建立以感興趣區域為基礎的標定功能進行數位放射影像偵測板的快速檢測與校正

The Rapid Detection and Calibration of Digital Radiographic Imaging Panels are achieved through the Establishment of a Labeling Function based on Region of Interest

蕭文田(Wen-Tien Hsiao) ； 蕭惠馨(Hui-Hsin Hsiao) ； 趙樹禮(Shu-Li Chao) ； 賴律翰(Lu-Han Lai) ； 杜俊元(Chun-Yuan Tu)

《台灣應用輻射與同位素雜誌》 19卷1期 (2023/03) Pp. 2223-2228

摘要

感興趣區域標定功能進行醫學影像診斷已相當廣泛應用，並成為現代數位醫學影像儀器必備的基本配備功能。此功能可標定出區域內像素灰階的平均值及灰階值的標準偏差和誤差，以作為臨床影像診斷的參考。本研究利用此功能建立醫學成像系統的快速校正方式，以解決臨床遇到影像有失真懷疑之問題。在成像系統校正前後，以儀器最低曝露劑量條件於43 ×43 cm數位放射影像板取得影像樣本，校正方式依據儀器專用校正軟體及程序進行，將所取得影像樣本進行全區域分區之感興趣區域標定，並將所得數值進行One way ANOVA統計分析，比較不同區域產生偏差及誤差之可能性，藉以瞭解硬體穩定性，分析問題區域。透過此方式建立數位放射攝影品質硬體之檢測，進而快速排除臨床現場問題。研究結果顯示，以中心射束為基準，對於所標定的各區域相對比較，p值皆小於0.001，校正程序也可在30分鐘內完成。此方式能快速檢測數位放射影像板品質衰退狀況，並解決系統頻繁使用所產生的失真衰退問題。所建立的檢測方式可提供做為醫療放射影像儀器品質管理檢測方式及故障排除之參考。

關鍵字

感興趣區域 ； 灰階值 ； 中心射束

並列摘要

The use of region-of-interest (ROI) labeling for medical image diagnosis has become widely applied and is a fundamental feature of modern digital medical imaging equipment. This function allows for the labeling of the mean grayscale value and the standard deviation and error of these grayscale values within the ROI, providing a reference for clinical image diagnosis. In this study, we utilized this feature to establish a rapid calibration method for medical imaging systems to address suspected image distortion issues encountered in clinical practice. We obtained digital radiography samples using the lowest exposure conditions on a 43x43 cm digital radiography plate before and after calibration of the imaging system. The calibration was performed using instrument-specific calibration software and program. We divided the obtained digital radiography samples into ROIs across the entire region, and the resulting values were subjected to one-way ANOVA statistical analysis to compare potential deviations and errors generated in different regions, assess hardware stability, and analyze problematic areas. Through this method, we established a hardware quality detection system for digital radiography that can quickly resolve clinical issues. The results showed that the p-values for the comparison of the various ROI labeled areas relative to the central ray were all less than 0.001, and the calibration procedure could be completed within 30 minutes. This method can rapidly detect the degradation of digital radiography plate quality and resolve the distortion degradation issues resulting from frequent system use. The established detection method can serve as a reference for quality management and troubleshooting of medical radiography equipment.

並列關鍵字

Region of Interest ； Grayscale Values ； Center Ray

參考文獻

Kaur, M., Wasson, V. (2015). ROI based medical image compression for telemedicine application. Procedia Computer Science, 70, 579-585.

Bushong, S. C. (2020). Radiologic science for technologists e-book: physics, biology, and protection. Elsevier Health Sciences.

Castleman, K. R. (1996). Digital image processing. Prentice Hall Press.

Barnes, G. T., Lauro, K. (1989). Image processing in digital radiography: basic concepts and applications. Journal of digital imaging, 2, 132-146.

Uffmann, M., Schaefer-Prokop, C. (2009). Digital radiography: the balance between image quality and required radiation dose. European journal of radiology, 72(2), 202-208.