Low Dose Brain CT, Comparative Study with Brain Post Processing Algorithm

Computed tomography (CT) scanners and CT exams increase continuously. The researchers aim to minimize the ionizing radiation dose by introducing new CT protocols, providing diagnostic CT images with a lower radiation dose to patients. However, such studies encounter difficulties, when the radiation dose is lowered, the quality of images becomes less and sometimes not diagnostic. In this study, the researcher aims to provide a low dose brain CT protocol, in order to then determine if the images match the quality criteria of Brain CT; and determine the diagnostic appearance of the images. Then, the researcher will compare the results obtained from the Brain CT, as well as the brain post-processing algorithm to determine which one provides a better diagnostic image, and a better match for the quality criteria of Brain CT, by the Numerical criterion (1: weak, 2: moderate, 3:perfect) which is used by expert medical imaging technologists, On a sample of 35 patients; the first brain CT was conducted by 22 milli-gray (mGy) volume computed tomography dose index (CTDIvol); the resulting image was noisy, with a poor match for quality criteria, then CTDIvol was raised to 25 mGy, then to 30 mGy, and finally to 33.8 mGy. At this point, the image was acceptable to complete the study. Four radiologists have been engaged to determine if the image provides diagnostic appearance, then six expert medical imaging technologists were involved to determine the quality criteria. These steps were followed for the Brain CT before and after applying the post-processing algorithm. Then the results were compared with the reference study of brain CT. The results for low dose brain CT were diagnostic and matching the quality criteria for brain CT. After applying the brain post-processing algorithm the image's diagnostic appearance was disturbed, the suggested protocol by the study provided a 47% dose reduction, from the standard protocol which uses 63 mGy. The problem of signal reduction is solved by using iDose4, which improves the signal to noise ratio (SNR).