Near-infrared optical spectroscopic system for characterizing cardiac substrates

Patients with atrial fibrillation (AF) require modification of ectopic electrical activity to avoid potentially fatal health complications. Catheter ablation therapy is a minimally invasive procedure to create tissue necrosis, called lesion, in areas of abnormal electrical activity. However, significant proportion of patients require repeat treatment from AF recurrences in part by electrical reconnection from incomplete lesions and conduction recovery. Current therapeutic approaches are limited by reliable methods to assess AF structural substrates and validate lesion sufficiency during procedures. In recent years, development of near-infrared optical spectroscopy has provided a non-invasive method to evaluate biological tissue. Near infrared spectroscopy (NIRS) is an optical technique that enables direct characterization of pathological tissue based on the absorption of major chromophores and light scattering. In this thesis, we explore the use of near-infrared optical spectroscopy to identify AF substrates and quantify lesion formation to improve treatment efficacy. First, we developed a near-infrared multispectral imaging system and present a model to assess lesion adequacy by direct visualization of cardiac lesions through an endoscope-integrated probe. Then, a custom single fiber integrated radiofrequency (RF) ablation catheter was fabricated to track irrigated lesion progression real-time on ex vivo swine hearts. A machine learning model was introduced to predict lesion size and transmurality. Additionally, we assess the feasibility of near-infrared spectroscopy by fabricating a NIRS-integrated open-irrigation RF ablation catheter and an algorithm to assess lesion dimensions based on key features derived from NIRS measurements. Using this model, we demonstrate real-time tracking of irrigated lesion delivery in both ex vivo and in vivo swine model. Lastly, we show left atrial endocardial mapping with NIRS-integrated RF mapping catheter to assess AF structural substrates. We present a classification algorithm for important AF structural substrates, such as pulmonary vein sleeve, normal myocardium, ablated tissue, and fibrosis, and a regression model to validate lesion adequacy. A near-infrared spectroscopy-based techniques to localize structural complexities and validate lesion sufficiency at the catheter tip could enhance the understanding of underlying AF substrates and improve treatment efficacy.