Time-resolved NIRS and non-destructive Titolo presentazione - - PowerPoint PPT Presentation

Titolo presentazione sottotitolo

Milano, XX mese 20XX

Time-resolved NIRS and non-destructive assessment of food quality

Lorenzo Spinelli, Alessandro Torricelli

Dipartimento di Fisica – Politecnico di Milano

Winter College on Applications of Optics and Photonics in Food Science 20 February 2019

Outline

2

• Lecture 1 (Alessandro Torricelli) 9-10 am
• Basics of TD NIRS
• Lecture 2 (Lorenzo Spinelli) 10-11 am
• Application of TD NIRS to food quality assessment
• Coffee break 11:00-11:30 am
• Discussion 11:30-12:30
• What does affect TD NIRS?
• Questions & Answers

Outline

3

• Lecture 1 (Alessandro Torricelli) 9-10 am
• Basics of TD NIRS
• Introducing the PHOOD lab @ PoliMi
• Modelling light propagation in food
• CW and TD NIRS
• Instrumentation for TD NIRS

Politecnico di Milano (PoliMi) Teaching & Research University

4

PoliMi since 1863 1863 – 2019 156 years Engineering Architecture Design

Photonics for Health, Food, and Cultural Heritage Dipartimento di Fisica – Politecnico di Milano

5

Professor Emeritus: Rinaldo Cubeddu Full professors: Antonio Pifferi Paola Taroni Alessandro Torricelli Gianluca Valentini Associate Professors: Andrea Bassi Daniela Comelli Davide Contini Cosimo D’Andrea Alberto Dalla Mora Assistant Professors: Rebecca Re Laura Di Sieno IFN-CNR Lorenzo Spinelli (CNR) Andrea Farina (CNR) Austin Nevin (CNR) Post-Docs: Lina Qiu Alessia Artesani + PhD Students (11) + Facilities (mechanic and electronic workshop)

Photonics for Health, Food, and Cultural Heritage Dipartimento di Fisica – Politecnico di Milano

6 Health

• In vivo Tissue Spectroscopy
• Optical Mammography
• Tissue Oximetry and Functional Imaging of the Brain
• Fluorescence Lifetime Imaging in biology and medicine

Cultural Heritage

• Photoablation and Material Processing
• Fluorescence Spectroscopy and Imaging
• Multispectral Imaging and Colorimetry

Food

• nondestructive assessment of

internal defects by pulsed NIR

• nondestructive maturity

assessment at harvest

Photonics for Health, Food and Cultural Heritage Laboratories

7

• functional near infrared spectroscopy
• fNIRS Lab
• diffuse spectroscopy
• DiffS Lab
• ptical mammography
• Mammot Lab
• molecular imaging
• Molim Lab
• near infrared spectroscopy for food
• NIRf Lab
• imaging spectroscopy for cultural heritage
• ARTIS Lab
• ultras for biomedicine
• UB Lab

Time-resolved systems

• mode-locking of dye, gas and solid state lasers
• time-correlated single-photon counting (TCSPC)
• time-gated imaging

Spectral-domain

• tunable laser sources
• broadband detectors

Spatial-domain

• scanning systems or camera
• multi-channel systems

Temporal-domain

• fast acquisition rate

8

Can light penetrate biological tissues?

St Joseph (1642) Louvre, Paris Georges de La Tour (1593 – 1652)

Thanks to Marco Ferrari (UnivAQ)

Light absorption in the near infra-red (NIR): biological tissue

9 The therapeutic and diagnostic window

Light absorption in the near infra-red (NIR): fruit

10

Visible (VIS) and near infrared spectroscopy (NIRS) : continuous wave (CW) approach

11

Rich Ozanich, Berkeley Instruments Inc., Richland, WA

VIS: 400-700 nm (nondestructive assessment of EXTERNAL properties) NIR: 700-3000 nm (nondestructive assessment of INTERNAL properties)

Visible (VIS) and near infrared spectroscopy (NIRS) : continuous wave (CW) approach

12

HL200 Ocean Optics ≈ 1000 € USB4000 Ocean Optics ≈ 2000 € Notebook ≈ 1000 € DA-meter, courtesy of P. Rozzi, Sinteleia (Italy) Spider, courtesy of Manuela Zude ATB Potsdam (Germany)

