MR microscopy can produce high-resolution pulmonary images in live rodents by synchronizing the image acquisition across multiple breaths [1]. The precision to control motion will most likely define the attainable resolution limits. This work evaluates how reliably the respiratory structures return to the same position from breath to-breath each time data are acquired.

Radio-opaque tungsten beads (0.28-mm diameter) were surgically glued on the abdominal surface of the diaphragm (Figure 1) of 10 anesthetized rats that were mechanically ventilated using a custom-built ventilator [2].

The rats were mechanically ventilated at 60 breaths/min (tidal volume: 1.8–3.0 ml). A solid-state pressure transducer on the breathing valve measured airway pressure.

Pediatric electrodes were taped on the footpads for ECG. These signals were also used to control the biological pulse sequences described below. The range of motion of the beads (relative to a reference vertebral bead) was evaluated using digital micro-radiography, by measuring the standard deviation of the distance between these beads and the reference bead on a series of 150–200 consecutive images.

We used two specific biological pulse sequences (Figures 2a and 2b)—ventilation-synchronous acquisition, and both ventilation-synchronous and cardiac-gated acquisition. The control and acquisition system was made up of three computers, a high-capacity angiography tube with 0.1-mm focal spot driven by a Phillips CXP 80 generator, and a 50-micron digital camera (Microphotonics XQUIS, Photonics Science, East Sussex, UK) (Figure 3). The programs that controlled the x-ray tube and camera according to the physiological parameters were developed using LabVIEW (National Instruments, Austin, TX).

Figure 1: Ventro-dorsal projection of the thorax in a rat. Note the three implanted beads on the diaphragm and a reference bead on a thoracic vertebra. The insert shows a close-up view of the left bead.

Figure 2a: Ventilation synchronous acquisition. For breath-hold images, the x-ray pulse occurred 150 (early breath-hold) to 250 ms (late breath-hold) after the beginning of inspiration (shown by dark-gray). For end-expiratory images, the x-ray pulse occurred 850 ms after the beginning of the inspiration. The camera was enabled at the same time as the x-ray pulse.

Figure 2b: Ventilation-synchronous and cardiac-gated images. A 110-ms acquisition window was defined (here represented for the breath-hold images) and any QRS occurring within this window triggered an x-ray pulse and enabled the data acquisition by the camera.

Figure 3: Schematic of the acquisition system controlled by three computers. One computer controls the ventilator and acquisition of physiological parameters. Another computer controls the x-ray rotor, enables the camera, and generates the x-ray pulse according to the defined biological pulse sequence. The third computer records the digital data coming from the camera.

Table 1: Mean differences (in microns) in distances between inspiratory breath-hold and end-expiration images, which represents the diaphragm excursion over the breathing cycle, providing an insight into the actual spatial resolution that would be achieved without control of breathing motion.

Table 2: Global results. For each biological pulse sequence, we present the standard deviation in microns averaged across all animals. Values listed are averages of the values obtained on the two perpendicular projections.

1. GA Johnson, GP Cofer, LW Hedlund et al., Magn Reson Med, 45: 365-370, 2001.

2. LW Hedlund, GA Johnson, ILAR J 43: 159-174, 2002.