利用田口方法與自製平版塊規評量核醫檢查之Tc-99m掃描影像品質的最小可分辨差異值

To quantify the minimum detectable difference of Tc-99m gamma camera scanned imaging using Taguchi analysis and an indigenous planar gauge

劉宛茹(WAN-JU LIU)

指導教授 ： 潘榕光

中臺科技大學/健康科學院/醫學影像暨放射科學系暨研究所/碩士(2020年)

摘要

目的：本實驗利用田口方法及自製平版塊規尋找核子醫學Tc-99m檢查最佳影像品質之設定，並計算最小可分辨差異值，以期能提升臨床檢查的影像品質。
材料與方法：實驗根據骨骼掃描常用的造影條件，選擇可能會影響影像解析度的六個因子：準直儀、距離、總計數、能窗、矩陣大小、放大倍率，進行田口L18的實驗。將上述因子帶入田口L18直交表進行設定，將99mTC-MDP染色，注入自製平版塊規，使用GE Infinia Hawkeye 4 SPECT/CT進行掃描；並將不同攝影條件所得出之影像，利用田口方法分析，尋找最佳之攝影條件。
結果：利用田口方法分析模擬核子醫學骨頭掃描之影像，發現影響影像解析度因子最主要為準質儀及距離。得出影像最佳化的條件設定為準質儀(LEHR)、距離(5 cm)、總計數(150k counts)、能窗(20%)、矩陣大小(128x128)、放大倍率(1.28)。將臨床常用設定影像與田口最佳化設定影像，使用MDD方法進行分析，臨床常用設定影像之MDD為2.94 mm，田口最佳化設定影像之MDD為2.15 mm，因此可知田口最佳化設定之影像解析度優於臨床常用設定。使用專家系統評估影像解析度，會因不同評分者對空間分辨率及其視覺能力的不同，產生較主觀評估，因此容易產生人為誤差；而利用MDD方法評估影像解析度，可以將解析度量化計算，因此所得出得結果較為客觀且更有說服力。
結論：所以利用田口方法及最小可分辨方法可找出影像解析度最佳之條件進行掃描，可提供醫師更為清晰的影像，輔助醫師進行影像判斷。

關鍵字

骨骼掃描 ； 田口方法 ； 影像最佳化 ； 最小可分辨差異值

並列摘要

Aim: the optimization of the T-99m gamma camera scanned images was analysis using Taguchi methodology and an indigenous planar gauge. The optimal images were further quantified by the calculation of minimum detectable difference (MDD).
Materials and methods: the factors for gamma camera scanning were optimized according to Taguchi unique L18 orthogonal array. Six factors were assigned to reorganize onto 18 different combinations and still hold the accuracy as those were evaluated in a full factor combination (2×3×3×3×3×3=486). The six factors were (A) collimator, (B) distance, (C) total counts, (D) window width, (E) matrix size, and (F) magnification, furthermore, the planar gauge was infused with Tc-99m-MDP radio-activated solution with dyeing in blue for easy recognizing in the measurement.
Results: the optimal setting for gamma camera in nuclear examination was LEHR collimator, 5 cm SID, 150k total counts, 20% window width, 128×128 matrix size, and 1.28 magnification. The dominant factor was either collimator or SID. The quantified MDD was 2.15 mm that was superior to the conventional setting, 2.94 mm in clinical examination. The quantified MDD can help greatly in the optimizing process. If without the MDD, the imaging quality can only be ranked by radiologists’ visual inspection that may cause biased and, thus, mislead the optimal suggestion.
Conclusion: the Taguchi optimization proves to be a useful technique in quantifying the MDD for gamma camera in the routine nuclear examination. The quantified MDD helps radiologists to hold solid idea in maintaining the quality assurance of gamma camera.

並列關鍵字

Bone scanning ； Taguchi ； imaging quality ； minimum detectable difference (MDD)

參考文獻

1. Krasnow AZ, Hellman RS, Timins ME, et al., Diagnostic bone scanning in oncology., Semin Nucl Med 1997;27:107-141.

2. D S Schauwecker., The scintigraphic diagnosis of osteomyelitis., American Journal of Roentgenology. 1992;158: 9-18.

3. 衛生福利部, 主要死亡原因死亡人數及死亡率, 2019-10.

4. Lionel S Zuckier, Patrick Martineau., Altered Biodistribution of Radiopharmaceuticals Used in Bone Scintigraphy., Semin Nucl Med. 2015 Jan;45(1):81-96.

5. Yang, David C, Ratani Rita S, Mittal Pardeep K, et al., Radionuclide Three-Phase Whole-Body Bone Imaging., Clin Nucl Med. 2002 Jun;27(6):419-26.