Unique im imaging contrast in in coronary ry artery ry dis - PowerPoint PPT Presentation

Unique im imaging contrast in in coronary ry artery ry dis iseases wit ith flu luorescence and ult ltrasound Laura Marcu Professor of Biomedical Engineering

Why FLIm Im-IVUS im imaging? IVUS FLIm - IVUS IVUS Morphology • Plaque burden • Calcification FLIm FLIm Composition • Collagen and Lipid content • Inflammatory cell infiltration • Maarek et al, Laser Surg Med , 2000 • Marcu et al, Ather Thromb Vasc Biol, 2001 • Marcu et al, Atherosclerosis , 2009 • Fatakdawala et al. J. Cardiovasc Trans Res , 2015

FLIm Im - IV IVUS The quest for demonstrating clinical value • Is FLIm-IVUS technologically possible? • What is FLIm-IVUS diagnostic value?

FLIm Im - IV IVUS The quest for demonstrating clinical value • Is FLIm-IVUS technologically possible? - Can be applied intravascularly in vivo? - Can FLIm operate fast enough? • What is FLIm-IVUS diagnostic value?

FLIm - IV IVUS catheter system Modified motor drive • 3.7 Fr catheter Image of • Derived from Boston fiber core 810 µm Scientific OptiCross ™ IVUS catheter 40 MHz J. Bec et al. Scientific Reports , 2017 transducer

How does FLIm work? IVUS Cable FLIm-IVUS motor drive Single excitation/ collection fiber optic UV beam Ultrasound ~250 m m – penetration Vessel wall depth Dichroic mirror Endogenous fluorophores • Collagen type I • Collagen type III • Elastin • Cholesteryl linoleate • Cholesteryl oleate 355 nm • LDL Pulsed laser • Fibrin <1 ns • Ceroid

How does FLIm work? IVUS Cable FLIm-IVUS motor drive Single excitation/ collection fiber optic UV beam Ultrasound Vessel wall Dichroic mirror FLIm detection module Imaging DM1 PMT PC 390 nm DM2 450 nm DM3 355 nm Sun et al. Optics 540 nm Pulsed Letters, 2008 laser Yankelevich et al. 630 nm Review of Scientific Instruments , 2014 Delay fiber Fast digitizer

FLIm Im parameters – Ex Ex-vivo h human coronary ry art rtery ry demo Pull Back (mm) 20 𝐽 𝑗 Intensity (u.l.) 𝐽𝑆 𝑗 = .6 20 mm length ratio (mm) 4 σ 𝑙=1 𝐽 𝑙 1 2 5 seconds 3 4 0 0 20 ∞ 𝑢𝐽 𝑢 𝑒𝑢 Mean ׬ 0 < 𝑢 >= (ns) ∞ 𝐽 𝑢 𝑒𝑢 Lifetime 8 ׬ 1 (mm) 0 2 3 4 0 2 Channel 4 Channel 1 Channel 2 Channel 3 630 nm 390 nm 450 nm 540 nm 25000 locations (pixels)

FLIm Im - IV IVUS pull-back ~100 nJ/pulse; 20 KHz laser rep. rate; 4 x averaging 1800 rpm, 166 samples/turn 20 mm length 5 seconds scan human coronary vessel

FLIm Im 3D surf rface maps Reconstruction using IVUS morphology information 630 nm 390 nm 450 nm 540 nm ns 8 5 2

In In viv ivo vali lidation: : Pig ig heart Circumflex coronary Fluorescent markers Stented 7Fr guide section Fluorescent markers emission spectra Intensity (a.u.) Imaging core tip Guide wire Wavelength (nm) J. Bec et al. Scientific Reports , 2017

In In viv ivo pig ig coronary ry • ~100 nJ/pulse • 20 KHz laser rep. rate • 4x averaging • 1800 rpm  166 samples/turn • 20 mm/ 5 s scan • Dextran 40 bolus • 8 s duration • 5 cc/s J. Bec et al. Scientific Reports , 2017

FLIm Im - IV IVUS The quest for demonstrating clinical value • Is FLIm-IVUS technologically possible? • What is FLIm-IVUS diagnostic value?

