Extraction of tube current values from DICOM CT images for patient dose estimation

Abstract

When CT examinations are conducted with automatic mA modulation (z-axis modulation), the tube current (mA) may vary from one CT slice to next, i.e., mA is no longer a fixed constant. Hence, patient dose calculation is no longer a straightforward process. The purpose of this article is to show how the mA information may be extracted from DICOM CT images without having to manually read off "mA values" from the displayed images one at a time. A statistical programming language called "R," which is capable of reading DICOM files, is employed to extract the mA values from a series of DICOM CT images. This task is carried out with a "script" designed to read the mA values from the DICOM CT images and generate a file in "delimited ASCII" format. This file can be imported into a spreadsheet program such as Excel (a Microsoft spreadsheet program) for further processing, calculation, and chart production. A CT examination of the chest was selected to carry out this operation for demonstration purposes. The "script" generated a delimited ASCII "file" which is a two column data sheet with the slice location and its corresponding mA values. After the file is imported into Excel for calculation, and with other pertinent scan parameters, the average mA can now be plugged into a CT dosimetry calculation program such as ImPACT for calculation of CTDI*w, CTDIvol, dose-length-product, and critical organ doses. Furthermore, a graphical presentation of the "mA" vs slice location can be produced with Excel. The chart generated by Excel shows the variation in tube current as a function of slice location. The area under the mA curve is equal to a rectangle of "average mA" x "distance". Here, the distance is the range covered by the CT scan. The script that the authors have written is able to extract mA values from DICOM CT images, generating a delimited ASCII file for further processing with Microsoft Excel spreadsheet program. After the DICOM images are imported into a personal computer, this semiautomated process of extracting mA values enabled the authors to perform dose calculation for patients undergoing CT examination scanned with automatic mA modulation control.