Quantification of dual-state 5-ALA-induced PpIX fluorescence: Methodology and validation in tissue-mimicking phantoms

TL;DR

The paper tackles quantitative imaging of 5-ALA–induced PpIX in gliomas by addressing dual-state emission ($PpIX_{620}$ and $PpIX_{635}$) that overlaps with autofluorescence and tissue optical distortions. It introduces a blind spectral unmixing pipeline based on non-negative matrix factorization and an empirical optical-correction framework, validated in novel tissue-mimicking phantoms that reproduce glioma-like $μ_a$ and $μ_s'$, dual PpIX emission, and background autofluorescence. The method achieves a strong correlation with ground-truth PpIX concentrations ($R^2 = 0.918 \pm 0.002$) and identifies an optimal correction strategy (abundance correction with a Hellinger distance objective and $\alpha = -1.0$). This work supports robust, quantitative PpIX fluorescence imaging with potential clinical translation to fluorescence-guided glioma surgery and adaption to both wide-field and probe-based devices, by using readily implementable diffuse reflectance–based inputs. The validated pipeline thus advances quantitative fluorescence imaging toward real-time, concentration-based assessment in neuro-oncology.

Abstract

Quantification of protoporphyrin IX (PpIX) fluorescence in human brain tumours has the potential to significantly improve patient outcomes in neuro-oncology, but represents a formidable imaging challenge. Protoporphyrin is a biological molecule which interacts with the tissue micro-environment to form two photochemical states in glioma. Each exhibits markedly different quantum efficiencies, with distinct but overlapping emission spectra that also overlap with tissue autofluorescence. Fluorescence emission is known to be distorted by the intrinsic optical properties of tissue, coupled with marked intra-tumoural heterogeneity as a hallmark of glioma tumours. Existing quantitative fluorescence systems are developed and validated using simplified phantoms that do not simultaneously mimic the complex interactions between fluorophores and tissue optical properties or micro-environment. Consequently, existing systems risk introducing systematic errors into PpIX quantification when used in tissue. In this work, we introduce a novel pipeline for quantification of PpIX in glioma, which robustly differentiates both emission states from background autofluorescence without reliance on a priori spectral information, and accounts for variations in their quantum efficiency. Unmixed PpIX emission forms are then corrected for wavelength-dependent optical distortions and weighted for accurate quantification. Significantly, this pipeline is developed and validated using novel tissue-mimicking phantoms replicating the optical properties of glioma tissues and photochemical variability of PpIX fluorescence in glioma. Our workflow achieves strong correlation with ground-truth PpIX concentrations (R2 = 0.918+-0.002), demonstrating its potential for robust, quantitative PpIX fluorescence imaging in clinical settings.

Figures (5)

• Figure 1: Comparison of optical properties (Top: Absorption coefficient $\mu_a$, Bottom: Reduced scattering coefficient $\mu_{s}'$) between measurements of a selection of our phantoms and reported values for glioma tissues from various sources, over a wavelength range of 400 - 1100 nm Yaroslavsky2002OpticalRangeGebhart2006InAdding-doublingDeRicBevilacqua1999InBrainHonda2018DeterminationAcid
• Figure 2: Raw fluorescence emission spectra for a selection of our phantoms spanning various concentrations of optical agents IL and Hb, with fixed PpIX concentration of 5 µg/mL. The overall intensity and spectral shape of PpIX fluorescence emission is markedly impacted by the concentrations of optical agents. Blue arrows indicate spectra were a mixture of the two PpIX forms PpIX$_{620}$ and PpIX$_{635}$ is noticeable, while red arrows indicate spectra where PpIX$_{635}$ seem to be dominant.
• Figure 3: Visualisation of the basis spectra discovered using the proposed unmixing approach with the Hellinger cost function and different regularisation strategies. Left: Basis spectra identified without regularisation. Middle: Tikhonov regularisation is applied to promote smoother endmembers. Right: A fifth end member is introduced to capture a sharp systematic feature. All end members are normalised for visualisation purposes.
• Figure 4: Visualisation of the discovered basis spectra and some exemplar reconstructions for each of our unmixing methods. Each endmember is normalised and the Hellinger basis is squared, for visualisation purposes. The reconstruction errors are also displayed for quantitative comparison.
• Figure 5: Visualisation of the correlations between the predicted and known PpIX concentrations in our phantom dataset for each of our unmixing and optical property correction methods. The dashed reference line represents the ideal 1:1 relationship, while the red line indicates the line of best fit through the predictions.