本文選題：口腔CT　切入點：金屬偽影　出處：《東北大學》2013年碩士論文　論文類型：學位論文

【摘要】：目前,口腔CT在齒科疾病的檢查中發(fā)揮著越來越重要的作用,它通過測量X射線在各個方向上透過人體斷層時的累積衰減系數(或稱投影),由計算機重建得到整個橫斷面上X射線衰減系數的分布,從而幫助醫(yī)生診斷病灶部位�？谇籆T相比全身CT具有劑量小、空間分辨率高等優(yōu)點,但是口腔CT在臨床檢查中面臨著很多問題和挑戰(zhàn),其中CT值的準確性和金屬偽影尤為突出。因此,本文開展了CT值與金屬偽影校正方法的研究,主要內容有：針對金屬偽影,本文主要從三個方面進行校正：一是投影數據補償,在圖像重建前,對金屬投影區(qū)域的數據采用插值的方法進行補償,以此來達到去除金屬偽影的目的。本文實現了線性插值與樣條插值；二是迭代重建校正,將金屬投影區(qū)域的數據舍棄掉,利用剩余數據進行迭代重建校正。對于該類方法本文實現了基于代數重建的ATR校正法以及基于統(tǒng)計迭代的MLEM方法；三是提出了一種基于投影數據加權的金屬偽影校正法,此方法主要利用三部分的投影數據,分別是模板圖像的投影數據、樣條插值后的投影數據、修正后的投影數據,對這三部分投影數據采用加權的方式得到修正后的投影數據,最后重建獲得校正后的圖像。該方法既實現了每個投影角度下投影數據的連續(xù)性,也實現了角度與角度間投影數據的連續(xù)性,有效的去除了金屬所引起的條狀偽影。針對目前商用CT值校正方法的缺點,本文提出一種適用于口腔CT的CT值校正方法,該方法充分考慮到口腔CT重建圖像中存在的噪聲和偽影,在進行CT值校正前加入了圖像預處理的步驟,有效降低了圖像噪聲以及偽影對CT值校正所帶來的誤差。將基于投影加權的校正方法應用于模擬金屬偽影圖像,圖像信噪比提升22%；將改進的CT值校正方法應用實驗室口腔CT,水模圖像感興趣區(qū)域CT值的準確性最高提升53%。
[Abstract]:At present, oral CT plays a more and more important role in the examination of dental diseases. By measuring the cumulative attenuation coefficient (or projection) of X-rays passing through human faults in all directions, the distribution of X-ray attenuation coefficients on the whole cross-section is obtained by computer reconstruction. Oral CT has the advantages of low dose and high spatial resolution, but it faces many problems and challenges in clinical examination. The accuracy of CT value and metal artifact are especially prominent. Therefore, this paper studies the correction method of CT value and metal artifact. The main contents are: for metal artifact, this paper mainly corrects from three aspects: first, projection data compensation, Before image reconstruction, interpolation method is used to compensate the data of the metal projection region to remove metal artifacts. In this paper, linear interpolation and spline interpolation are realized, and iterative reconstruction correction is used. The data of the metal projection region is discarded and the residual data is used for iterative reconstruction correction. For this kind of method, the ATR correction method based on algebraic reconstruction and the MLEM method based on statistical iteration are implemented in this paper. Third, a metal artifact correction method based on weighted projection data is proposed. This method mainly uses the projection data of three parts, namely, the projection data of template image, the projection data after spline interpolation, and the revised projection data. The modified projection data is obtained by weighted method, and the corrected image is reconstructed. This method not only realizes the continuity of the projection data at each projection angle, The continuity of projection data between angle and angle is realized, and the strip artifacts caused by metal are removed effectively. In view of the shortcomings of current commercial CT correction methods, this paper proposes a CT correction method suitable for oral CT. In this method, the noise and artifacts in the reconstruction image of oral CT are fully considered, and the steps of image preprocessing are added before the correction of CT value. The error caused by image noise and artifacts to CT value correction is reduced effectively. The projection weighted correction method is applied to simulate metal artifact image. The improved CT correction method is applied to the laboratory oral CT.The accuracy of CT value of the region of interest in the water model image is increased by 53%.
【學位授予單位】：東北大學
【學位級別】：碩士
【學位授予年份】：2013
【分類號】：R816.98;TP391.41
，

本文編號：1636099