Light propagation in diffusive media: absorption and scattering

13 clear diffusive

Scattering: related to tissue structure Scattering coefficient: µs = 1/ls (cm-1) Absorption: related to tissue components Absorption coefficient: µa = 1/la (cm-1)

 interplay between light absorption and light scattering

IIN IOUT = IIN exp(-µa L) (a) clear medium L

Lambert-Beer law Light attenuation in a THIN diffusive medium

14

L I I A

a in

• ut

µ =         = ln CL A ε =

Lambert Beer

C

a

ε µ =

Lambert-Beer Clear medium (b) turbid medium L IIN IOUT = IIN exp(-µt L) µt = µa+ µs

( )L

I I A

s a in

• ut

µ µ + =         = ln

• Valid only if µtL < 1 (Single scattering regime)
• Scattering coefficient is unknown
• Pathlength is unknown

Scattering coefficient Diffusive medium

Modelling photon migration in diffusive media Radiative Transport Equation (RTE)

15 s Conservation of energy in a small volume dV in a given direction s Photons out Photons in (1) Photons scattered to another direction (3) Photons scattered from another direction (s’) to direction of interest (s) (4) dV (5) Light source Photons absorbed (2)

ε

π

+ ∫ )

• ′

µ + µ − µ − ∇

• −

= ∂ ∂

4 s s a

d ˆ ˆ ( v v v ˆ v Ω n p n n n t n s s s

n = photon (angular) density

[ n dV dv is the expected number of photons in the volume dV about r, with velocity in dv about v, at time t]

The RTE for the Radiance

16

) + ∫ ) )

• µ

+ ) µ + µ − ) ∇

• −

= ∂ ) ∂

′ ′

t q t L p t L t L t t L , ˆ , ( d , , ( ( , ˆ , ( ) ( , ˆ , ( ˆ , ˆ , ( v 1

4 s s a

ˆ ˆ ˆ

s r r s r s r s s r

s s s

π

Ω

[ W m -2 sr-1] [ W m -3 sr-1]

) , ˆ , ( v ) , ˆ , ( t n h t L s r s r ν = ) , ˆ , ( ) , ˆ , ( t h t q s r s r ε ν =

Solutions of the RTE

17 RTE Expansion methods

• PN approximation:

P0 = Diffusion (1989) P1 = Diffusing Wave P3 = … Stochastic methods

• Monte Carlo (1987)

Discretisation methods

• Discrete ordinates
• 2-flux or Kubelka-Munk
• Adding-double method
• Finite Element Method (1993)

Hybrid methods

• Paasschens (1997)

Note: The year represents the first use of the method in the field of Biomedical Optics

PN Approximation

18

• Expansion of the RTE terms into spherical harmonics to separate the

position and directional variables

) ˆ ( ) , ( 4 1 2 ) , ˆ , (

, ,

s r s r

m l m l l l l m

Y t l t L φ π ∑ ∑ + =

∞ = + − =

• PN approximation truncates the expansion to the N-th term obtaining a

set of N+ 1 independent equations

• P1 approximation is typically used in Photon Migration studies

P1 Approximation

19

µ π π g p 4 3 4 1 ) ˆ ˆ ( + ≈

• ′ s

s

( )

t S t q , 4 1 ) , ˆ , (

0 r

s r π ≈

( ) ( ) s

r J r s r ˆ , 4 3 , 4 1 ) , ˆ , (

• +

Φ ≈ t t t L π π ) + ) Φ µ − )

• −∇

= ∂ ) Φ ∂ t S t t t t , ( , ( v , ( , ( v 1

a

r r r J r

∫

= Φ

π 4

Ω d t L t ) , ˆ , ( ) , ( s r r

∫

=

π 4

Ω d t L t ) , ˆ , ( ˆ ) , ( s r s r J

Inserting in the RTE and integrating over Ω we get Inserting in the RTE, multiplying by and integrating over Ω we get

