最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影

行政院國家科學委員會專題研究計畫 成果報告

最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡 對比劑造影

研究成果報告(精簡版)

計 畫 類 別 ： 個別型

計 畫 編 號 ： NSC 100-2320-B-011-001-

執 行 期 間 ： 100 年 08 月 01 日至 101 年 07 月 31 日 執 行 單 位 ： 國立臺灣科技大學電機工程系

計 畫 主 持 人 ： 沈哲州

計畫參與人員： 大專生-兼任助理人員：游聲彰 大專生-兼任助理人員：邢永騏

報 告 附 件 ： 出席國際會議研究心得報告及發表論文

公 開 資 訊 ： 本計畫可公開查詢

中 華 民 國 101 年 08 月 06 日

中 文 摘 要 ： 超音波諧波影像常用以提升微氣泡對比劑的偵測能力，但其 效果常限於組織諧波與諧波溢漏；本計劃提出最佳發射相位 法利用組織諧波與溢漏諧波之間的相位關係來抑制背景組織 區域的諧波信號，以提高影像的 CTR 值。本研究中除驗證本 最佳發射相位法的理想效果之外，也嘗試將所需要發射之任 意連續波形以臨床儀器常用之二元碼實現以探討此技術之實 用性。我們將透過 Sigma-Delta 及編碼微調兩步驟，把任意 連續波形調整為二元碼，並分別使用單一探頭及陣列探頭發 射編碼後波形以驗證是否仍可產生組織諧波抑制之效果。實 驗結果顯示二元化之波形仍可產生相當之組織抑制效果，但 因其溢漏諧波成分之相角及強度均隨發射相位呈現不規則變 化而使得組織抑制效果並不如原連續波形之規律，故臨床應 用時必須特別依賴事前校正來選擇最佳發射相位而影響其實 用性。

中文關鍵詞： 諧波影像、諧波溢漏、組織諧波、二元碼、Sigma-Delta 法、編碼微調、最佳發射相位法

英 文 摘 要 ： Ultrasonic harmonic imaging is routinely used to improve detection of contrast agent microbubbles.

However, harmonic contrast detection is limited by tissue harmonics and leakage harmonics. We have previously proposed the method of optimal transmit phasing to increase CTR by relatively phasing the two harmonic components in the tissue background to

cancel each other. Since most clinical systems can only transmit binary waveforms, the binary

transformation of continuous transmit signal using the Sigma-Delta method is studied for optimal

transmit phasing. Our results indicate that, though tissue suppression is still achievable with binary transmit waveforms, the extent of tissue suppression becomes uncertain with transmit phasing.

Consequently, the pre-calibration becomes very

critical to determine the optimal transmit phase. It may clinically limit the practicability of optimal transmit phasing.

英文關鍵詞： harmonic imaging、harmonic leakage、tissue harmonic, binary code, Sigma-Delta、code tuning, optimal

transmit phasing

行政院國家科學委員會補助專題研究計畫 行政院國家科學委員會補助專題研究計畫 行政院國家科學委員會補助專題研究計畫

行政院國家科學委員會補助專題研究計畫期 期 期末成果 期 末成果 末成果 末成果報告 報告 報告 報告

最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影

計畫類別：■ 個別型計畫 □ 整合型計畫 計畫編號：NSC 100－2320－B－011－001

執行期間： 100 年 8 月 1 日至 101 年 7 月 31 日

計畫主持人：沈哲州 共同主持人：

計畫參與人員：游聲彰 刑永騏

成果報告類型(依經費核定清單規定繳交)：■精簡報告 □完整報告

本成果報告包括以下應繳交之附件：

□赴國外出差或研習心得報告一份

□赴大陸地區出差或研習心得報告一份

■出席國際學術會議心得報告及發表之論文各一份

□國際合作研究計畫國外研究報告書一份

處理方式：除產學合作研究計畫、提升產業技術及人才培育研究計 畫、列管計畫及下列情形者外，得立即公開查詢

□涉及專利或其他智慧財產權，□一年■二年後可公開查詢

執行單位：國立台灣科技大學電機系

中 華 民 國 101 年 7 月 31 日

行政院國家科學委員會專題研究計畫期末成果報告

最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影

計畫編號：NSC 100-2320-B-011-001 執行期限：100 年 8 月 1 日至 101 年 7 月 31 日

主持人：沈哲州 國立台灣科技大學電機系

一一

一一、、、、中文摘要中文摘要中文摘要中文摘要

超音波諧波影像常用以提升微氣泡對比劑的 偵測能力，但其效果常限於組織諧波與諧波溢 漏；本計劃提出最佳發射相位法利用組織諧波與 溢漏諧波之間的相位關係來抑制背景組織區域的 諧波信號，以提高影像的 CTR 值。本研究中除驗 證本最佳發射相位法的理想效果之外，也嘗試將 所需要發射之任意連續波形以臨床儀器常用之二 元碼實現以探討此技術之實用性。我們將透過 Sigma-Delta 及編碼微調兩步驟，把任意連續波形 調整為二元碼，並分別使用單一探頭及陣列探頭 發射編碼後波形以驗證是否仍可產生組織諧波抑 制之效果。實驗結果顯示二元化之波形仍可產生 相當之組織抑制效果，但因其溢漏諧波成分之相 角及強度均隨發射相位呈現不規則變化而使得組 織抑制效果並不如原連續波形之規律，故臨床應 用時必須特別依賴事前校正來選擇最佳發射相位 而影響其實用性。

