Effect of wavelength selection on the accuracy of blood oxygen saturation estimates obtained from photoacoustic images

Abstract

In photoacoustic tomography (PAT) the image contrast is due to optical absorption, and because of this PAT images are sensitive to changes in blood oxygen saturation (sO 2 ). However, this is not a linear relationship due to the presence of a non-uniform light fluence distribution. In this paper we systematically evaluate the conditions in which an approximate linear inversion scheme -which assumes the internal fluence distribution is unchanged when the absorption coefficient changes -can give accurate estimates of sO 2 . A numerical phantom of highly vascularised tissue is used a test case for this assumption. It is shown that using multiple wavelengths over a broad range of the near-infrared spectrum yields inaccurate estimates of oxygenation, while a careful selection of wavelengths in the 620-920nm range is likely to yield more accurate oxygenation values. (We demonstrate that a 1D fluence correction, obtained from the average decay rate in the image, can further improve the estimates.) However, opting to use these longer wavelengths involves sacrificing signal-to-noise ratio in the image, as the absorption of blood is low in this range. This results in an inherent trade-off between uncertainty in the sO 2 estimates due to fluence variation and error due to noise. This study shows that the depth to which sO 2 can be estimated accurately using a linear approximation is limited in vivo, even with idealised measurements, to at most 3mm. In practice, there will be even greater uncertainties affecting the estimates, eg. due to bandlimited or partial-view acoustic detection.

Optimal wavelengths for optoacoustic measurements of blood oxygen saturation in biological tissues

“…The problem of selecting the optimal wavelength for two-wavelength OA measurements of blood oxygen saturation has been raised in a number of studies [29][30][31][32][33]. In [29] the use of the condition number of the absorption spectra matrix as an indicator of the stability for linear least squares spectral unmixing was proposed.…”

“…In [31] an algorithm for optimal wavelength selection, based on the Cramer-Rao lower bound, uses a biological environment model as input data together with initial assumptions about the tissue composition. In most studies [30,31,33,34] the fluence distribution is considered to be known and is calculated from the biological tissue scattering and absorption parameters, which are accepted as being homogeneous in the investigated biological tissue. The optical fluence is calculated either by the Monte Carlo simulation [33], or is given by an exponential decay function depending on depth [31,34], with the optical parameters taken from [35,36].…”

“…In most studies [30,31,33,34] the fluence distribution is considered to be known and is calculated from the biological tissue scattering and absorption parameters, which are accepted as being homogeneous in the investigated biological tissue. The optical fluence is calculated either by the Monte Carlo simulation [33], or is given by an exponential decay function depending on depth [31,34], with the optical parameters taken from [35,36]. However, biological tissues are not uniformly absorbing and scattering objects, and the large variations of the absorption and scattering parameters of the biological tissue do not allow modeling the fluence with sufficient accuracy, thus resulting in errors in the determination of blood oxygen saturation.…”

Estimating blood oxygenation from photoacoustic images: can a simple linear spectroscopic inversion ever work?

“…For this reason, in the next section, consideration is given as to whether more accurate estimates could be obtained by selecting wavelengths beyond 600 nm, where blood absorption and its spectral dependence is relatively low. 44…”

Generation of anatomically realistic numerical phantoms for photoacoustic and ultrasonic breast imaging

“…This is particularly true when task-based image quality measures are employed. 18 While some recent works in quantitative PAT have employed realistic numerical phantoms derived from μCT rat and mouse brain images, 19,20 those phantoms were utilized for small animal imaging and therefore provided limited guidance for clinical system designs. As such, there remains a need for the development of anatomically realistic numerical breast phantoms suitable for clinical purposes to advance PACT and USCT technologies.…”