) µ′ + µ − ) Φ ∇ − = ∂ ) ∂ t t t t , ( ) ( , ( 3 1 , ( v 1

s a

r J r r J

(1) (2)

s s

) 1 ( µ − = µ′ g

Reduced scattering coefficient [ m -1]

s ˆ

Fluence Flux

Diffusion Equation

20 Under the assumptions

) µ′ + µ − ) Φ ∇ − = ∂ ) ∂ t t t t , ( ) ( , ( 3 1 , ( v 1

s a

r J r r J

s

v , ( , ( 1 µ′ << ∂ ) ∂ ) t t t r J r J

The relative variation of the flux is smaller than the scattering rate! v= 0.03 cm/ ps, µ’s= 10 cm -1, v µ’s = 0.3 ps-1

) Φ ∇ − = ) t D t , ( , ( r r J

( )

3

1/

s

D

′ µ

=

We obtain the Fick’s Law Diffusion coefficient [ m]

) = ∂ ) Φ ∂ + ) Φ µ + ) Φ ∇ − t S t t t t D , ( , ( v 1 , ( , (

a 2

r r r r

Substituting into (1) we obtain the Diffusion Equation

Steady state (or continuous wave, CW) Diffusion Equation

21

) = ) Φ µ + ) Φ ∇ − r r r ( ( (

a 2

S D

Point source solution for infinite medium

) = ) r r ( ( δ S         − µ − − = ) Φ r r r r r D exp D 4 1 (

a

π

r r 0 O r-r 0 Neglect time dependence

CW NIRS 1/2

22 Point source solution

) − − ) = ) ( ( ( z z S δ δ r

r z r z z0 z0 = (µs’) -1 = ls z0

( ) ( ) ( ) ( )2

2 2 2 a 2 2 2 2 a

D exp D 4 1 D exp D 4 1 ( z z r z z r z z r z z r + +         + + µ − − − +         − + µ − = ) Φ π π r

Collimated source

) Φ = ) Φ = ) Φ r,z ,z r, ( ( ( ϕ r

Cylindrical coordinate system with radial simmetry

CW NIRS 2/2

23 r z r

( ) ( ) ( ) ( )2

2 2 2 a 2 2 2 2 a

D exp D 4 1 D exp D 4 1 ( z z r z z r z z r z z r + +         + + µ − − − +         − + µ − = ) Φ π π r

( )

2 2 2 2 a 2 2 a

D exp 2 1 (

1 D

z r z r z D R

z r z

+         + µ − = = Φ ∇ − = )

        + + µ

π r r

Number of photons per unit area exiting from the tissue at a distance r from the source Fluence Reflectance

a a 2

exp D 1 ( 2

1 D

r R r

z r

π µ − )

      µ   ≈ +      

r

2 2 2

r z r

≈

+

Things to know about CW NIRS 1) Coupling of absorption and scattering

24 µ’s r µa r µ’s , µa r log10R log10R

a ' s a

D µ 3µ = µ

2 2 2 2 a 2 2 a

D exp 1 D 2 1 ( z r z r z r z R +         + µ −           + + µ = ) π r

Coupling of µa and µs’ !! In a homogeneous medium it is not possible to estimate both µa and µs’ with a single distance CW measurement

Things to know about CW NIRS 2) Penetration depth depends on µs’, ρ, and µa

25 12 mm 25 mm 15 mm 9 mm 35 mm 15 mm

F.Martelli et al. Scientific Reports 6:27057 (2016)

Things to know about CW NIRS 3) no difference between superficial and deep layers

26

0.01 mm-1

10 mm Reflectance Contrast = (RPERT-RHOM)/RHOM 10 mm 10 mm 10 mm

0.04 mm-1 (+400%)

10 mm 10 mm r r (mm) r (mm) Single distance r r

0.0125 mm-1 (+25%)

Things to know about CW NIRS 4) multi distance CW might be a solution

27 r 35 mm 15 mm 5 mm 3.5 mm 10 mm 10 mm Short is short enough, long is long enough Scalp signal clearly separated form cortex SRS = space resolved spectroscopy

Things to know about CW NIRS 4bis) «strange» multi distance CW approaches

28 Measure of Tissue Oxygenation Index (TOI)

2 a a

D exp D 2 1 ( r r z R         µ −         µ ≈ ) π r

a s

2 ( ( ln (

µ ′ 3µ

+ ≈ ∂ ) ∂ ) − = ) r r r r r A R A

( ) ( )

λ λ h k − ≈ µ 1

'

s

( )

( )

      µ

− ∂ ∂ − ≈ r r A h k 2 10 ln 1 3 1

a

λ

λ Approximation on spectral dependence of scattering coefficient

( ) ( )

           µ       µ

+ ≈ + ≈ Hb O HHb k k Hb O HHb k k

Hb O HHb Hb O HHb 2 2 a 2 1 a

2 2 2 1 2 1

λ ε λ ε λ λ ε λ ε λ

   = = ... ...

