tvercaut
commented
1 year ago In case it's easier to run iad without plotting etc, here is a small example illustrating the fitting issue with iad version >= 3.11.5
Input file: pbinput.rxt
IAD1
#
# Tests using measured values for M_R and M_T
# by Anisha Bahl
#
1.41 # Index of refraction of sample
1.0 # Index of refraction of top slide
2.65 # [mm] Thickness of sample
0.0 # [mm] Thickness of slides
10.0 # [mm] Diameter of illumination beam
0.98 # Reflectance of calibration standard
1 # [mm] Number of spheres used during each measurement
# Reflection sphere properties (unused if n_spheres=0)
100.0 # [mm] Sphere Diameter
21.8 # [mm] Sample Port Diameter
18.3 # [mm] Entrance Port Diameter
8.0 # [mm] Detector Port Diameter
0.98 # Reflectivity of the sphere wall
# Transmission sphere properties (unused if n_spheres=0)
100.0 # [mm] Sphere Diameter
18.3 # [mm] Sample Port Diameter
0.0 # [mm] Entrance Port Diameter
8.0 # [mm] Detector Port Diameter
0.98 # Reflectivity of the sphere wall
2 # [mm] Number of measurements
# M_R M_T
640.0 0.63162425 0.18490826999999999 #0.9
630.0 0.63364865 0.18168763 #0.9Command line call
iad -X -i 8 -g 0.9 pbinput.rxtoutput
# Inverse Adding-Doubling 3-12-0 (21 Jun 2023)
#
# Beam diameter = 10.0 mm
# Sample thickness = 2.650 mm
# Top slide thickness = 0.000 mm
# Bottom slide thickness = 0.000 mm
# Sample index of refraction = 1.4100
# Top slide index of refraction = 1.0000
# Bottom slide index of refraction = 1.0000
#
# Fraction unscattered refl. in M_R = 100.0 %
# Fraction unscattered trans. in M_T = 100.0 %
#
# Reflection sphere
# sphere diameter = 100.0 mm
# sample port diameter = 21.8 mm
# entrance port diameter = 18.3 mm
# detector port diameter = 8.0 mm
# wall reflectance = 98.0 %
# standard reflectance = 98.0 %
# detector reflectance = 0.0 %
#
# Transmission sphere
# sphere diameter = 100.0 mm
# sample port diameter = 18.3 mm
# entrance port diameter = 0.0 mm
# detector port diameter = 8.0 mm
# wall reflectance = 98.0 %
# standard transmittance = 100.0 %
# detector reflectance = 0.0 %
#
# M_R and M_T were measured using the comparison (dual-beam) method.
# Single sphere corrections were used with light incident at 8 degrees from the normal.
# The inverse routine adapted to the input data.
#
#
# AD quadrature points = 12
# AD tolerance for success = 0.00010
# MC tolerance for mu_a and mu_s' = 0.010 %
# Photons used to estimate lost light = 100000
#
# Measured M_R Measured M_T Estimated Estimated Estimated
##wave M_R fit M_T fit mu_a mu_s' g
# [nm] [---] [---] [---] [---] 1/mm 1/mm [---]
640.0 0.6316 0.6956 0.1849 0.2171 0.0202 2.6936 0.9000 # *
630.0 0.6336 0.5697 0.1817 0.2151 0.0204 2.7408 0.9000 # * 
scottprahl
anishabahl
Dear Sir/Madam/Mx, I have found the IAD software to be invaluable to my PhD (in King's College London supervised by Prof Tom Vercauteren) and am very grateful for its availability and documentation! We have found that using the latest versions of the software (versions 3.11.5 onwards) gives odd fits to the measured data in the output files. Examples of this can be found attached and in the MWE_phantom_IAD_version_discrepancy.ipynb here along with some example measurements. When using the dual-beam settings (as is relevant to our set-up) the absorbance and reduced scattering coefficients remain largely the same, however. We would be very grateful for any explanation of this or advice for how to improve these results! Many thanks, Anisha Bahl
