Share Email Print
cover

Proceedings Paper

Quantitative subsurface fluorescence imaging enabled by spatial frequency domain imaging for enhanced fluorescence-guided surgery (Conference Presentation)
Author(s): Mira Sibai; Dennis J. Wirth; Frédéric Leblond; David W. Roberts; Brian C. Wilson
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Intra-operative fluorescence-guided resection (FGR) enables maximum safe resection of glioma by providing real-time tumor contrast. In its most widely used form, FGR is mediated by the preferential overproduction of the fluorophore protoporphyrinIX (PpIX) in malignant tissue after an oral dose of its precursor 5-Aminolevulinic Acid (ALA)1. ALA-PpIX-FGR has been shown to significantly increase completeness of tumor resection. However, the subjective visual assessment and the variable intrinsic optical attenuation of tissue limit this technique to delineating only high-grade tumors that display strong fluorescence residing at the tissue surface. We have shown that wide-field quantitative assessment by extracting 2D maps of PpIX concentration in the tissue, [PpIX], significantly improves the accuracy in detecting diffuse tumors, thereby potentially extending FGR to patients with low-grade tumors. In this approach, hyperspectral fluorescence imaging is coupled to a custom-built spatial frequency domain imaging (SFDI) system. SFDI enables the recovery of tissue optical properties maps, μ_a and μ_s^'. These are used to correct the fluorescence images. The corrected hyperspectral fluorescence images are then spectrally unmixed to separate true PpIX fluorescence from that of its photoproducts and from autofluorescence. Quantitative fluorescence imaging was validated against the clinically used spectroscopic probe by comparing the recovered optical properties and [PpIX] in vivo of a rat brain tumor model. This quantitative approach was also applied to a near infrared fluorophore ZW-800 on tissue-simulating phantoms. ALA-PpIX-FGR, as it is currently implemented, is inaccessible to infiltrative residuals lying beyond the resection cavity because of the limited penetration depth of the blue excitation light used. This is problematic as these infiltrative tumors are the main cause of reccurence. Enhanced sub-surface tumor detection was shown feasible by exciting PpIX’s secondary absorption peak of 635 nm intra-operatively on patients with various intra-cranial pathologies. However, resolving strong fluorescence of a deep-seated tumor from weak fluorescence of a shallow tumor was not possible. That is because the detected fluorescence intensity is heavily dependent on fluorophore concentration, depth, and fluorophore distribution, while also being convolved with tissue turbidity. The aim of this work, therefore, is to extend quantitative ALA-PpIX-FGR to identify sub-surface tumors by resolving tumor depth from fluorophore concentration. This should assist the surgeon in making an informed decision as for whether to further resect or not. A new quantitative depth imaging method was developed by exploiting SFDI’s depth-encoding capabilities in fluorescence mode. The result is a series of spatially modulated fluorescence images, where the modulation amplitude decays with increasing spatial frequency at a rate dependent on fluorophore depth. After recovering depth, a diffusion-based fluorescent light transport model is applied to extract fluorophore concentration. The algorithm was validated using tissue-simulating phantoms and an ex vivo tissue model indicating that the maximum depth recovered is highly dependent on fluorophore concentration as well as on tissue turbidity. For the [PpIX] and optical property maps relevant for glioma tissue, our quantitative depth fluorescence technique can predict depths up to 9 mm ± 0.4 mm, while recovering [PpIX] with an accuracy of 15% for concentrations as low as 2.5 µg/ml.

Paper Details

Date Published: 24 May 2018
PDF
Proc. SPIE 10685, Biophotonics: Photonic Solutions for Better Health Care VI, 1068517 (24 May 2018); doi: 10.1117/12.2307010
Show Author Affiliations
Mira Sibai, Univ. of Toronto (Canada)
Princess Margaret Cancer Ctr., Univ. Health Network (Canada)
Dennis J. Wirth, Thayer School of Engineering at Dartmouth (United States)
Frédéric Leblond, Ecole Polytechnique de Montréal (Canada)
David W. Roberts, Dartmouth Hitchcock Medical Ctr. (United States)
Brian C. Wilson, Princess Margaret Cancer Ctr., Univ. Health Network (Canada)
Univ of Toronto (Canada)


Published in SPIE Proceedings Vol. 10685:
Biophotonics: Photonic Solutions for Better Health Care VI
Jürgen Popp; Valery V. Tuchin; Francesco Saverio Pavone, Editor(s)

© SPIE. Terms of Use
Back to Top