關鍵詞關鍵詞關鍵詞

關鍵詞：諧波影像諧波影像諧波影像諧波影像、、、、諧波溢漏諧波溢漏諧波溢漏、諧波溢漏、、組織諧波、組織諧波組織諧波組織諧波、、、、二元二元二元二元 碼碼碼

碼、、、、Sigma-Delta 法法法、法、、編碼微調、編碼微調編碼微調編碼微調、、、、最佳發射相位最佳發射相位最佳發射相位最佳發射相位 法法法

法

Abstract

Ultrasonic harmonic imaging is routinely used to improve detection of contrast agent microbubbles.

However, harmonic contrast detection is limited by tissue harmonics and leakage harmonics. We have

previously proposed the method of optimal transmit phasing to increase CTR by relatively phasing the two harmonic components in the tissue background to cancel each other. Since most clinical systems can only transmit binary waveforms, the binary transformation of continuous transmit signal using the Sigma-Delta method is studied for optimal transmit phasing. Our results indicate that, though tissue suppression is still achievable with binary transmit waveforms, the extent of tissue suppression becomes uncertain with transmit phasing.

Consequently, the pre-calibration becomes very critical to determine the optimal transmit phase. It may clinically limit the practicability of optimal transmit phasing.

Keywords: harmonic imaging、harmonic leakage、

tissue harmonic, binary code, Sigma-Delta、code tuning, optimal transmit phasing

二二

二二、、、、緣由與目的緣由與目的緣由與目的 緣由與目的

超音波由於具有非侵入性、安全性及即時性 等優點，因而被廣泛運用於醫學領域中；舉凡心 臟影像、腫瘤偵測、血液流速等，皆可藉由超音 波輔助得到所需之資訊。近期超音波影像的發展 趨勢以諧波影像為主，其主要是透過設計對應之 帶通濾波器，濾出所需的諧波信號再加以成像；

相較於傳統使用基頻信號成像之方法，其組織諧 波較不易受相位偏差的影響，在對比劑諧波上亦 有較佳的影像對比度。

由於諧波影像是利用介質對聲波的非線性反

應所產生的信號來成像[1]，因此信號中含有對比 劑諧波及組織諧波的成分；當聲壓加大時，對比 劑及組織的諧波信號皆會增強，故常使對比劑諧 波影像的 CTR 值降低。另外，當對比劑諧波頻帶 中含有基頻溢漏成分時，亦會降低其 CTR 值[2,3]。

在本實驗室先期的研究中，已提出利用最佳 發射相位法對組織區域的諧波信號做抑制來提高 對比劑諧波影像之 CTR 值[4]；然而在激發波形的 設計上為任意波形，實現上須經由任意波形產生 器發射，在臨床的運用及評估上也因而受限(現有 之臨床儀器多以發射二元碼為主)。本研究將嘗試 以 Sigma-Delta 法為基礎之二元碼轉換[5]來實現 最佳發射相位法，以探討其在臨床上之實用性。

三三

三三、、、、研究原理研究原理研究原理研究原理

A.最佳發射相位法

以連續波發射信號為例，當發射信號 x(t)中存 有諧波溢漏成分時可用基頻頻率成分與第二諧波 頻率成分組合來表示如下，且分別對應的振幅為

a

) ) 2 ( 2 cos(

) 2

cos(

)

(

x

=

+

+

+

式中之φ_f表示為基頻信號的相位，而

φ

y

x(t)+b

x

=b1

a

cos(2π(2f

φ

a

2

cos(2π(2f

φ

φ

係數 b1及 b2分別為線性信號及第二諧波信號的振 幅。因此回傳信號經帶通濾波器後，式子等號右 邊第一項及第二項分別代表為諧波溢漏信號及組 織諧波信號。如圖(一)所示，若基頻發射信號相 位為

φ

φ

φ

2 。此時若在二倍頻帶(2f0)部分有相位為

φ

_φ

_φ

_π

故可以對組織背景信號產生抑制作用以提高對比 劑諧波影像之 CTR 值；此抑制作用於組織諧波信 號及溢漏諧波信號強度相近時將達到最佳效果。

在一般的超音波系統中以發射脈衝波(pulse wave) 為 主 ， 假 設 基 頻 頻 譜 為 一 高 斯 形 狀 如 圖 (二)，此時其組織諧波頻譜亦近似於高斯；雖然 諧波頻帶與溢漏頻帶重疊範圍較寬，但仍以諧波 溢漏信號與第二諧波信號的大小值相同時(fL處) 其相消效果最佳。

