| ||||||||||||||||||||
Data Supplement for Circulation: Volume 106, Issue 22 -- Page 2771
This item has the following additional materials available:
Data Files
Figure I (JPEG file). Schematic of the OCT system (adapted from Rollins et al.19). The high-speed OCT system used in this study has three noteworthy features. First, a broadband near-infrared semiconductor light source is used to illuminate the sample. Second, it employs a fiber optic Michelson interferometer for time-gating of backreflected sample light. Third, femtosecond pulse-shaping technology is used in the rapid scanning delay line to generate an A-scan acquisition rate of 4kHz. B-scans are acquired by scanning the sample beam laterally with a galvanometer-mounted mirror. Volume scans are acquired by mechanically translating the probe along the z-axis on a precision micrometer translation stage.
Figure II (JPEG file). In each stage (14/15, 16, 18, 20 and 22) two embryos were collected to visualize cardiac development with OCT. The two embryos of each stage were scanned separately in two different planes. One embryo was scanned from cranial to caudal direction to generate B-scan OCT images corresponding anatomically to the transverse plane, the second embryo was scanned from dorsal to ventral while the B-scan OCT images here were corresponding anatomically to the frontal plane. During the scanning procedure the embryos were lying on their left side in a petri dish filled with saline. The embryos were separated directly after scanning and labeled in petri dishes with “frontal plane” or “transverse plane” to ensure correct embedding of the embryos in the corresponding plane to the OCT images. The embryos were immersion-fixed at 4°C overnight in 4% paraformaldehyde. The following day they were dehydrated, cleared, and embedded in paraffin (two embryos of each stage were embedded in two different planes in paraffin to match exactly the planes of B-scan OCT images). The embryos were embedded horizontally and vertically in such a way that after cutting the paraffin blocks matching histological sections could be found easily on the generated slides. The paraffin blocks were sectioned at 7 μm. The slides were postfixed in Bouin’s solution and stained with hematoxylin-eosin for comparison with acquired OCT images. Correspondence was determined by the best match between OCT images and light microscopy observations of the histology.
Movie I (AVI file). Flip-book movie of B-scans from the HH-stage 15 chick embryo which is shown in Figure 1 G and Figure 2 A in the printed manuscript. The movie consists of 53 B-scans, with each B-scan displayed for one-eighth of a second. Depicted is the early heart tube in the frontal plane. The images scan through the heart from posterior (starting with the inflow limb or the cardiac loop) to anterior showing the outflow limb of the cardiac loop and presumptive right ventricle. Cardiac structures such as the compact outflow myocardium, the cardiac jelly and endocardium can be identified easily. The movie was created in Matlab (The MathWorks).
Movie II (AVI file). Three-dimensional rotation of the early heart tube shown in Figure 2 A with 3D volume reconstruction. The outflow limb and presumptive right ventricle of the heart tube are depicted. The same cardiac structures described in Movie 1 are well visualized. The 3D images and slices were created in Slicer Dicer (PIXOTEC, LLC).
Movie IIIA and Movie IIIC (AVI files). Flip book movie of B-scans of another HH-stage 15 chick embryo taken in the frontal plane (A). Using the volume data in A, serial sections were taken in the sagittal plane (B) and in the transverse plane (C) and displayed in flip book format. All structures of the early heart tube are depicted in detail, i.e. inflow and outflow limb of the cardiac loop, myocardium, the cardiac jelly, endocardium, inner and outer curvature of the cardiac loop, and the ventricle. Figure 1A in the manuscript shows the geometric and anatomic orientation of the volume data. In our native B-scans, we sampled in 3 micron intervals on the y-axis, yielding 7 samples per full-width half-maximum duration of the axial point spread function. Along the x-axis, we sampled in 2 micron intervals, yielding 14 samples per full-width half-maximum duration of the lateral point spread function. When we performed volume scans by moving our probe in 10 micron intervals along the z-axis, we acquired 3 samples per full-width half-maximum duration of the lateral point spread function. That we sampled 2-3 times less densely along the z-axis explains why there is degradation of image quality and pixilation in the movies of sequential images in the sagittal and transverse planes, but not in the movie of sequential frontal plane images. The frontal plane images are native B-scans, which are densely sampled along both axes of the image. The sagittal and transverse movies are composed of images that are densely sampled in one axis but sparsely sampled in the other. This is the origin of the pixilation. The movies were created in Matlab (The MathWorks).
Movie IV (AVI file). Dynamic motions of the heart of a HH-stage 15/16 chick embryo captured with OCT at a speed rate of 8 frames/second. Note the filling of the outflow limb at end-systole and the relaxation of the cardiac tube in end-diastole. Still frames of this movie are presented at the bottom of Fig 1 (H through K). The images show the heart in end-systole (H), early (I) and end-diastole (J), and at early systole (K).
Prepared by: the Data Supplement Manager
| ||||||||||||||||||||