Self-calibrating time-resolved near infrared spectroscopy

TY - JOUR

T1 - Self-calibrating time-resolved near infrared spectroscopy

AU - Wojtkiewicz, Stanislaw

AU - Gerega, Anna

AU - Zanoletti, Marta

AU - Sudakou, Aleh

AU - Contini, Davide

AU - Liebert, Adam

AU - Durduran, Turgut

AU - Dehghani, Hamid

PY - 2019/5/1

Y1 - 2019/5/1

N2 - Time-resolved near infrared spectroscopy is considered to be a gold standard technique when measuring absolute values of tissue optical properties, as it provides separable and independent information about both tissue absorption and scattering. However, time-resolved instruments require an accurate characterization by measuring the instrument response function in order to decouple the contribution of the instrument itself from the measurement. In this work, a new approach to the methodology of analysing time-resolved data is presented where the influence of instrument response function is eliminated from the data and a self-calibrating analysis is proposed. The proposed methodology requires an instrument to provide at least two wavelengths and allows spectral parameters recovery (optical properties or constituents concentrations and reduced scatter amplitude and power). Phantom and in-vivo data from two different time-resolved systems are used to validate the accuracy of the proposed self-calibrating approach, demonstrating that parameters recovery compared to the conventional curve fitting approach is within 10% and benefits from introducing a spectral constraint to the reconstruction problem. It is shown that a multiwavelength time-resolved data can be used for parameters recovery directly without prior calibration (instrument response function measurement).

AB - Time-resolved near infrared spectroscopy is considered to be a gold standard technique when measuring absolute values of tissue optical properties, as it provides separable and independent information about both tissue absorption and scattering. However, time-resolved instruments require an accurate characterization by measuring the instrument response function in order to decouple the contribution of the instrument itself from the measurement. In this work, a new approach to the methodology of analysing time-resolved data is presented where the influence of instrument response function is eliminated from the data and a self-calibrating analysis is proposed. The proposed methodology requires an instrument to provide at least two wavelengths and allows spectral parameters recovery (optical properties or constituents concentrations and reduced scatter amplitude and power). Phantom and in-vivo data from two different time-resolved systems are used to validate the accuracy of the proposed self-calibrating approach, demonstrating that parameters recovery compared to the conventional curve fitting approach is within 10% and benefits from introducing a spectral constraint to the reconstruction problem. It is shown that a multiwavelength time-resolved data can be used for parameters recovery directly without prior calibration (instrument response function measurement).

UR - http://www.scopus.com/inward/record.url?scp=85066973524&partnerID=8YFLogxK

U2 - 10.1364/BOE.10.002657

DO - 10.1364/BOE.10.002657

M3 - Article

C2 - 31149386

VL - 10

SP - 2657

EP - 2669

JO - Biomedical Optics Express

JF - Biomedical Optics Express

SN - 2156-7085

IS - 5

ER -