Premium
Poster — Thur Eve — 09: Effects of Small Sample Size on Diffuse Reflectance Spectroscopy for the Identification of Brain Tumours
Author(s) -
Cappon D,
Hayward J,
Farrell T,
Fang Q
Publication year - 2010
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.3476114
Subject(s) - diffuse reflectance infrared fourier transform , spectroscopy , diffuse reflection , monte carlo method , materials science , attenuation coefficient , optics , reflectivity , signal (programming language) , absorption (acoustics) , radius , analytical chemistry (journal) , nuclear magnetic resonance , nuclear medicine , biomedical engineering , chemistry , physics , medicine , mathematics , statistics , catalysis , biochemistry , computer security , chromatography , quantum mechanics , photocatalysis , computer science , programming language
An optical system using diffuse reflectance spectroscopy (DRS) and time‐resolved laser induced fluorescence spectroscopy (TR‐LIFS) is being developed to identify surgical margins during brain tumour resection. Since initial trials will be performed ex vivo on small samples, we investigated our ability to perform DRS for these small sample volumes. We examined both the effects of decreasing sample volume on the DRS signal and decreasing probe size on our ability to recover optical properties. Sample volume effects were studied by performing DRS measurements on bovine brain tissue samples of various sizes and comparing the signals to those for very large samples. A decrease in DRS signal strength was observed at sample volumes of less than 500 mm 3 . Probe size effects were studied using Monte Carlo simulations to obtain the spatially‐resolved reflectance using the optical properties of human brain tissue. Varying amounts of noise were added to the reflectance, and optical properties were recovered by fitting to Monte Carlo generated reflectance curves. The recovered properties were compared to their actual values as the maximum radius used for fitting was reduced. Our ability to recover the absorption coefficient was improved when fitting to absolute reflectance data.