Optical properties of biological tissues: a review
2013 Phys. Med. Biol. 58 R37

Steven L. Jacques
Oregon Health & Science University, Portland, OR
jacquess@ohsu.edu

Free PDF available at http://iopscience.iop.org/0031-9155/58/11/R37
or at Jacques_PMB2013.pdf.

This webpage contains the data in Tables 1 and 2 that pertain to scattering , as well as the raw scattering spectra used to generate the tables, which were compiled from 56 references cited in the article.

Also, this webpage contains the data in Table 3 that pertain to tissue absorption properties.

This review did not cover all published data. The author continues to find both old and new references that should be added to this listing. (This task is planned.)

The data are available here in two formats:

1. as .csv files

table1_JacquesPMB2013.csv

table2_JacquesPMB2013.csv

table3_JacquesPMB2013.csv

2. as .mat Matlab files

table1_PMB2013.mat

Table 1: Parameters specifying the reduced scattering coefficient of tissues.
The data are listed as an array structure                                    
   tissue(k) for k=1:56 citations, with fields:                              
   tissue(k).nm    = wavelengths [nm]                                        
   tissue(k).musp  = reduced scattering coeff. [cm^-1]                       
   tissue(k).name  = tissue description                                      
   tissue(k).ref   = reference cited                                         
   tissue(k).type  = tissue type                                             
   tissue(k).abafm = [a b aa fray bmie] (note: aa = a' in Eq. 2)                                      

table2_PMB2013.mat

Table 2 "Average parameters for reduced scattering coefficient, mu_s [cm^-1], for tissues"
for i=1:7 --> skin, brain, breast, bone, soft, fibrous, fatty                             
   name(i).s = name of tissue type                                                        
   mn(i,:) = mean values of [a, b, aa, fray, bmie] (note: aa = a' in Eq. 2)                                        
   sd(i,:) = std deviations                                                               
   nn(i,:) = number of cited values averaged                                              

table3_PMB2013.mat

Table 3: In vivo tissue parameters governing optical absorption.          
in struct array tissue(1:47) with fields:                                
  tissue(i).ref = cited reference                                        
  tissue(i).HGb = concentration HGb [uM]                                 
  tissue(i).B   = % blood volume fraction                                
  tissue(i).S   = % HGb oxygen saturation                                
  tissue(i).W   = % water volume fraction                                
  tissue(i).F   = % fat voume fraction                                   
  tissue(i).M   = % melanosome vol.fraction in 60um epidermis            

The raw spectral data

The spectra from 56 references were acquired either from published tables or from graphs, digitizing some or all the data points in the graph. The following figures show the spectra (symbols) and the fits (lines), using two equations:

Eq. 1: µs' = a (λ/500nm)-b
(shown as solid lines) where the fitting parameters are:

Eq. 2: µs' = a' (fray(λ/500nm)-4 + fmie(λ/500nm)-bmie)
(shown as dashed lines) where the fitting parameters are:

Discussion

The literature commonly describes tissue scattering using Eq. 1, µs' = a (λ/500nm)-b. Fitting with the separate Rayleigh and Mie components of scattering as in Eq. 2, µs' = a' (fray(λ/500nm)-4 + fmie(λ/500nm)-bmie), is only slightly better. However, Eq. 2 does more reliably extrapolate to shorter wavlengths below 400 nm and longer wavelengths above 1300 nm.

The key advantage to using either Eq. 1 or Eq. 2 is that data at a limited number of wavelengths can guide the prediction of scattering behavior at other wavelengths.

Table 3 lists the values of parameters that affect absorption by a tissue, i.e., blood (B), oxygen saturation (S), water content (W), fat content (F), and melanin content described as the melanosome volume fraction (M) in a 60-µm-thick epidermis. This M can be multiplied by 0.0060 cm to yield the volume of melanosomes per cm2 skin surface, which is a metric not directly tied to an assumed epidermal thickness. These parameters can scale the known absorption spectra of oxy-hemoglobin, deoxy-hemoglobin, water, fat, and melanin.

Figure A (below) shows all the spectra on one graph.

Figure B shows the a and b of Eq. 1 as b vs a. As the scattering strength increases (bigger a) the scattering power increases (bigger b). Fibrous tissues with collagen content show higher a and b.

Figure C plots the bmie of Eq. 2 versus the b of Eq. 1. While in some cases these are in agreement (along dashed black line of unity slope), bmie often falls below the dashed line (bmie < b), since Eq. 2 lets the λ-4 behavior of the Rayleigh scattering account for the rise in scattering at shorter wavelength.

Figure D shows the ratio bmie / b versus fray. As collagen content increases and more Rayleigh scattering occurs (fray increases), the b value keeps climbing to account for the rise in scattering at short wavelengths, while bmie lets the λ-4 behavior of the Rayleigh scattering account for the rise. Hence, the ratio falls.

IN SUMMARY

The optical properties of biological tissues can vary, and the successful measurement of these properties can also vary. Optical property measurements can be frustratingly tricky sometimes. The literature on tissue optical properties is still a bit ragged, and more work is needed. But those who have measured and published values for tissue optical properties have succeeded in defining the approximate range of property values for various types of tissues, which is a great aid to those who are developing novel optical devices or therapeutic protocols.