2Hb

O HHb

k k

2 2 2 2 2

SO Hb O HHb Hb O Hb O HHb Hb O TOI

k k k

= + = + =

Hamamatsu SRS (NIRO 300) and others: multi distance with LONG distances ONLY Still unable to uncouple superficial from deep signal!

Things to know about CW NIRS 4ter) «strange» multi distance CW approaches

29

KUL SRS: multi distance with SHORT distances ONLY Signal only from superficial layers!!

D62

Time Domain Diffusion Equation

30

) = ∂ ) Φ ∂ + ) Φ µ + ) Φ ∇ − t S t t t t D , ( , ( v 1 , ( , (

a 2

r r r r

Point source solution for infinite medium

) = ) , ( , ( r r δ t S         µ − − − = ) Φ vt 4Dvt exp Dvt) v(4 , (

a 2 3/2

• r

r r π t

r r 0 O r-r 0

TD NIRS 1/2

31 Point source solution

) − − ) = ) , ( , ( , ( z z t S δ δ r

r z r z z0 z0 = (µs’) -1 = ls z0

( ) ( )

        µ − + + − −         µ − − + − = ) Φ vt 4Dvt exp Dvt) v(4 vt 4Dvt exp Dvt) v(4 , (

a 2 2 3/2

• a

2 2 3/2

• z

z r z z r t π π r

Collimated source

) Φ = ) Φ = ) Φ r,z,t ,z,t r, t ( ( , ( ϕ r

Cylindrical coordinate system with radial simmetry

TD NIRS 2/2

32

( ) ( )

        µ − + + − −         µ − − + − = ) Φ vt 4Dvt exp Dvt) v(4 vt 4Dvt exp Dvt) v(4 , (

a 2 2 3/2

• a

2 2 3/2

• z

z r z z r t π π r

( ) (

)

( ) (

)

, , , ,

, ( , ( , (

ρ ρ = = =

) Φ ∇ − ) = )

z r z r

t D t t R r r J r

r z ρ

        µ − + − = ) vt 4Dvt exp t Dvt) (4 , (

a 2 2 5/2

• 3/2
• z

z t R ρ π ρ

Photon Diffusion TD solutions for other geometries

33 Infinite Semi-infinite Slab Parallelepiped Sphere Cylinder

F .Martelli et al. Photon migration through diffusive media: Theories and software (SPIE book, 2010)

2 layer inhomogeneity N layers

Time domain NIRS

34

Things to know about TD NIRS 1) Uncoupling of absorption and scattering

35 In a homogeneous medium it is possible to estimate both µa and µs’ with a single distance TD measurement We can work on normalized TD curves

Things to know about TD NIRS 2) Penetration depth depends on µs’ and t (NOT on µa or ρ)

36 Null distance approach feasible in the TD!

F.Martelli et al. Scientific Reports 6:27057 (2016) A.Torricelli et al. Phys Rev Lett 98 (1995)

Things to know about TD NIRS 3) difference between superficial and deep layers

37

0.990 0.995 1.000 1.005 1.010 1.015 1.020

• 20

20 40 60

time (s) intensity (a.u)

late time-gate early time-gate

Depth resolution is related to photon time-of-flight I

time

ρ

Steinbrink et al. Phys Med Biol 46:879-896 (2001) Del Bianco et al. Phys Med Biol 47:4131-4144 (2002)

• mean penetration depth does NOT depend on µa and ρ
• mean penetration depth does depend on µs

’ and t

Early photons Late photons

Things to know about TD NIRS 4) multi distance TD ensures better accuracy

38

ρ = 2, 3, 4, 5, 6 cm

0.0 0.1 0.2 0.3 0.4 0.5 700 750 800 850 900 950 1000 wavelength (nm) absorption (cm-1) 2 3 4 5 6 2 4 6 8 10 700 750 800 850 900 950 1000 wavelength (nm) reduced scattering (cm -1) 2 3 4 5 6

Heterogeneous sample

Obviously the multi distance approach introduces further requirements on the setup (e.g. Multi detector, optical switch, ...)

