Micro-CT with respiratory and cardiac gating

Abstract

Cardiopulmonary imaging in rodents using micro-computed tomography (CT) is a challenging task due to both cardiac and pulmonary motion and the limited fluence rate available from micro-focus x-ray tubes of most commercial systems. Successful imaging in the mouse requires recognition of both the spatial and temporal scales and their impact on the required fluence rate. Smaller voxels require an increase in the total number of photons (integrated fluence) used in the reconstructed image for constant signal-to-noise ratio. The faster heart rates require shorter exposures to minimize cardiac motion blur imposing even higher demands on the fluence rate. We describe a system with fixed tube/detector and with a rotating specimen. A large focal spot x-ray tube capable of producing high fluence rates with short exposure times was used. The geometry is optimized to match focal spot blur with detector pitch and the resolution limits imposed by the reproducibility of gating. Thus, it is possible to achieve isotropic spatial resolution of 100 microm with a fluence rate at the detector 250 times that of a conventional cone beam micro-CT system with rotating detector and microfocal x-ray tube. Motion is minimized for any single projection with 10 ms exposures that are synchronized to both cardiac and breathing motion. System performance was validated in vivo by studies of the cardiopulmonary structures in C57BL/6 mice, demonstrating the value of motion integration with a bright x-ray source.

Fig. 1

The x-ray tube (a) and detector (b) are stationary. The mouse is supported in an acrylic tube (c) placed on a support which is rotated by a computer-controlled stepping motor (d). The tube and detector are supported on a gantry (e) constructed from extruded aluminum to limit the impact of building vibration. The relative position of the elements of the scanner is easily adjusted. The valve for scan synchronous ventilation (f) and ECG leads are supported from the top of the gantry.

Fig. 2

Relative fluence rate is plotted as a function of focal spot dimension for limiting resolution at 25, 50, and 100 μm. The fluence rate is normalized to the maximum that is attained for resolution at 100 μm with a focal spot of 0.3 mm. The normalized exposures for 80 kVp, 10 ms are plotted for the 0.3 and 1.0 mm focal spots at the sdd required for resolution of 100, 50, and 25 μm and the maximum current available for the two focal spots (100 and 620 mA).

Fig. 3

A schematic of the system is shown (a) and the signal wave forms during sampling in (b). The system is controlled by three computers each running LABVIEW. PC1 controls the ventilator and monitors the physiologic signals from the animal. PC2 acts as the sequencer for the system. It receives triggers from PC1 that control the x-ray generator and the stepping motor for the gantry. PC3 receives the trigger from PC2 that controls the camera acquisition, integration, and readout. (b) shows a capture of the monitoring application during sampling. Trace 1 shows the pressure at the ventilator. In this example a window is enabled at end expiration (trace 2) which is then logically combined with the ECG (trace 3) to allow exposures 1 and 2 (EX1, EX2). Trace 4 shows the digital signal to advance the table to the next projection.

Fig. 4

The limiting resolution imposed by the focal spot and geometry was verified in projection images for the 0.3 mm (a) and 1 mm (b) focal spots using a digital camera with a 0.05 mm pitch. (c) The MTF of the entire micro-CT system measured in reconstructed images with a disk phantom as described in ASTM.

Fig. 5

Coronal slices are shown at the same level from four studies done on the same mouse. The lower images show the left lung from each image magnified by 3×. (a) and (e) are from the ungated study. (b) and (f) were obtained with ventilatory synchronization; (c) and (g) were obtained with ventilatory synchronization and cardiac gating; (d) and (h) were obtained on the same animal after sacrifice with an anesthetic overdose.

Fig. 6

Axial and coronal 400 μm slabs extracted from the study with ventilatory and cardiac gating providing detailed anatomy of the thorax: vena cava (VC), descending aorta (DA), right (RB) and left (LB) pulmonary bronchi and associated bronchial vessels, and the gall bladder.