透過您的圖書館登入
IP:3.238.161.165
  • 學位論文

X光射束跟效應下醫學影像改善之探討

Heel Effect Improvement Assessment of X-ray Medical Images

指導教授 : 蘇振隆
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


目前使用X射線來進行醫療診斷的需求日益增加,同時對於影像的品質也日益重視。由於X光管的跟效應會造成影像惡化現象,進而影響影像之判讀。本研究之目的,為瞭解跟效應的分佈情況及發展一處理系統來改善其產生之影像。 在研究中,由不同整流系統及陽極靶角度的X光管,在不同的管電壓和管電流下來照射軟片。計算由軟片上所產生的黑化度偏差來評估跟效應,並探討所造成的影像惡化,然後使用均勻化修正來補償復原影像。在跟效應探討中,X光管設定於20mAs,並改變管電壓於50至120kV的範圍:發現採用單相全波電源供應系統者,其跟效應對軟片黑化度所產生的偏差,於靶角度為17度者,其範圍介於3%至6.09%。陽極靶角度為12度者其黑化度偏差範圍介於8.03%至12.89%。採用靶角度為17度的變頻電源供應系統者,其軟片黑化度偏差範圍則介於2.14%至3.85%。在管電壓90kV管電流於50至500mA的設定範圍:採用全波高壓直流供應系統者,其跟效應所產生的黑化度偏差,於靶角度為17度者,其範圍介於1%至3.6%。陽極靶角度為12度者,其黑化度偏差範圍介於4.5%至11.5%。採用靶角度為17度的變頻電源供應系統者,其軟片黑化度偏差範圍則介於3%至4.2%。平均而言,改變管電流其跟效應的變化較大。改變管電壓其跟效應的變化較不明顯。變頻電源供應系統者,其跟效應變化現象較小。12度陽極靶之跟效應變化較17度為大。 在影像復原探討中,使用改良後均勻化修正方法後:於17度陽極靶變頻系統所產生的影像灰階平均及標準偏差由46336±509改善至46336±285。於12度陽極靶全波系統所產生的影像灰階平均及標準偏差由36862±860改善至36862±287。因此對不同系統其標準偏差均可明顯下降。此影像處理時間約需2至3秒,對於改善跟效應影像所需時間單相全波與變頻則沒有明顯差別。 由實驗結果顯示,在不同高壓直流供應系統,所產生的跟效應對軟片黑化度的偏差影響不同。12度陽極靶之跟效應變化較17度為明顯。在影像復原上使用均勻化修正對於12度與17度X光陽極靶所產生的影像都可有效的改善影像品質。此方法未來將可應用於醫療影像儲存暨傳輸系統(PACS)中的X光影像系統。

並列摘要


The need of X-ray diagnosis system is increasing and then the task of quality control for images becomes more important. The degrading image due to the heel effect of X-ray can affect the interpolation of image. The purpose of this study is to find factors that cause the heel effect and to develop a system to restore the image. For three different X-ray machines, four parameters which including the tube current, the tube voltage, the target angles of anode, and the power supply type of X-ray system, may affect the heel effect is investigated. The results show that the deviation of the grey scale of the film varies from 3% to 6.09% when the tube voltage changed from 50kV to 120kV for a system with full wave rectification power supply and 17o target angle. Moreover, it varies from 8.03% to 12.89% for a system with 12o target angle; and varies from 2.14% to 3.85% for a system with high frequency rectification power supply and 17o target angle. Again, the deviation of the grey scale of the film varies from 1% to 3.6% when the tube current changed from 50mA to 500mA for a system with full wave rectification power supply and 17o target angle. Besides, it varies from 4.5% to 11.5% for a system with 12o target angle, and varies from 3% to 4.2% for a system with high frequency rectification power supply and 17o target angle. The modified uniformity correction method is introduced to restore the degraded image which due to the heel effect. The degraded image is produced from a phantom which exposure by a X-ray system. The results show the mean and standard deviation of gray scale for region of interesting (ROI) of image is improved from 46336±509 to 46336±285 for a system with high frequency rectification power supply and 17o target angle. It is improved from 36862±860 to 36862±287 for a system with full wave rectification power supply and 12o target angle. After this process, the standard deviation of grey scale for ROI is reduced for both systems. The consuming time of image restoration for the compensation of heel effect is only 2~3 seconds for both X-ray machines. We concluded that the heel effect is varied when different tube voltage and tube current applied. The power supply systems and the target angle of the X-ray machine are also affecting. To compensate for the heel effect and improve the image, the technique of uniformity correction we introduced here is effective, useful, and potential to be applied in picture archiving and communication system (PACS) of X-ray diagnosis system.

參考文獻


[20]張榮華,許賓杰,”HITACHI DHF-153HII X光品質探討”,中華民國放射線技術第十二屆第三次會員大會暨年會學術研討會,2000,台北。
[1]Bontrager KL, ”Textbook of Radiolographic Positioning and Related Anatomy” p36, Mosby, 2001, USA.
[2]Fritz SL and Livingston WH ”A comparison of computed and measured heel effect for various target angles”, Medical Physics, 9 (2), pp216-9, 1982.
[4] West JD; mayor MB; Collier JP ”Potential errors inherent in quantitative densitometric analysis of orthopedic radiographs A study after total hip arthroplasty” Journal of Bone and Joint Surgery, 69(1), pp58-64, 1987.
[6]Larsson JP, Persliden J, Sandborg M and Carlsson GA, ”Transmission ionization chambers for measurements of air collision kerma integrated over beam area Factors limiting the accuracy of calibration”, Physics in Medicine and Biology, 41 (11), pp2381-98, 1996.

延伸閱讀