Evolution of TD NIRS systems

39 Generation Light sources Photo-detectors Acquisistion system 1st (1990-2000) > 500,000 € gas lasers dye laser solid state laser microchannel plate photomultiplier (PMT) electronic chain for TCSPC with NIM module 2nd (2000-2010) > 100,000 € semiconductor laser heads with external RF driver compact metal channel dynode PMT TCSPC electronic board 3rd (2010-2015) > 50,000 € supercontinuum fiber laser hybrid PMT TDC module with USB controller

TD NIRS systems Time correlated single photon counting (TCSPC)

40 ++ temporal resolution +++ sensitivity

http://www.becker-hickl.com/literature.htm#handb

Becker-Hickl GmbH

• cost

1st generation TD NIRS Broadband laboratory set-up

41 Fully automated system spectral range: 540 -1100 nm

Pifferi et al., Review of Scientific Instrument 78, 053103 (2007)

2nd generation TD NIRS Time resolved optical mammograph

42

2nd generation TD NIRS Time resolved optical mammograph

43

Patient # 4 7 , oblique view

age: 36 y thickness = 5.7 cm Lesion size = 3.0 cm Lesion type = tumor

S0 2 tHb R L R L 52%-89% 17 - 91 µM 62%-95% 16 - 66 µM

Type View Cases Detection rate Failures Corrected detection rate Cancer 2 41 73% 80% 1 9 89% 4 96% 6 11% Cyst 2 59 72% 8 83% 1 5 78% 3 90% 18 22% Fibroadenoma 2 17 33% 2 39% 1 5 43% 5 50% 29 57%

Taroni et al., TRTC 4:527-537 (2005).

Clinical study (225 lesion)

2nd generation TD NIRS Multichannel dual wavelength system

44

clock µCHIP delay 2x2 fused splitter 50% 50% 2x4 fused splitter R1 R2 R3 R4 S16 S9 S8 S1 sync 820 nm 690 nm Laser driver variable ND variable ND 1x9 fiber switch 1x9 fiber switch 4 anodes PMT-1 4 anodes PMT-2 4 anodes PMT-3 4 anodes PMT-4 4 ch router-1 4 ch router-2 4 ch router-3 4 ch router-4 8 ch amp-1 8 ch amp-2 F1 F16 clock TCSPC-1 TCSPC-2 TCSPC-3 TCSPC-4

Contini et al., Opt Expr, 14: 5418-5432(2006).

2 wavelengths 16 channels 50 ms acquisition time “Medical device” approved for clinical investigations

2nd generation TD NIRS Multichannel dual wavelength system

45

In collaboration with: I.Gilioli, S.Franceschetti, F . Panzica, E.Visani @ IRCCS Besta Milan, Italy

healthy ULD patients

A, D: O2Hb and HHb time-courses in the most reactive channel and the corresponding GLM activation maps. B, E: BOLD signal extracted from the active cluster and fMRI maps.

2nd generation TD NIRS The BabyLux device for blood oxygenation and blood flow

46

• The BabyLux project aims to provide a precise, non-

invasive and robust integrated system to continuously monitor cerebral oxygen metabolism and blood flow in extremely preterm newborns.

• It will enable neonatologists to prevent the neurological

damage due to lack of oxygenation in the brain that not infrequently is accompanied at premature birth.

