Confocal microscopy-based estimation of intracellular conductivities in myocardium for modeling of the normal and infarcted heart

Figures

• 

  Fig. 1.. Composition and 3D reconstruction of myocardium from control animal. (A) Signals for WGA, DAPI, Cx43, vimentin, and α-SMA labeling are depicted in green, blue, magenta, cyan, and yellow, respectively, in a slice from an image stack. Arrows indicate interlaminar cleft spaces. Most Cx43 signal is found at end-to-end and side-to-side myocyte junctions. (B–E) Zoom-ins for the white boxes in (A) display representative microstructural features. (B) Cx43 labeling at the end-to-end junction of two myocytes. (C) End-to-end Cx43 labeling and adjacent vimentin-positive cell fragment in the interstitium. (D) End-to-end Cx43 labeling and vimentin-positive cell. (E) Side-to-side Cx43 labeling between two myocytes. (F) Reconstructed myocytes from the image stack. Myocytes were colored individually. The black line segments mark the edges and corners of the image stack. (G) Reconstructed non-myocyte contributors to the myocardium with ES in blue, vessels in red, and fibroblasts in cyan. The ES visualization is clipped in half for visibility of other tissue constituents.

• 

  Fig. 2.. Composition and 3D reconstruction of myocardium from MI animal. (A) WGA, DAPI, Cx43, Vimentin, and α-SMA in green, blue, magenta, cyan, and yellow, respectively in a slice from MI image stack. Zooms into boxes in (A) depict (B) end-to-end Cx43 labeling between two myocyte pairs, (C) complex end-to-end junction between myocytes with Cx43 labeling, (D) α-SMA positive cell near a myocyte membrane, and (E) side-to-side myocyte coupling with Cx43 labeling. (F) Reconstructed myocytes from the image stack. (G) Reconstructed non-myocyte contributors to the myocardium with ES in blue, vessels in red, fibroblasts in cyan, and myofibroblasts in yellow.

• 

  Fig. 3.. Representative myocyte reconstructions. (A) Reconstructions of myocytes from control image stacks of 100 μm in length and width with myocytes individually colored. The reconstructions illustrate the dense packing and alignment of myocytes. (B) Reconstruction myocytes from MI image stacks depict a larger variety including pronounced size differences and disarray among myocytes.

• 

  Fig. 4.. Electrode placement and field calculation in conductivity model derived from control image stack. (A) Reconstructed myocytes with electrodes visualized as red planes placed. Only half of the myocytes are shown. We applied electric fields along normal, longitudinal, and transverse directions. The orientation of arrows shows the correspondence to x, y, and z directions, respectively. (B) Potential distribution due to applied field with red, white, and blue corresponding to 1, 0, and −1V. (C) Current density distribution shows sparse regions of non-zero values in x and z directions in contrast to current density along y.

• 

  Fig. 5.. Electrode placement and field calculation in conductivity model derived from MI image stack. (A) Reconstructed myocytes with electrodes for fields applied along x, y, and z directions. In part, a loss of direct connection occurred between myocytes as well as between electrodes and myocytes due to large ES. (B) Potential distribution through the reconstruction along the three axes. (C) Current density visualized on myocytes.

• 

  Fig. 6.. Streamline visualization of simulated current densities in myocardium. Current density streamlines in a control stack for the electric field applied along (A) normal, (B) longitudinal, and (C) transverse direction, i.e. x, y, and z direction, respectively. The streamlines were seeded at the center of the stack on a plane parallel to the electrodes and trace the paths of current flow. A high density of streamlines identifies locations with a high current density.

• 

  Fig. 7.. Statistical analyses of tissue features and conductivities. We used violin, box, and strip plots. (A) Vmyo, (B) Vfibrosis, and (C) Vgj differed in control (blue, n = 22) and MI (red, n = 28) samples. (D) σmyo,l was decreased in MI samples. (E) The difference of σmyo,t for control and MI samples was marginal. (F) σmyo,n was decreased in MI samples. **: p < 0.01, *: p < 0.05.

• 

  Fig. 8.. Regression analyses of tissue features and conductivities. Linear regression lines in black, blue, and red, wherever present, applied to all samples, control samples (n = 22) alone, and MI samples (n = 28) alone, respectively. (A) The relationship of σmyo,l with Vmyo was strongly positive. (B) σmyo,l−Vfibrosis was strongly negative. (C) σmyo,l−Vgj was strongly positive. σmyo,t did not exhibit relationships with (D) Vmyo and (E) Vfibrosis. (F) σmyo,t had a weak positive relationship with Vgj for MI samples alone. (G) σmyo,n increased with increasing Vmyo and (H) decreased with increasing Vfibrosis. (I) σmyo,n increased as Vgj increased.