Abstract
Time-resolved temporal point spread function (TPSF) measurement of near
infrared spectroscopic (NIRS) data allows the estimation of absorption and reduced scattering properties of biological tissues. Such analysis requires an iterative calculation of the theoretical TPSF curve using mathematical and computational models of the domain being imaged which are computationally complex and expensive. In this work an efficient methodology for representing the TPSF data using a superposition of cosines calculated in frequency domain is presented. The proposed method is outlined and tested on a finite element realistic models of human neck and head. Using an adult head model containing
~140k nodes, the TPSF calculation at each node for one source is accelerated from 3.11 s to 1.29 s within an error limit of ± 5% related to the time domain calculation method.
infrared spectroscopic (NIRS) data allows the estimation of absorption and reduced scattering properties of biological tissues. Such analysis requires an iterative calculation of the theoretical TPSF curve using mathematical and computational models of the domain being imaged which are computationally complex and expensive. In this work an efficient methodology for representing the TPSF data using a superposition of cosines calculated in frequency domain is presented. The proposed method is outlined and tested on a finite element realistic models of human neck and head. Using an adult head model containing
~140k nodes, the TPSF calculation at each node for one source is accelerated from 3.11 s to 1.29 s within an error limit of ± 5% related to the time domain calculation method.
| Original language | English |
|---|---|
| Pages (from-to) | 41-54 |
| Number of pages | 14 |
| Journal | Biomedical Optics Express |
| Volume | 9 |
| Issue number | 1 |
| Early online date | 4 Dec 2017 |
| DOIs | |
| Publication status | Published - Jan 2018 |