All procedures were conducted in accordance with institutional guidelines and AVMA Guidelines for the Euthanasia of Animals. In accordance with Section 3.4.4 of the AVMA Guidelines, all quail embryos were decapitated at stage HH38 on day 10 of incubation.
2.1. Procurement and Incubation of Quail Eggs
Fertilized quail eggs were purchased from Ozark Egg Company, Stover, MO. Early incubation efforts targeted approximately 40 eggs per cohort to support imaging throughput with an expected incubation success rate of 70%. Subsequently, a twice-weekly incubation schedule was adopted, with 20 eggs set for incubation on Mondays and Tuesdays to allow for completion of incubation on the following Thursday and Friday, corresponding to 10 days of embryonic development. This was performed over a month-long period, correlating to ~160 eggs incubated. This large imaging throughput was initially necessary to adapt to the minuscule manipulations required for the proper visualization of the quail embryos. Due to this, many of the early images did not correspond to visualized structures usable for data acquisition. However, following this initial period, the echocardiographic experience of the primary investigator, guided by the pediatric cardiologist, was sufficient to obtain imaging that was consistent with scans capable of data analysis.
An incubation period of 10 days was selected to allow the longest developmental period before visualization. Additionally, the longer development was chosen as, at this stage, we would expect the manifestation of congenital heart defect abnormalities. This stage was also chosen due to developmental constraints associated with quail embryogenesis. Japanese quail begin feather development at approximately embryonic day 9, with substantial feather growth continuing after day 10, impairing visualization [
6]. During experimentation, we observed that feather formation significantly impairs echocardiographic visualization of the developing heart by introducing artifacts and acoustic shadowing, thereby reducing image quality and data reliability.
To control for confounding variables, the fertilized eggs were homogeneously distributed on an incubation tray in a Genesis model 1588 (Hova-Bator Incubator, Savannah, GA, USA), maintained at 38 °C. Each egg was marked on the superior surface to ensure consistent orientation during post-incubation access. Throughout the incubation period, distilled deionized water was added to the chamber’s reservoir to maintain appropriate humidity levels for proper growth conditions.
2.2. Quail Embryo Preparation for Echocardiography
Preparation of the quail embryos for echocardiography required careful temperature control throughout the procedure to preserve embryonic viability. Although embryos were humanely terminated in accordance with institutional and AVMA regulatory guidelines following imaging, it was essential to maintain normothermic incubation-like temperatures during preparation and scanning. To achieve this, several procedures were performed as follows. Eggs not actively being prepared or scanned were kept in the incubator until use. Upon removal for shell preparation, eggs were immediately placed on a heating pad. During echocardiographic imaging, temperature was maintained by circulating warm air through a custom 3D-printed imaging incubator, which supported the egg in an aluminum foil cradle with heated airflow beneath it by the Airtherm SMT Model 98736-3 heater (World Precision Instruments, Sarasota, FL, USA) (
Figure 1). A temperature sensor was secured underneath the aluminum foil cradle to monitor heating throughout the course of imaging.
Egg shell removal was performed on a heating pad under direct illumination to enhance visualization (
Figure 2). A molded aluminum foil cradle was used to stabilize the egg and prevent its rolling during manipulation. The peeling process was initiated by gently tapping the inferomedial, wider aspect of the egg with the blunt end of a dull instrument, just below the reference line on the superior aspect of the egg marked during incubation. This initial opening facilitated a controlled shell removal process using forceps. Care was taken to avoid damaging the thick outer membrane beneath the shell, which is adherent to the underlying chorioallantoic membrane containing the embryo’s vasculature, to prevent rupture of these vessels and resultant exsanguination. A visualization window approximately 3 cm × 2 cm, making up most of the superior aspect of the egg, was created. The outer membrane was then carefully separated from the chorioallantoic membrane, as disruption of the latter results in rapid exsanguination and significantly diminishes the duration of viable imaging. Small tears in the outer membrane were used as initiation points to facilitate this controlled separation process.
Following shell and membrane removal, the egg was prepared and positioned for echocardiographic imaging. A double-layered aluminum foil insert was secured within the 3D-printed incubator, and the prepared egg was positioned upon this heated opening. A square sheet of plastic wrap was placed tautly over the visualization window to form a stable imaging interface for the echo probe. Gauze was placed between the foil borders and the positioned egg to further stabilize the egg and prevent its shifting during scanning. Throughout imaging, 38 °C air was continuously circulated beneath the foil cradle to maintain appropriate temperature conditions.
2.3. Echocardiographic Imaging of Quail Embryos
Echocardiography can provide comprehensive information on cardiac anatomy, physiology, and mechanical properties. In humans, echocardiography consists of transmitting high-frequency sound waves, which are reflected off different tissue moieties (e.g., myocardium, blood, valves, etc.) back to an ultrasound transducer that receives the reflected signal off all tissues. Software dynamically processes the incoming signals from the different tissues and generates a real-time image based on the known acoustic impedances of each tissue. B-mode produces 2-D views of the heart (short- or long-axis), allowing the assessment of cardiac chamber dimensions, physiology, and visualization of cardiac anatomic structure such as papillary muscle and valves.
