brain tumour. A low dose of 5- aminolevulinic acid (5- ALA) is given orally to the patient. 5- ALA promotes accumulation of protporphyrin IX (PpIX) in the tumour cells which is a fluorescing substance. To measure the fluorescence emission the neurosurgeon examines the brain tissue with the help of an optical probe and spectroscopy system developed at the Department of Biomedical Engineering at Linköping University. The aim of this study was to develop an optical phantom which mimics brain tumour tissue. The phantom allows measuring and performing characterisation experiments in a controlled environment. Material and methods: To develop the tissue simulating phantom black ink and 20 % intralipid was used to obtain the optical properties of high grade glioma brain tumour. The optical properties of each substance were measured with collimated transmission spectroscopy and compared to the data reported in the literature. Agar gel was used to obtain similar photobleaching properties of tissue. Different PpIX concentrations were added to the phantom and fluorescence characterisation measurements were performed. Fluorescence quantification and the photobleaching as a function of concentration were investigated. An algorithm was established to predict the fluorescence concentration in the phantom based on the photobleaching decay characteristics. Results: An optical phantom with an error of ± 4 % was developed for the specified wavelength. The fluorescence ratio measurements showed a power and integration time independent relationship while they were linearly correlated (R2 = 0.79) with the PpIX concentration which could be quantified with a resolution of 4 μg ml-1. The fluorescence ratio values showed among the concentrations significant differences (p < 0.001) by using the Kruskal-Wallis. Three parameters of laser light pulsation, initial concentration of the fluorophore and excitation power affected the photobleaching proportionally. The developed algorithm can predict the fluorophore concentration in the phantom with an error of 9.2 %. The decay times for 63 % for loss of the initial intensity in the phantom (9 s - 43 s) were comparable with clinical data (19 s - 39 s). With the assumption that the photobleaching characteristics were the same in brain tumour tissue as in the phantom an already photobleached clinical measurement was reconstructed and the initial PpIX concentration was predicted. Conclusion: This study provides insight into a detailed development process of a brain tumour simulating optical phantom. The fluorescence ratio measurements showed a significant difference (p < 0.001) among different PpIX concentrations. Furthermore, the fluorescence ratio measurements showed a power and integration time independent linear relationship (R2 = 0.79) to quantify the PpIX concentration with a resolution of 4 μg ml-1. Additionally, main excitation parameters which affected the photobleaching kinetics in vivo were investigated in the phantom. The developed method to predict the fluorescence concentration in the phantom, based on the photobleaching measurements, was feasible to predict the initial intensity from a clinical measurements in a reasonable range but further investigations are required.
|
|---|