B.二元編碼轉換

假定所設計之任意波形為 x(t)，信號於發射時 需先經過探頭之頻率響應 ht(t)，故實際發射於受 測物體的信號可表示為 x(t)

⊗ h

⊗ h

⊗ h

1.Sigma-Delta 調變

此步驟為初步將任意波形轉換為二元編碼，

其發射後之頻譜與原始波形大致接近。由於調變 目的為將波形轉換為二元碼，故在調變前須先將 原始任意連續信號 x(t)以取樣頻率 fs轉為離散信 號 x(n)，而 x(n)再經 Sigma-Delta 調變後產生二元 碼 y(n)。圖(三)為 Sigma-Delta 調變之方塊圖，其 中 e(n)為量化誤差值。

(圖三)

根據此方塊圖，可將輸入及輸出信號關係整 理為下式:

y(n)=x(n-1)+e(n)-e(n-1)；

假設輸入信號 x(n)之振幅範圍為±V，則輸出信號

y(n)為{+1,-1}，y(n)在透過重建濾波器將高頻雜

(圖二) (圖一)

2.編碼微調

一般而言，取樣時選擇較高的取樣頻率可使 二元化前後通過探頭發射之波形與頻譜更為相 近，然而大部分系統可接受的取樣頻率皆有一定 的限制，因此經 Sigma-Delta 調變所得的 y(n)仍需 經過調整；透過此演算法可將 N 個位元之 y(n)調 整為一新的編碼 y’(n)。在此演算法中，一次可做 調變的編碼最多為 N2個連續位元(本研究所採用 之 N2為 12)，並以此為單位反覆移動循環，產生 新的編碼。我們可定義 d[y(n)]為 y(t;fs)及 x(t)間的 相似性，則 d[y(n)]=∥[y(t;fs)-x(t)

⊗ h

∥．∥為 L²norm。假設 y1(n)為調整後的編碼，若

d[y

y(n)。若繼續對 y

y

四 四四

四、、、、實驗架構實驗架構實驗架構實驗架構

A.超音波儀器之重要規格說明:

本研究採用之儀器為 Ultrasonix RP system，

並搭配 C7-3 之陣列探頭，最佳發射及接收頻帶為 3~7MHz。此儀器可使用於臨床及學術研究，並具 有可直接擷取 RF 信號加以分析之特色。量測時 以 C 語言為其控制程式，透過參數的調整可改變 探頭的聚焦深度、取樣頻率(發射端取樣頻率可調 為 40MHz 或 80MHz，接收端取樣頻率則固定為 40MHz)；可自行輸入設計後之二元碼，以發射所 需之特定頻率信號。接收時，可接收回傳之 RF 信號，並使用 Matlab 程式對信號加以分析處理。

B.實驗方法

1.求探頭頻率響應之架構與求法

以陣列探頭 C7-3 為例；透過程式控制模式發 射一個位元碼至線仿體，意即產生一脈衝信號，

則所接收之回傳信號即為探頭頻率響應之迴旋積

h

⊗ h

H(f)=A(f)e

2.單一探頭驗證實驗架構

本實驗採用單一探頭發射接收的方式來驗證 發射高斯及二元碼波形之諧波信號，實驗選用一 中心頻率為 3.5MHz 的探頭來發射與接收信號，

並在離探頭表面聚焦深度 70mm 處放置線仿體(釣 魚線)作為反射體。發射信號為中心頻 2.5MHz，

頻寬為 80%之高斯波形，其相位由-170 度每隔 10 度改變至 180 度，由 80MHz 任意波形產生器(AFG) 產生，透過微波功率放大器將訊號放大後經由雙 工器(Diplexer)而傳送至探頭發出信號，再以同一 探頭接收回傳信號經由雙工器而傳送至超音波脈 衝接收器來放大，最後透過 20MHz 之類比數位轉 換卡，將實驗資料傳到電腦分析。

3.B-mode 影像模式架構

我們也可透過 B-mode 影像模式來觀察組織 諧波部分不同發射相位之差異性；我們使用 Ultrasonix RP system 並以程式控制模式發射 36 個 不同相位編碼後之波形，聚焦深度設定為 30mm。

如圖(四)所示，使用一圓柱形仿體，並於仿體中 挖洞，洞內灌入蒸餾水；將探頭放置在仿體側面，

並於交接處塗抹凝膠使其阻抗匹配，探頭表面距 洞約為 30mm。接收到的訊號將透過帶通濾波器，

將諧波成分濾出成像。

五五

五五、、、、實驗結果實驗結果實驗結果實驗結果

A.探頭之頻率響應

圖(五)(a)、(b)分別為陣列探頭C7-3與單一探 頭頻率響應之波形及頻譜，本研究編碼部分將採 用此頻率響應加以設計。

B.波形設計之結果

圖 ( 六 ) 為 將 高 斯 -170 度 波 形 加 以 編 碼 之 結 (圖四)

