Multimodal Optical Imaging Combined with Radiomic Analysis for Fibrotic Cardiac Tissue Investigation

Abstract

Understanding the process of fibrotic scarring of the myocardium is critical for the diagnosis and risk stratification of life-threatening cardiac dysfunction. Complex changes in structure, composition, and conductivity occurring at different stages of fibrogenesis diversify the biomedical characteristics of the myocardium. We present a multimodal optical imaging approach including cardiac optical mapping (COM), optical coherence tomography (OCT), multiphoton microscopy (MPM), and line scan Raman microspectroscopy (LSRM) for multiparametric assessment of the myocardium with radiomic analysis to link electrophysiologic, morphologic, functional, and molecular changes in ischemic cardiac tissue and validate our results with histology. COM is used to map the electrical behavior across myocardial tissue. Second harmonic generation and two-photon excitation fluorescence imaging as MPM techniques provide additional unique contrast of collagen, the extracellular matrix, and cardiac cells, such as cardiomyocytes playing a critical role in cardiac fibrosis. Our machine learning model based on radiomic features extracted from MPM data addresses the need for automated fast high-throughput classification between healthy and pathologic cardiac tissues and achieved an accuracy of 0.99. In addition, LSRM assesses the molecular contrast and is used to evaluate the development stage of fibrotic scarring and multiclass classification by utilizing partial least-squares discriminant analysis, achieving sensitivity and specificity values of 0.94. OCT is used for fast navigation through the sample, for intermodal referencing, and easy coregistration between the complementary imaging techniques operating at different fields of view and resolutions ranging from cm² down to μm².

1

Workflow for myocardial analysis, integrating COM, OCT, MPM, and LSRM. Image processing includes radiomics of label-free SHG and TPEF images and PLS-DA of Raman spectra, for multiparametric data interpretation at different magnifications.

2

Multiscale multimodal analysis of a tissue pair (a, d) from a scar border region from the LV myocardium. (a) Cardiac tissue after noninvasive multimodal optical imaging performed on the red rectangle. (b) Low-magnification mode maximum intensity projection of the OCT image. (c) MT image of (b). Black arrows point to fibrotic tissue visible in the OCT and MT. Dashed and solid blue rectangles show the position of high-magnification OCT scans (g) and (l), respectively. (f) and (k) High-magnification OCT B-scans with blue lines showing the depth position of the OCT images in (g) and (l), respectively. Black lines in (g) and (l) show the lateral position of the OCT B-Scans (f) and (k), respectively. The purple rectangle in (g) shows the area scanned by LSRM. (h) and (m) MPM images with SHG channel (green) and TPEF channel (red) coregistered with the OCT images (g) and (l) and validated with MT in (i) and (n). (e) and (j) AT maps at pacing frequencies of 3 and 1.5 Hz, respectively. Isochrones are spaced 5 ms. The yellow arrow points to a possible re-entrant circuit. White arrows indicate a microvascular structure, and red arrows indicate a triangular-shaped collagen formation. White scale bars 200 μm, black 1 mm, and blue 5 mm.

3

Coregistered optical imaging of infarcted cardiac samples with radiomic analysis based on MPM contrast. (a) Representative MPM images used for training and testing of the ML model (upper row) for binary classification with the corresponding MT images (lower row). The SHG channel is colored in green, and the TPEF channel is colored in red. Scale bars 200 μm. (b) Box plot of ML training shows probability values ρ of each MPM image.

4

Multiclass Raman spectroscopy results: (a) Mean Raman spectra from muscle, necrotic, granulated, and fibrotic tissues. (b) PLS-DA score plot of LV1 vs LV2 showing clustering of necrotic, muscle, granulated, and fibrotic samples. (c) Loadings of LV1 and d) LV2.