Electrophysiological mapping of the epicardium via 3D ‐printed flexible arrays

FIGURE 1. Patient‐specific flexible electrode array design and characterization. (a) Example workflow to create mapping array for the right atrium (RA). A cardiac MRI was performed to obtain detailed whole‐chamber resolution of the myocardium. Heart chamber segmentation generated a 3D model of the epicardial surface with the RA highlighted in blue. A computer‐aided design (CAD) model was generated that matched the anatomic features of the epicardial surface, which was then used to 3D print a flexible epicardial shell. (b) Image of flexible linear electrode array, interelectrode spacing was optimized in the preoperative CAD environment. (c) Electrode arrays were attached to the surface of the epicardial shell to create a patient‐specific array capable of whole‐chamber epicardial mapping. (d) Mapping electrodes adopted the geometry of the epicardial surface to optimize signal acquisition. (e) (left) 3D model of the whole heart depicting the segmentation of the RA, (center) presurgical CAD rendering of the flexible epicardial shells and, (right) postsurgical photo of the ex vivo porcine heart overlaid with custom whole‐chamber flexible array. Ao, aorta; LA, left atrium; LV, left ventricle;…

FIGURE 2. Electromechanical characterization of 3D printed flexible whole‐chamber epicardial arrays. (a) Impedance spectrum of flexible electrodes after 1000 cycles of ±50% strain (b,c), in vivo characterization of relationship between systemic mean arterial blood pressure (MAP) and external force applied by the epicardial array over the right and left atrium, respectively. The region highlighted in blue indicates the range of force (1 ± 0.15 N) necessary to ensure proper electrode contact with the epicardium. Forces above 4 N caused significant hemodynamic compromise. (d) Epicardial arrays were 3D printed with flexible photopolymer resin allowing for dynamic adaptation to changes in epicardial morphology over the course of a cardiac cycle. (e) Strain analysis during the cardiac cycle demonstrated percentage deformation of the device over the period of a cardiac cycle (red, ±1 SD). Analysis of atrial anatomy of different hearts demonstrates a high degree of morphological variability. (f,g) Anterior view of right and left atrial segmentations obtained from three different cardiac MRIs. (h,i) Gaussian curvature analysis of respective right and left atrial epicardial shells highlight the…

FIGURE 3. In vivo cardiac mapping, pacing, and ablation with whole‐chamber electrode arrays. (a) CAD rendering of multiaxis articulating arm for positioning of whole‐chamber mapping arrays. (b) Articulating arms facilitate placement of the mapping array on the epicardial surface. (c) Surgical deployment of the mapping array over the left atrium via median sternotomy. (d) Example high‐density (16 × 16) flexible electrode array for the left atrium. Scale bar, 10 mm. (e) Demonstration of geometric conformity between the flexible array and LA epicardial surface. Scale bar, 10 mm. (f) Representative unipolar left atrial electrogram from the flexible array. (g) Isochrone activation map during sinus rhythm determined by (dV/dt)max. (h) Time‐lapse visualization of individual electrode activation at 5 ms time intervals demonstrates sequential action potential wave propagation during sinus rhythm. In addition to electrophysiological mapping, the electrode arrays were capable of delivering focused energy for the purpose of extra stimulus pacing and RF ablation. (i) Isochronal activation during extra stimulus pacing via two adjacent electrodes demonstrated retrograde action potential propagation…

FIGURE 4. Cardiac mapping of the left atrium during chronic AF. (a) Schematic showing pacing‐induced chronic AF in a porcine model. Yorkshire swine underwent right atrial pacing at 350 beats/min for 6 weeks prior to terminal cardiac mapping. (b) Isochrone activation maps demonstrate random fibrillatory activity and loss of coordinated action potential wave propagation when compared to sinus rhythm consistent with atrial fibrillation. (c) Representative electrograms from the black dashed box in (b) are characterized by low‐amplitude and high‐frequency signals. (d) Time‐lapse visualization of individual electrode activation at 10 ms time intervals demonstrates random fibrillatory pattern of atrial activation. (e) Frequency domain visualization of individual electrode dominant frequency (DF). (f) Representative power spectrums of three electrodes identified in e with varying DF.

FIGURE 5. Serial mapping of the ventricular epicardium during myocardial ischemia and beat‐to‐beat capture of premature ventricular complexes. (a) Representative optical images of the anterior right ventricle (RV) and left ventricle (LV) after median sternotomy. (b) The anterior interventricular artery (AIA) was permanently ligated with suture and ischemia confirmed by visible blanching of the ischemic area at risk (white dotted area). (c) Serial ventricular mapping of the anterior RV and LV was performed using flexible ventricular arrays. (d) Baseline isochrone map of ventricular epicardial activation during sinus rhythm. (e) Changes to body surface ECG morphology were noted after AIA ligation with increasing evolution of ST‐elevation over 60 min. (f) Isochrone maps obtained at baseline, and every 15 min after AIA ligation, provide spatial information on the evolution of myocardial damage after ischemic injury. Mapping of the ventricular epicardium was also performed with identification of premature ventricular contractions (PVC). (g) PVCs were observed on surface ECG, while epicardial mapping captured beat‐to‐beat differences in wave propagation during (h), sinus rhythm and (i) PVCs,…

FIGURE 6. Ex vivo epicardial mapping of human‐dilated cardiomyopathy. (a) Donor hearts from patients with history of ventricular tachycardia (VT) were mounted on a Langendorff‐isolated heart perfusion system for epicardial mapping. (b) Representative optical image of a Langendorff‐perfused human heart depicts severe dilation of the LV. (c) Representative electrogram depicting capture of irregular baseline rhythm, extra stimulus pacing, and subsequent induced monomorphic VT. Ventricular mapping captured differences in epicardial wave propagation between (g) baseline and (h) VT rhythms. Spatial activation maps at select time intervals demonstrate differences in the direction of action potential propagation at (f) baseline originating from the ventricular apex, compared to (g) VT originating from the lateral border of the LV. LA, left atrium; LAA, left atrial appendage; LV, left ventricle; PA, pulmonary artery.