果。圖中藍線表原始之高斯信號，綠線及粉紅線 則分別表示初步Sigma Delta及編碼調整後的結 果；三組信號皆與探頭之頻率響應做迴旋積，以 模擬實際發射之波形。由此圖可見編碼調整後的 波形與頻譜較Sigma Delta後皆更接近原始信號，

然而由於諧波影像需預留頻寬以擷取二次諧波之 信號，故其重建濾波器(探頭頻率響應)在高頻仍有 相當增益，因此二元碼波形在高頻部分的頻譜與 原高斯皆有落差。

(a) C7-3陣列探頭

0 50 100 150 200 250 300 350 400

-15 -10 -5 0 5 10 15 20 25

0 2 4 6 8 10 12 14 16 18 20

-40 -35 -30 -25 -20 -15 -10 -5 0

Frequency(MHz)

dB

(b) 3.5MHz單一探頭

0 50 100 150 200 250 300 350 400

-8 -6 -4 -2 0 2 4 6x 10^-14

Time(us)

Amplitude(V)

0 1 2 3 4 5 6 7 8 9 10

-40 -35 -30 -25 -20 -15 -10 -5 0

Frequency(MHz)

dB

(圖五)

300 400 500 600 700 800

-8 -6 -4 -2 0 2 4 6

8 Orignal Transmit

Sigma Delta Code Tuning

0 2 4 6 8 10 12 14 16 18 20

-100 -50 0 50

Frequency (MHz)

dB

Orignal Transmit Sigma Delta Code Tuning

(圖六)

C.發射不同相位波形實驗結果

1.單一探頭驗證結果

圖(七)(a)為發射高斯信號時，改變傳輸相位 所得之組織信號頻譜圖，由圖中可見當改變傳輸 波形相位時，在基頻頻帶內的振幅並無太大的變 化；但在第二諧波頻帶(5MHz)附近於120度可見 一明顯的抑制點；相反的，在相位為-60度時其諧 波成分較其他相位皆來的高。將36個相位之二倍 頻能量做相加；如圖(b)所示，可明顯的看出其組 織區諧波強度隨發射相位成規律之週期變化。

(a) (b)

Phase(degree)

Frequency(MHz)

-150 -100 -50 0 50 100 150

0

1

2

3

4

5

6

7

-7 -6 -5 -4 -3 -2 -1 0

-150 -100 -50 0 50 100 150

1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4 1.45 1.5

Phase(degree)

Second Harmonic Intensity

(圖七)

將發射波形改為二元碼重新實驗，並繪出改 變傳輸相位所得之組織信號頻譜圖，圖(八)(a)基 頻頻帶內的振幅強度仍大致相同，但第二諧波頻 帶附近的振幅卻出現不規則的變化，因此在圖(b) 中之二倍頻能量雖然有隨相位之明顯變化，但其 諧波能量之大小與發射相位並無規律之關連性。

(a) (b)

Phase(degree)

Frequency(MHz)

-150 -100 -50 0 50 100 150

0 1 2 3 4

5 6 7

-5 -4 -3 -2 -1 0

-150 -100 -50 0 50 100 150

1 1.5 2 2.5 3 3.5 4

Phase(degree)

Second Harmonic Intensity

(圖八)

2.B-mode影像結果

使用陣列探頭發射不同相位編碼後波形至仿 體，並將回傳信號中基頻及二倍頻成分分別做濾 波成像。圖(九)(a)為二次諧波成像36相位之組織 部分(圈選區域)強度，雖然強度上有明顯差異但並 未隨相位不同而有規律的改變；取其抑制及非抑 制相位影像，影像中黑色圓形區域為洞隙灌入水 之部分，較亮區域則為組織，如圖(b)、(c)可發現 組織區域在諧波強度上有明顯之改變。圖(d)為基 頻成像36相位之組織強度，其強度較無明顯改 變，觀察與二次諧波成像抑制及非抑制相同相位 之影像如圖(e)、(f)，在組織亮度上亦無明顯的變 化。

(a) (b) phi=80° (c) phi=-120°

-150 -100 -50 0 50 100 150 3

3.5 4 4.5 5 5.5 6 6.5 7 x 10^-3

Phase(degree)

B-mode (Positive)

Width (mm)

Depth

10 15 20 25 30

0 10 20 30 40 50 60

5 10 15 20 25 30

B-mode (Positive)

Width (mm)

Depth

10 15 20 25 30

0 10 20 30 40 50 60

5 10 15 20 25 30

(d) (e) phi=80° (f) phi=-120°

-150 -100 -50 0 50 100 150

0.01 0.0105 0.011 0.0115 0.012 0.0125 0.013 0.0135 0.014 0.0145

Phase(degree)

B-mode (Positive)

Width (mm)

Depth

10 15 20 25 30

0

10

20

30

40

50

60

5 10 15 20 25 30

B-mode (Positive)

Width (mm)

Depth

10 15 20 25 30

0 10 20 30 40 50 60

5 10 15 20 25 30

(圖九)

3. in-vivo影像

我們使用紐西蘭兔對其心臟部份進行造影，為 避免心臟的快速運動造成比較上的困難，RP影像 系統採用在同一影像掃瞄線上連續發射兩種相位 波形的成像方式以限制兩種相位波形對應之影像 只差異一個PRI的時間。(圖十)結果顯示兩者影像 在對比劑區域(紅框處)之強度差異不大，但抑制相 位確實可以對心肌組織(藍框處)造成相當的抑制 效果，因此抑制相位發射信號相較於非抑制相位 發射信號的CTR可約有5 dB的提昇效果!

