Radiation dose reduction in fluoroscopic procedures: left varicocele embolization as a model

Abstract

Objectives

To investigate the effect of a radiation reduction program on total dose, fluoroscopy dose per second corrected for body habitus and degree of collimation in left varicocele embolizations (LVE).

Methods

A radiation reduction program for LVE was implemented, consisting of a technique minimizing fluoroscopy time, using low-dose presets, virtual collimation, and virtual patient positioning. Height, weight, fluoroscopy time, kerma area product (KAP) and reference air kerma (Ka,r) were recorded for 100 consecutive cases satisfying the inclusion criteria. For each patient, a device specific dose correction factor, determined using a phantom, was used to standardize the KAP to that of the cylindrical diameter of the standard man and a collimation index was derived from the KAP and Ka,r.

Results

Median fluoroscopy time was 3 minutes (mean 4.5, range 1-23.8). Median KAP was 0.54 Gy/cm² (mean 0.82, range 0.12-6.52). There was a significant decrease in KAP/second corrected for cylindrical diameter (p < 0.001) and the collimation index (p < 0.001) over time.

Conclusions

This study shows that a dedicated dose reduction program can achieve very low total radiation dose rates for LVE. The significant decrease in collimation index and standardized KAP per second during this study suggest a learning curve for collimation.

Key Points

• Left varicocele embolization can be performed with very low radiation doses.

• The most effective real time radiation reduction measure is collimation.

• There is a learning curve to achieving optimal collimation.

Appendices

Appendix 1

A device-specific correction factor K(cm^-1) relating radiation dose to thickness of area irradiated was derived using a variation of the method described by Chapple [16]. A water bath 30 centimeters in height was placed on the tabletop allowing determination of KAP for different depths of water starting from zero and increasing by three centimeter increments. The exact level of the water was determined using a ruler taped to the inside of the water bath. Each depth of water was irradiated for five seconds on the low-dose fluoroscopy setting at 7.5 pulses per second to allow autocalilbration of tube output. The KAP was the then recorded for five minutes of continuous fluoroscopy. The table detector distance (set arbitrarily at 32 cm) and collimation were unchanged throughout the procedure. The water phantom thickness and the KAP were plotted on a scatter diagram to yield an exponential gradient K, which was used to correct the radiation doses of the varicocele embolization procedures to that of the reference man of 73 kg weight and 176 cm height.

Appendix 2

The accuracy of the KAP meter for different irradiated areas was investigated using a 23 cm thick water phantom. Each area was irradiated for five seconds on the normal dose fluoroscopy setting at 7.5 pulses per second to allow autocalibration of tube output. The KAP and Ka,r were then recorded for two minutes of continuous fluoroscopy and the results plotted against the area irradiated which was calculated from measurements made using the inbuilt software (Fig 5).

Fig. 5

Plot of KAP and Ka,r against area irradiated demonstrating linear increase in KAP over the entire range of areas irradiated in the study. The slight decrease in Ka,r with increasing area is caused by the automatic brightness control compensating for the increased scatter radiation with larger field size

The results demonstrated a linear increase in KAP over areas from 6.5-180 cm² which covered the entire range used for left varicocele embolization in this study.