• Started
• n

1st January 2014, 40 months 9 partners

GA no. 620996 CIP ICT-PSP

http://www.babylux-project.eu

Microvascular, local, cerebral blood oxygen saturation blood flow

Photonics for Health  Photonics for Food

47 H BC The majority of foods and vegetables produce are diffusive media, like human tissues

1st generation TD NIRS Broadband laboratory set-up

48 Fully automated system spectral range: 540 -1100 nm

Pifferi et al., Review of Scientific Instrument 78, 053103 (2007)

2nd generation TD NIRS Dual-wavelength transportable system

49

laser heads 750 nm 670 nm d r i v e r fiber optic swtch PMT amp TCSPC sync filters and optics Cubeddu et al., Appl Spectroscopy 55:1368-1374 (2001) Torricelli et al. Sens. & Instrumen. Food Qual. 2:82–89 (2008)

3rd generation TD NIRS Time resolved multi wavelength spectrometer

50

Laser source: Supercontinuum fiber laser − spectral range: 450-1600 nm − power: 6 W − frequency: 40 MHz Wavelength selection: Filter wheel − spectral range: 540-900 nm Detector: Hybrid PMT − no afterpulse − time response: 250 ps TCSPC SYNC CFD Time resolution: Time-Correlated Single-Photon Counting: − high dynamic range − suitable for faint signal − time resolution: up to 1 ps SYNC

Supercontinuum laser

Filter wheel grin fiber ∅ =100μ m

Sample Large area detector

Objective 10x Step-index fiber ∅ =1mm Filter wheel

Supercontinuum laser Supercontinuum laser

Filter wheel grin fiber ∅ =100μ m

Sample Sample Large area detector Large area detector

Objective 10x Step-index fiber ∅ =1mm Filter wheel

Photonics for Food @ PoliMi & IFN-CNR Main applications

51 Non destructive optical characterisation of internal optical properties and correlation with quality parameters

• Basic studies in apples, kiwifruits, nectarines, tomatoes, …
• Changes in optical properties during growth in Elstar apples and Tophit plums
• Texture in Jonagored apples, Braeburn apples and Pink Lady apples during storage

Non destructive detection of internal disorders and defects

• Browning in Granny Smith apples, Braeburn apples and Conference pears
• Watercore in Fuji apples
• Mealiness in Braeburn apples and Jonagored apples
• Chilling injuries in Jubileum plums and Morsiani nectarines

Non destructive assessment of fruit maturity at harvest and correlation with quality parameters

• Basic studies in apples, kiwifruits, nectarines, peaches, mangoes, …
• Sensory attributes, aroma composition, ethylene production Ambra nectarines
• Softening prediction (based on biological age) in Spring Bright nectarines

and in Tommy Atkins mangoes

The End

52

Outline

53

• Discussion 11:30-12:30
• What does affect TD NIRS?
• Questions & Answers

What does affect TD NIRS?

54

• Instrument response function
• Signal to noise ratio
• Temporal stability
• Range of optical and geometrical parameters
• Availability of systems

Instrument response function 1/4

55 x(t) h(t) y(t) = x(t)*h(t) Linear causal systems: the output y(t) is the convolution of the input x(t) with the impulse response h(t) Laser Injection fiber Sample Collection fiber Detector Timing electronics δ(t) TRS(t) δ(t) IRF(t) h1(t) h2(t) R(t) h3(t) h4(t) h5(t) TRS(t) = δ(t) * [h1(t)*h2(t)*R(t)*h3(t)*h4(t)*h5(t)] = δ(t) * [h1(t)*h2(t)*h3(t)*h4(t)*h5(t)]*R(t) TRS(t) = IRF(t) * R(t) h1(t) h2(t) h3(t) h4(t) h5(t) IRF(t) R(t) TRS(t) = R(t)*IRF(t)

Instrument response function 2/4

IRF(t) R(t) TRS(t) = R(t)*IRF(t) Examples of IRF Pulsed laser diode Hybrid PMT 1mm ∅ GIGA POF, 2m length Pulsed laser diode Metal Channel Dynode PMT 3mm ∅ step index fiber bundle, 1.5m length Supercontinuum fiber laser SiPM 1mm ∅ GIGA POF, 2m length

Instrument response function 3/4

Effect of IRF

• Limiting the maximum µa and the minimum µs’ that can be fitted
• Increase in absorption or decrease in scattering results in narrowing of the TRS curve
• If IRF is broad we can not resolve those changes

0.001 0.010 0.100 1.000 1000 2000 3000 4000 5000

time (ps) intensity (a.u.)

laser pulse TRS data model

Instrument response function 4/4

Effect of IRF

• Reducing depth discrimination
• IRF introduces uncertainty in the time of arrival of photons (i.e. photons detected at

time τ can be originated at time τ±∆T, with ∆T depending on the IRF width)

• Coupling superficial and deep contributions

I

time

ρ

Early photons and Late photons well separated

I

time Narrow IRF (<250ps) Broad IRF (>500ps) Early photons and Late photons might overlap!

