Optimization in 64-mdct of the chest using tube current modulation based on noise index : phantom study

Abstract

Tube current modulation is the great potential method for the radiation dose reduction. Noise Index is an indicative parameter of the image quality controlled by the AEC system for the noise level in the image associated with radiation dose. The purpose of this study is to determine the radiation dose when varying Noise Index and the optimal Noise Index for the acceptable image quality of the chest phantom. The standard and large sizes Lung Man Chest phantom with +100 HU of 12, 10, 8, 5, and 3 mm in diameter of spheres simulated nodules were scanned with various reconstruction filters, 0.625-5.0 mm of slice thickness, Noise Index 10-20 and 75-380 of mA. The CTDIvol (mGy) and DLP (mGy.cm) were recorded from the CT console. The quantitative image quality was assessed by the contrast to noise ratio (CNR) values. The nodule detectability and spatial resolution were independently evaluated by two radiological technologists for qualitative image quality. With the variation on Noise Index of 10-20 and the slice thickness of 0.625-5.0 mm on the standard size phantom, the radiation dose was decreasing, CTDIvol from 16.52 to 3.38 mGy. The CTDIvol and DLP slightly decreased at the thin slice thickness and rapidly decreased at thick slice thickness on varying Noise Index. There were no variation of CTDIvol and DLP at the thin slice thickness for the large size phantom, the range of CTDIvol was 16.52-7.09 mGy. The STD filter offered the highest percent CNR when compared to the CHEST filter at 50-60%, the LUNG and BONE+ offered the lowest percent CNR respectively. The LUNG filter produced the best spatial resolution image. The scoring on image quality by two observers for standard, and large size phantoms were similar with good agreement. Noise Index and slice thickness are the major parameters affecting the radiation dose. Increasing of Noise Index 10-20 results in decreasing radiation dose to 18.2, 32.1, 64.9, and 65.9% for 0.625-5.0 mm slice thickness respectively. The slice thickness had a major impact on radiation dose for the large size phantom with the reduction from 12.82% to 57.07 % for 2.5-5.0 mm of slice thickness respectively. The STD filters were designed for good spatial resolution with reasonably low image noise. The Noise Index has little affected on CNR. When varying Noise Index from 10-20, the reduction in CTDIvol was 9.91-3.38 mGy for standard size phantom, and 16.52-7.09 mGy for large size phantom. The factors affecting on radiation dose and image quality were Noise Index, slice thickness and reconstruction filters. The selection on Noise Index depends on clinical applications. Using the Noise Index of 20 at 75-380 mA with LUNG and Bone+ filters resulted in acceptable subjective image quality whereas Noise Index 15-17.5 at 75-380 mA with STD filters resulted in acceptable objective image quality for routine chest CT.