(圖十)

六 六六

六、、、、討論討論討論討論

最佳發射相位法中對組織區域之諧波信號抑 制需滿足以下兩條件:其一為溢漏諧波信號與組 織諧波信號互為反向而於向量相加時相消產生抑 制；另外兩者強度亦相近以達最佳抑制效果。

檢 驗 所 設 計 之 高 斯 發 射 信 號 其 基 頻 成 分 (2.5MHz) 及 溢 漏 諧 波 成 分 (4.5MHz) 之 相 角 如圖 (十一)(a)，溢漏諧波的相角大小和基頻有同樣之

變化趨勢，意即確實可利用調整發射基頻之相位 達到相位反向之目的；然而編碼後發射之二元碼 相角如圖(b)所示，其溢漏諧波相角出現不規則變 化，此部分可能造成相角相消時之誤差。

(a) (b)

-150 -100 -50 0 50 100 150

-3 -2 -1 0 1 2 3 4 5 6 7

Phase(degree) 2.5MHz 4.5MHz

-150 -100 -50 0 50 100 150

-3 -2 -1 0 1 2 3 4 5 6 7

Phase(degree) 2.5MHz 4.5MHz

(圖十一)

此外檢查發射高斯各相位頻譜之強度如圖 (十二)(a)可發現在不同相位時其溢漏諧波強度皆 相同，故不同之基頻發射相位僅改變溢漏諧波相 位，此時組織區域諧波之抑制可以控制在某一特 定頻帶(如圖(二)之fL處)；然而編碼後之頻譜強度 如圖(b)，雖然不同相位之基頻強度仍大致相近因 可預期不同基頻發射相位所產生的組織諧波強度 大致相同，但溢漏頻帶的強度卻隨不同相角有極 大的落差，如此便無法確切控制組織諧波與溢漏 諧波產生最明顯相消的頻率位置，此時諧波信號 強度抑制之頻率位置與效果均會因不同發射相位 而不規則變動，因此必須依賴事前校正來找出使 組織區域諧波強度最低之發射相位。

(a) (b)

(圖十二)

七 七 七

七、、、、參考文獻參考文獻參考文獻參考文獻

[1] T. Christopher ,”Finite Amplitude Distortion- Based Inhomogeneous Pulse Echo Ultrasonic Imaging ” IEEE Trans. Ultrason., Ferrroelect., Freq. Control., vol. 44 ,pp.

125–139, Jan. 1997.

[2] P. N. Burns, “Harmonic imaging with ultrasound contrast agents,” Clin. Radiol., vol. 51, Suppl. 1, pp. 50–55, 1996.

[3] P. H. Chang, K. K. Shung, S. J. Wu et al., “Second harmonic imaging and harmonic Doppler measurements with Albunex,” IEEE Trans. Ultrason. Ferroelect., Freq.

Control., vol. 42, no. 6, pp. 1020–1027, 6ov. 1995.

[4] C.-C. Shen and Y.-C. Hsieh, “Optimal transmit phasing on

tissue background suppression in contrast harmonic imaging”, Ultrasound Med. Biol. 2008 (Accepted).

[5] Sheng-Wen Huang, Pai-Chi Li, “Arbitrary Waveform Coded Excitation Using Bipolar Square Wave Pulsers in Medical Ultrasound ” IEEE Trans. Ultrason., Ferrroelect., Freq.

Control., vol. 53 no.1 ,pp.106-116,Jan. 2006

1

國科會補助 國科會補助 國科會補助

國科會補助專題研究計畫項下出席國際學術會議心得報告 專題研究計畫項下出席國際學術會議心得報告 專題研究計畫項下出席國際學術會議心得報告 專題研究計畫項下出席國際學術會議心得報告

日期：101 年 07 月 31 日

一、參加會議經過

IUS 會議是醫用超音波相關研究人員的主要會議之一，也是本人每年都會參 與的國際會議，本年度會議於美國奧蘭多舉行，奧蘭多位於佛羅里達州中部，

與台灣之間並沒有直達班機，因此最後不得不在美國境內多次轉機才能抵達，

單程搭轉機時間總計約 24 小時。

二、與會心得

由於本人較專長於非線性諧波影像技術，因此對於非線性諧波相關的研討會 議程特別感興趣，大會中本人注意到一篇由日本熊本大學M. Tanabe教授所屬研究 團隊之相關發表「 A Novel Imaging Method of Coded THI Using Multi Chirp Signals 」其想法 十分有趣，它的主要研究動機是一般諧波影像考慮其訊雜比與穿透深度不足的限 制後往往不得不以編碼發射波形(coded excitation)加強其信號，編碼發射又以 計畫編號 NSC 100－2320－B－011－001