Combined FLIm + IV IVUS Parameters Cla lassifi fication Results fr from as earlier stu tudy Support Vector Machine Classifier N = 16 ex-vivo coronary; 8 plaque classes (184 regions) • • Diffuse Intimal Thickening Thin Cap Fibroatheroma • • Pathological Intimal Thickening Fibrocalcified tissue • Thick Cap (>250 µm) Fibroatheroma • Fibrotic Tissue • • Thick Cap with Macrophage /Lymphocyte Lipid rich core of Fibroatheroma IVUS FLIm IVUS + FLIm Sensitivity 45 % 69 % 89.3 % Specificity 94 % 97 % 98.8 % Positive Predictive Value 61 % 88 % 88.8 % Bimodal FLIm + IVUS has better diagnostic value Ref. Fatakdawala et al. Journal of Cardiovascular Translational Research, 2015

FLIm Im - IV IVUS The quest for demonstrating clinical value • Is FLIm-IVUS technologically possible? • What is FLIm-IVUS diagnostic value? - What does FLIm “see”? - Can FLIm parameters predict specific biochemical makeup of coronary disease? - How does FLIm and IVUS complement each other?

Connecting dots: : FLIm Im-IVUS vs His istopathology Image human coronary vessels 2 mm rings Faxitron 0.25 mm levels sections H&E Histology Imaging planes for co- registration with histology 8 ns FLIm-IVUS Imaging 1 mm 25000 pixels!!! 2 ns CH3 Special staining actin (  - Movat’s Picrosirius Oil red O CD 68 red SMC) Pentachrome

Histologic grading by o’clock 540 nm

Resolving ECL + FC in in the in intima ECL+FC (N=56) FC only (N=868) Normal (N=7935) Movats Pentachrome CD68 Total arteries = 24

Resolving superficial calc lcium – lu lumen surf rface Superficial Ca (N=261) No superficial Ca (N=15544) ECL (N=928) ECL Calcium on the surface Lifetime Channel 2 Lifetime Channel3 Total arteries = 24

Development of f a foam cell ll predictor Foam cells – 540 nm Lifetime • 21 arteries in training set 8 • Random forest regression classifier 7 Lifetime (ns) • Parameters used in classifier: lifetime 6 and intensity ratios at 460 and 540 nm 5 wavelength bands 4 𝜍 = 0.71 3 2 0 5 10 20 30 40 % of foam cells in intima- from pathologist tabulation

Foam cell ll predictor testing Foam cell predictor Oil red O 0 7 Lifetime 540 nm Pullback distance (mm) 3 200 µm 16 Foam cell % 25 200 µm 0 ° 360 ° 2 Angle

Conclusion IVUS • Developed a fully integrated FLIm-IVUS catheter system (3.7 Fr) and demonstrated functionality of this bi-modal catheter system in-vivo in pig coronaries • Currently we develop an further optimized catheter version (3.2 Fr) based on freeform optics employing a 100 µm fiberoptic • Seek FDA approval for first in human studies 3.2 Fr • FLIm parameters can be used to automatically predict specific biochemical features (e.g. FC, ECL, etc). They can be depicted as abundance maps based on pathology and classification • FLIm derived information has the potential to augment IVUS images for a better prediction of distinct pathophysiologies of coronary vessels

Acknowledgements Biophotonics Laboratory Collaborators Jeff Southard Ken Margulies L. Maximilian Buja Deborah Vela Julien Bec Jennifer Phipps Ben Sherlock Shamira Sridharan Jakob Unger Jeny Shklover Alba Alfonso Clay Sheaff Anne Haudenschild Funding Jim McMasters Xiangnan Zhou Cai Li Brent Weyers Tianchen Sun R01 HL067377 Su Hyun Lyu Emmery Leighton Michael Agung