ρ

Signal to noise ratio

Effect of SNR

• Limiting the dynamics of the TRS curve
• Effect on fitting results and contrast
• Limiting the penetration depth
• Effect on contrast

107 collected photons 106 105 104 103

Temporal stability

• Time drift (in the range of 10-100 ps!) introduces errors in the fitted µa and µs’
• ↓ τ0  ↑ µs’  ↑ µa
• ↑ τ0  ↓ µs’  ↓ µa

How to reduce the effect of temporal instability? Acquire several IRF curves and/or measure calibrated phantoms during the experiment effect of time drift effect of time drift after correction

Range of optical and geometrical parameters

TD NIRS measure the broadening of a light pulse while traveling in a medium Broadening is affected by

• distance between source and detector

> 1 cm

• value of the reduced scattering coefficient

> 5 cm-1

• value of the absorption coefficient

< 0.5 cm-1

• temporal width of the IRF

<< 500 ps If possible the TD NIRS system should be tailored to the application

Range of optical and geometrical parameters

0.0 0.1 0.2 0.3 0.4 0.5 650 700 750 800 850 900 950 1000

wavelength (nm) absorption (cm

• 1)

apple kiwifruit 5 10 15 20 25 650 700 750 800 850 900 950 1000

wavelength (nm) transport scattering (cm

• 1)

apple kiwifruit Cubeddu et al., Applied Optics 40:538-543 (2001)

Optical properties can change with wavelength!

Availability of TD NIRS system

63

Where to buy a TD NIRS system? How much for a TD NIRS system ?

Nowhere !!?? A lot !!!

Evolution of TD NIRS systems

64

• A. Pifferi et al.,“New frontiers in time-domain diffuse optics, a review,” J. Biomed. Opt. 21(9), 091310 (2016), doi: 10.1117/1.JBO.21.9.091310.

Next generation TD NIRS Compact two wavelengths system

65

M.Buttafava et al., “A compact two-wavelengths Time-Domain NIRS system based on SiPM and Pulsed Diode Lasers”, IEEE Photonics Journal 9(1), 7800114 (2017)

size 200 x 160 x 50 mm3 total power consumption lower than 10 W (ready for battery operation)

TD NIRS expectations

66

• A. Pifferi et al.,“New frontiers in time-domain diffuse optics, a review,” J. Biomed. Opt. 21(9), 091310 (2016), doi: 10.1117/1.JBO.21.9.091310.

Access to TD NIRS systems at PHOOD lab @PoliMi

67

LaserLab-EUROPE project

Center for Ultrafast Science and Biomedical Optics Politecnico di Milano - Dipartimento di Fisica Milan, Italy

European Large Scale Facility since 2002

Access to infrastructure (limited to European researchers) Full reimbursment of travel and accomodation expenses http://www.laserlab-europe.net/ alessandro.torricelli@polimi.it

Summary

68 Single channel TD NIRS

• No need to use data from literature (estimate of µs’ is equivalent to pathlength)
• Absolute value for absorption coefficient: µa , not just ∆µa
• Discrimination of superficial and deep signals by means of early and late photons
• Less influenced by contact, movement artefact, signal changes

 NOT YET COMMERCIALLY AVAILABLE (just wait ...) Single channel CW NIRS

• Coupling of absorption and scattering parameters
• Coupling of superficial and deep signals
• Need of photon pathlength from literature for quantitative analysis
• Strongly influenced by contact, movement artefact, signal changes

 NOT USEFUL Multi-channel CW NIRS

• Discrimination of absorption and scattering parameters
• Uncoupling of superficial and deep signals (if short and long channels are used!)
• Somewhat influenced by contact, movement artefact, signal changes

 POTENTIALLY USEFUL

Acknowledgments Collaborators & Grants

69