計畫名稱 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 出國人員

姓名 沈哲州 服務機構及

職稱 國立台灣科技大學 會議時間

2011 年 10 月 18 日 至

2011 年 10 月 21 日

會議地點 美國 奧蘭多(Orlando)

會議名稱 2011 IEEE international ultrasonics symposium (2011 IUS) 發表論文

題目

Clutter suppression for harmonic Doppler detection in swept-scan high-frequency system

附件四

2

啾聲信號編碼最為普遍，但在諧波影像中除非以脈衝反相技術完全消去基頻信號 的干擾，否則只要對影像解析度有所要求(也就是發射頻寬要夠大)，則此時基頻 信號必定會對啾聲編碼諧波信號產生干擾，該干擾的後果是產生解碼不良而造成 軸向旁瓣(range sidelobe)。

M. Tanabe教授是採用在頻譜上split and merge的方式來嘗試解決這個問 題，所謂的split是將原本大頻寬的原啾聲發射信號分拆成兩個以上的小頻寬啾聲 信號來發射，其優點是單個小頻寬啾聲發射較不會產生基頻與諧波頻帶之間重疊 干擾，因此個別分開處理時可以維持良好的解碼品質，之後可以再將它們merge在 一起回復原本大頻寬發射時的高軸向解析度。他們的影像結果如下，左圖為使用 大頻寬啾聲發射的諧波影像，右圖為分兩個小頻寬啾聲發射再合併的諧波影像，

可以看出原本軸向旁瓣造成之殘影已經獲得大幅改善!

三、攜回資料名稱及內容

大會會議議程一本以及會議論文CD一片

Clutter suppression for harmonic Doppler detection in swept-scan high-frequency system

Che-Chou Shen¹, Hsin-Hsien Wu¹ and Chih-Kuang Yeh²

1

Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan

2

Department of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, Taiwan

Abstract

—In high-frequency swept-scan Doppler system, the spectral broadening of tissue clutter limits the detectability of low-velocity flow signal. Conventionally, the scanning speed of transducer has to be reduced to alleviate the clutter interference but at the cost of imaging frame rate. With third harmonic (3f

) transmit phasing, the tissue harmonic clutter is suppressed and the cut-off frequency of wall filtering can be reduced to preserve low- velocity flow without compromising the frame rate. Our results indicate that the 3f

transmit phasing effectively reduces the harmonic clutter magnitude and thus improves the flow signal-to-clutter ratio. Compared to the conventional counterpart, the clutter-suppressed color flow and power Doppler images show fewer clutter artifacts and is capable of detecting more low-velocity flow of microbubbles. The resultant color-pixel-density also improves with clutter suppression. For the swept-scan high-frequency system, 3f

transmit phasing is capable of providing effective clutter suppression. With the same achievable scanning speed, the resultant Doppler image has higher sensitivity for low-velocity flow and is less susceptible to clutter artifacts.

Keywords-Harmonic Doppler, 3f0 transmit phasing, Tissue spectral broadening, Clutter suppression, microbubble contrast agents

I. I

Ultrasound high-frequency (> 20 MHz) imaging system has been shown to be highly potential for applications in ophthalmology [1], dermatology [1], and pre-clinical small animal imaging [2] due to its capability to portray fine anatomy with high image resolution. High-frequency Doppler system also improves the detectability of microcirculation in these applications due to the elevated magnitude of blood signal at high frequency [3]. Nevertheless, strong interference from tissue clutter is still one major limitation in high- frequency Doppler detection, especially in the presence of tissue motion. Consequently, the signal contrast of flow to clutter in Doppler detection (i.e., the flow signal-to-clutter ratio (SCR)) remains low. The problem becomes even more severe when a swept-scan high-frequency imaging system is utilized. In the swept-scan system, the image is constructed by scanning the single-element transducer laterally at constant speed and continuously producing image lines over the region of interest during the movement of the transducer. The swept- scan system remains popular for high-frequency imaging since the high-frequency array is not always readily available. As a

result of the swept-scan process, the tissue clutter will spectrally broaden in the Doppler domain even in the absence of tissue motion and thus interfere with the low-velocity blood flow. Specifically, the high-pass wall-filtering aimed at removing the broadened tissue signal also inevitably rejects low-velocity flow signal. Since the extent of tissue spectral broadening is proportional to the mechanical scanning speed, the swept-scan Doppler system suffers from a trade-off between the highest achievable frame rate and the lowest detectable flow velocity.