Using the Fujifilm Visual Sonics Vevo 3100 ultrasound system (FUJIFILM VisualSonics Inc., Bothell, WA, USA), real-time echocardiographic imaging of the quail heart on embryonic day 10 (HH38) was performed. The MX550D transducer was utilized for image acquisition, which is the gold standard for small animal model micro-sonography. The transducer has an axial resolution of 40 µm, lateral resolution of 80 µm, center frequency of 40 MHz (broadband 25–55 MHz) and a depth of up to 15 mm. We used 3 main modes of echocardiography: brightness mode (B-mode) which produces 2-dimensional (2D) views of the heart (long-axis or short-axis) and Doppler echocardiography which included both color Doppler and pulse wave Doppler (PWD). For 2D images the frequency was 40–60 MHz, with frame rates ranging from 150 to 200 frames per second (fps) with a gain of 25–30 dB and a dynamic range of ~50 dB. For color Doppler the frequency was 30 MHz, 20–30 fps frame rate, and a pulse repetition frequency (PRF) of 10–15 kHz. For the PWD, the frequency was 30 MHz, dynamic range was 30 dB and PRF was 25 kHz. During PWD assessment we only selected the images where the Doppler was placed in line with the blood flow, aiming for an angle of insonation that is less than 40° and therefore angle correction was not required for any of the images chosen.
To begin, a small amount of ultrasound gel, the size of a quarter, was applied to the plastic wrap covering the visualization window and evenly distributed using the echo probe to ensure proper acoustic coupling. The imaging depth was ~14 mm. Quail embryo heart rates are extremely fast, around 600–750 bpm, when compared to the heart rate of human embryos, i.e., 150–170 bpm. Multiple scans were performed on each embryo to obtain consistent morphological and functional data. Gain and depth settings were conservatively modified throughout the course of imaging and across embryos. Like humans, there were variances across the embryos that required subtle manipulation of gain/depth to maximize morphological structure visualization.
Scans were acquired in both long-axis and short-axis views. The long-axis view provided a detailed visualization of cardiac morphology, allowing measurement of myocardial thickness, right ventricular outflow tract and pulmonary outflow velocities, and pulmonary arterial diameter. The short-axis view allowed for additional measurement of the myocardial thickness, as well as primary measurement of right ventricular outflow tract velocity. Attempts to utilize the M-Mode function of the Vevo 3100 system were unsuccessful. M-Mode is universally utilized to quantify cardiac function. However, the size of the quail embryonic heart and the rapid heart rate limited the reproducibility of this parameter. Future work will seek to incorporate this imaging modality and standardize its implementation.
The process of scanning was initiated by lowering the echo probe gently onto the embryo’s surface to calibrate the image on the Vevo 3100 system. All imaging was performed with the echo probe perpendicular to the surface of the embryo. Further manipulation was needed to obtain the long-axis view, which served as an anchor image for the rest of the scanning process. The scan selection was based on visualization of specific landmarks (pulmonary artery (PA) being the anterior outflow and aorta being the posterior outflow) rather than relying on signal-to-noise ratio. At the long-axis view, a B-Mode scan was performed to visualize all the relevant anatomical structures, such as the cardiac chambers and outflow tracts. The probe was then adjusted to prioritize visualization of the ventricular wall to measure the ventricular outflow dimensions, followed by Doppler. However, the visualization of the ventricular outflow in the long-axis was sometimes inconsistent. The probe was further manipulated cranially along this same axis to visualize the pulmonary artery, allowing B-Mode measurement of diameter and Doppler assessment of outflow velocity.
Rotation of the echo probe 90° from the long-axis view provided the short-axis view. From this view, fine adjustments were required to frame the heart properly. Following visualization of this view, both B-Mode and Doppler scans were obtained to provide analysis for ventricular wall thickness and ventricular outflow tract velocity, respectively. The Doppler probe on the Vevo 3100 screen was localized to the right ventricle for outflow tract velocity measurement during scanning. All processes and steps were attempted on each viable embryo. However, not every embryo showed sufficient visualization of each desired feature. This is due to expected variations in where the embryo was sitting, or if the shell was blocking visualization.
Due to the small size of the quail embryos, with structures often less than 0.5 mm thick, precise echo probe manipulation was critical. To improve operator skill and efficiency, 2–4 chicken embryos were incubated and scanned as practice models alongside the quail embryos but were not utilized for data collection. Familiarity with human echocardiographic procedures, as well as consultation with a pediatric cardiologist, facilitated efficient imaging on the Vevo 3100 system due to the inherent ability of not having to reposition the probe for every scan.
2.4. Quantification of Embryonic Cardiac Morphology and Function
In totality, around 100 embryo scans were analyzed for morphological assessment. For each morphological feature, the 10 highest quality B-Mode scans were selected for analysis. An objective system was used to determine “highest quality” by most prominent visualization of cardiac morphological structures of interest. This objective system was informed by collaboration with a pediatric cardiologist with the primary criteria being visualization of cardiac structures based on specific anatomical landmarks such as the developing lungs, ventricular outflow tracts, great vessels, etc. Each embryo was scanned to visualize as many morphological features as possible. However, due to variability across embryos and time restrictions, most embryos allowed for analysis of one morphological feature.
Due to this, 10 measurements of each morphological structure were nested in each embryo and performed across 10 embryos per feature, totaling 100 data points for all ventricular wall thicknesses and pulmonary artery diameter measurements. These measurements were taken from one hyperechoic edge to the opposing hyperechoic edge, in both systole and diastole. Because right ventricular wall thickness was visualized in both long- and short-axes, this process was performed in the long-axis and repeated upon manipulation to the short-axis. Pulmonary artery thickness was measured using the same procedure in the long-axis only. For each outflow velocity Doppler analysis, 3–7 measurements were nested in each embryo and performed across 10 embryos. Pulmonary artery outflow velocity was measured distal to the right ventricle in the long-axis view, whereas right ventricular outflow tract velocity was measured in the center of the right ventricular chamber in the short-axis view. All measurements were performed by a single observer and intra-observer error is reported in the results section.
For each site, the mean (SD), median, and IQR range (minimum–maximum) were calculated to summarize the distribution of measurements. In addition, we fit a linear mixed-effects model for each assessment with random intercepts for subject to assess site-specific measurement repeatability using intraclass correlation coefficients (ICC) and measurement variability with the coefficient of variation (CV).