An overview of the proposed process flow is shown in Fig. 1. Contrast-enhanced X-ray pulmonary angiography images were acquired. The acquired images are integrated into the image processing flow. The image of the base mask, obtained before the injection of the contrast agent, is subtracted from the subsequent consecutive images. ROI was determined in each right and left lung region, as shown in Fig. 2. The TICs of two ROIs are obtained. The time window for the analysis of the obtained TIC is optimized by the new algorithm. The TIC parameters are calculated in each region for the optimized time window. Finally, the right-left ratio is calculated.
To obtain objective, quantitative and reproducible automated methods, several key functionalities are necessary. These include precision, reproducibility, broad application to different types of diseases, and rapid computation that is clinically acceptable for the procedure. To meet these requirements, in this work, we developed an algorithm to determine the ROI size, ROI location, automated time window optimization, stable parameter selection in the TIC and minimizing computation time.
Determination of return on investment
For a quantitative assessment of the right-left ratio of pulmonary blood flow distribution, rectangular ROIs are placed in the right and left regions, as shown in Fig. 2. When an image is acquired, the field of view and distance from the source imager are usually adjusted so that the entire lung is maximally included in the image to minimize radiation dose to the patient. Therefore, the ROI can be expanded to cover left and right lungs; these ROIs are close to the vicinity of the edges of the image, as shown in Fig. 2. The gap between the right ROI and the left ROI in the middle of the image is increased as much as possible so that the main pulmonary trunk and the tip of the catheter are excluded, but the whole lung region is included. A larger discrepancy between left and right ROIs is also beneficial when evaluating many complex pediatric treatments, such as Blalock-Taussig shunts. In this study, the size and location of the ROI are fixed for all analyzed cases. In the 1024 by 1024 images, the width of the ROI is 350 and the height of the ROI is 820. The coordinates are shown in Figure 2; Right ROI (X1, Y1, X2, Y2) = (9, 103, 358, 922) and Left ROI (X3, Y1, X4, Y2) = (665, 103, 1014, 922). ROI selection is not affected by dynamic acquisition because diaphragm movement is not critical for a short period of time within 200 ms. The X-ray image acquisition angle of the cranial (CRA) and caudal (CAU) directions can be applied as well as the anteroposterior (AP) directions. However, the left anterior oblique (LAO) and/or right anterior oblique (RAO) directions cannot be used.
Contrast-enhanced XA images are incorporated into image processing. The image of the base mask, obtained before the injection of the contrast agent, is subtracted from the subsequent consecutive images. The TIC of two ROIs was obtained by averaging all pixel values in each ROI. By using this averaging approach to calculate the TIC, the computation time is greatly reduced. The original calculation requires image-based processing of all pixels in the image, which is image width by image height (for example, 1024 by 1024 pixels). However, the current ROI-based processing approach only requires two calculations (left and right ROI).
Time window optimization
The right-left ratio of pulmonary blood flow distribution is calculated only in the specific time window to measure equivalent blood flow with LS which has different tracer kinetic patterns3,9,10. In X-ray angiography, the time window should be set to the torrent period during the second cardiac cycle after contrast injection. The torrent period is a short period during which the contrast agent is torrentially flushed from the pulmonary arteries to the capillary bed. The second cardiac cycle is used to eliminate variance in contrast agent concentration because the contrast agent is not well mixed and distributed unilaterally in the pulmonary trunk during the first cardiac cycle immediately after injection of contrast. This unilateral distribution leads to lateral flow in the first cardiac cycle. By using the second cardiac cycle, this variance is reduced and a stable measurement is obtained.
In this paper, the average TIC combining both right and left regions is used, and the maximum slope time of the combined TIC is detected. If lateral flow occurred due to unilateral distribution in the pulmonary trunk, the combined TIC would have a small slope because the total amount of contrast flow was low; therefore, the timing of a lateral flow would not be detected. If the contrast is well mixed, the contrast agent flows to the right and left regions simultaneously, the total amount of contrast flux is large, and the combined temporal signal density curve should have a steep slope.
The duration of the time window is fixed at less than 200 ms; six frames in the case of data acquisition at 30 frames/s. Indeed, the period between the moment when the contrast agent arrives at the level of the first branch of the pulmonary artery and the moment when the contrast agent fills the entire pulmonary field is approximately 200 ms.
We observed that the start time of contrast flow from the pulmonary trunk was slightly different between the right and left sides. The difference is up to 100 ms. This difference does not affect LS measurements that count temporally accumulated tracer11.12. On the other hand, this impacts the proposed method because the proposed method does not measure the accumulation but measures the net increase in the TIC in a short window of time. In this paper, a new automated algorithm is proposed to obtain stable results even in cases where the start time of the contrast flow is slightly different. In this algorithm, the time window is optimized for right and left lung regions independently. First, a representative time slot of six frames is detected using the combined TIC above. Second, it is extended by eight frames: four frames before and after the six representative frames. A total of 14 frame lengths are determined as the candidate time slot. Third, from this time window of 14 candidate images, six images that show the maximum slope of the TIC are selected from each right and left region independently. These steps are illustrated in Fig. 3. In summary, optimized time windows are selected for each right and left region independently within the same cardiac cycle.
Calculation of parameters
The right-left ratio of pulmonary blood flow distribution is calculated by the net increase in signal intensity and is an equivalent model with scintigraphy3. In this article, a stable selection of parameters is investigated. If only two points are used to measure the net increase in signal intensity, it is easily affected by several noise factors, such as body motion, heart motion, and image acquisition noise. Therefore, in this article, the six points of the time window are used to calculate the slope using a linear fit. This approach is equivalent to the scintigraphy method, and it makes the algorithm stable and robust.