To avoid the interference from the tissue spectral broadening in the swept-scan system, inter-frame wall- filtering has been proposed to subtract the tissue clutter in one frame from that in the next frame [4]. Nonetheless, when the mechanical scanning is not perfectly stable or when the tissue motion cannot be ignored, the subtraction could fail due to the change of tissue clutter between consecutive frames. For better separation between tissue and blood flow, an alternative is to suppress the tissue clutter magnitude directly in the radio- frequency domain. Tissue suppression has been widely discussed in ultrasound nonlinear imaging of encapsulated contrast microbubbles to provide good image contrast in B- mode. In some of the reported methods, the specific response of contrast microbubbles to acoustic excitation is exploited.

For example, radial modulation of bubble’s oscillation with dual-frequency excitation can help to identify the bubble signal [5]. Harmonic generation of bubbles at sub-harmonic frequency also helps to exclude the background tissue [6].

Even in second harmonic imaging which is readily available in most high-end commercial scanner, tissue suppression is also feasible by pre-modification of the transmit waveforms to produce cancellation of tissue harmonic generation [7]. In the previous studies, the third harmonic (3f

) transmit phasing has been developed to provide effective tissue harmonic suppression by using dual-frequency transmit waveform at both fundamental and 3f

frequencies to produce cancellation pair of tissue harmonic signal [8]. In this study, we focus on the feasibility of 3f

transmit phasing to suppress tissue clutter for improved harmonic detection of low-velocity flow of contrast microbubbles. High-frequency power Doppler (PD) imaging and color flow (CF) imaging were performed in phantom experiments to compare the detectability of contrast flow without and with 3f

transmit phasing.

This work was supported by National Science Council of Taiwan under grant NSC 100-2320-B-011-001

10.1109/ULTSYM.2011.0101

II. C

D

In the swept-scan Doppler system, the tissue clutter signal spectrally broadened and thus is present at higher frequencies than dc as shown by the solid green line in Fig. 1. Due to the spectral overlap between the tissue signal and the flow signal, the cut-off frequency of wall filtering has to increase to be f

so that the wall filtering rejects most of the tissue clutter.

However, it should be noted that the flow signal with Doppler frequency lower than f

will also be removed by wall filtering and thus is not detectable in the final Doppler image. On the contrary, when the clutter suppression is applied to reduce the tissue clutter magnitude as shown by the dashed green line in Fig. 1, the spectral overlap between the flow signal and the tissue clutter is largely alleviated. In this case, the cut-off frequency of wall filtering could decrease back to be f

more low-velocity flow signal remains preserved after wall filtering. This helps to visualize the slow blood flow in Doppler imaging, especially when the detection of microcirculation is of main concern.

To achieve the clutter suppression in harmonic Doppler detection, the method of 3f

transmit phasing is one of the possible candidates. It is based on transmit waveforms comprising fundamental signal and 3f

signal to generate mutual cancellation of tissue harmonic signal. Specifically, the combination of the fundamental transmit signal cos(2 πf

φ

) and the 3f

transmit signal cos(2 π(3f

θ+φ

) can produce two different components of tissue harmonic signal at second harmonic frequency: the frequency-sum component and the frequency-difference component. One produces the harmonic whose frequency is the sum of transmit frequencies. The other results in harmonic at the difference frequency of the transmit signals. Note that the symbol φ

is the phase of the fundamental transmit signal and the symbol θ represents the relative phase between the 3f

transmit signal and the fundamental transmit signal. It has been shown in [9] that, by selecting an optimal θ, the phase of the frequency-sum and the frequency-difference component can be manipulated to be 180 ° out of phase for second harmonic suppression of tissue background. In addition to the 3f

transmit phase, the 3f

transmit amplitude also plays an important role in tissue harmonic suppression. The amplitude of 3f

transmit signal has to be adequately selected such that the generation of the frequency-sum and frequency-difference components are approximately equal in magnitude for maximal cancellation of these two components.

Fig. 1 Illustration of tissue clutter suppression on the cut-off frequency of wall filtering in Doppler estimation.

Fig. 2. Schematic of the experimental setup.

III. M

High-frequency swept-scan imaging system

Fig. 2 shows the schematic diagram of the experimental setup for harmonic Doppler imaging. A 38-MHz transducer with -6-dB fractional bandwidth of 95 % is used for transmit and receiving (custom-designed model, University of Southern California, Los Angeles, CA, U.S.A.). The transducer has a focal length of 12.7 mm and a diameter of 4 mm. A motion stage (model HR8, Nanomotion, Yokneam, Israel) was utilized to move the transducer in the lateral direction for swept-scan imaging. An arbitrary function generator (model 2040, Tektronix, Beaverton, OR, U.S.A.) was used to generate the transmit pulse. The transmit pulse is the combination of the fundamental signal at 16 MHz and the 3f

signal at 48 MHz. A power amplifier (model 325, ENI, Rochester, NY, U.S.A.) in connection with a custom-made expander were utilized to amplify the transmit pulse and to prevent the large noises from leaking into the receiving end. Then, the transmit pulse was coupled to the transducer by a diplexer (model DIP- 3, Matec Instruments NDT, Northborough, MA, U.S.A.). The focal MI at fundamental frequency is limited to below 0.18.

The received echo was sent to a low-noise amplifier for amplification (Ritec BR-640A, Warwick, RI, U.S.A.) and was digitized by a 200-Msamples/s digitizer (model CS12400, Gage Applied Tech Inc., IL, U.S.A.). To select the optimal 3f

transmit amplitude and phase for maximal tissue harmonic suppression, calibration was performed by measuring the tissue harmonic signal in the water tank from a cellulose wire of 200- μm diameter. The experimental 3f

transmit amplitude and phase was determined to be 0.2 and 40 degrees, respectively. A speckle-generating agar phantom with uniform distribution of graphite powder (Sigma-Aldrich, 282863, St.

Louis, MO, U.S.A.) was utilized as the imaged object in Fig.

2. A tubular flow channel of 1-mm diameter was fabricated inside the phantom with a Doppler angle of about 60 degrees.

A syringe pump (model 100, KD Scientific, Holliston, MA, U.S.A.) is used to regulate the flow of contrast microbubbles inside the channel. The axial velocity of the bubble flow was 3 mm/s. In the study, custom-made high-frequency microbubbles were used with a volume concentration of 0.1

%. The microbubbles have a mean diameter of 1.6 μm and a

resonance frequency of about 7 ~ 11 MHz [9].The

transducer’s scanning speed was 1.6 mm/s, and the pulse-

repetition frequency was 800 Hz.

Three-order high-pass Chebyshev type II digital filters were used for wall-filtering the demodulated two-dimensional signals in the slow-time direction. The cut-off frequency of the wall filter was calculated from the scanning speed and the beam width of the transducer. Therefore, the theoretical cut- off frequency of the wall filter is about 16.6 Hz. In this study, wall filtering with this theoretical cut-off frequency is referred to as the full wall filtering. For comparisons, wall filters with a half and a quarter of this theoretical cut-off frequency are also utilized and are referred to as the half wall filtering and the quarter wall filtering, respectively. After wall filtering, the Doppler signals were grouped into packets of 20 ensembles for CF and PD estimations using conventional autocorrelation method [10]. In the final PD image, only the pixels with Doppler power higher than the noise level by a 5-dB threshold were displayed. In the corresponding CF images, the flow velocity estimated to be lower than the cut-off velocity of wall filtering is further removed.

IV. R

Fig. 3 shows the radio-frequency spectra of received echoes from the wire target without and with 3f

transmit phasing. Note that the additional 3f

transmit signal is clearly present in the spectrum with 3f

transmit phasing. The averaged second harmonic magnitude ranging from 27 MHz to 40 MHz is shown to be suppressed by about 9.5 dB with 3f

transmit phasing. Therefore, it is evident that the 3f

transmit phasing can provide effective suppression of tissue clutter magnitude. For the flowing microbubbles, the corresponding Doppler spectra are demonstrated in Fig. 4. It appears that the Doppler spectral distribution with 3f

transmit phasing remains similar to that without 3f

transmit phasing. This indicates that, though the microbubbles have experienced a different driving waveform due to the additional 3f

transmit signal in 3f

transmit phasing, the velocity of bubble flow is not markedly modified. It should be noted that, however, the 3f

transmit phasing does result in a slight decrease of Doppler magnitude from microbubbles which is consistent to that observed in B- mode images .

Fig. 5 shows the swept-scan power images in gray-scale before and after wall filtering with a 30-dB dynamic range.

For both conventional image (i.e., without 3f

transmit phasing) and clutter-suppressed image (i.e., with 3f

transmit phasing), the wall filters with various cut-off frequencies are applied. The “1”, “1/2”, “1/4” and “0” represent the full, half, quarter and no wall filtering, respectively. The panels in Fig. 5 from left to right demonstrate that the wall filtering gradually removes the tissue clutter when the cut-off frequency increases. Compared to the conventional image, the clutter- suppressed image with the same wall filtering generally has a lower background clutter level. Fig. 6 shows the flow signal- to-clutter ratio (SCR) of the images in Fig. 5 as a function of different wall filtering. The SCR is estimated by the ratio of mean power in the flow region and the tissue clutter region, indicated by the boxes in the leftmost panels in Fig. 5. For both conventional image and clutter-suppressed image, the SCR generally increases with the cut-off frequency of wall filtering. Nonetheless, since the high cut-off frequency of wall

filtering also inevitably leads to loss of flow signal, it should be noted that the full wall filtering actually results in a slightly lower SCR compared to the half wall filtering. Moreover, the clutter-suppressed image consistently has a higher SCR than its conventional counterpart. In other words, the 3f

transmit phasing effectively improves the SCR by suppressing the tissue clutter signal. Note that the clutter-suppressed image with quarter wall filtering already has a superior SCR to the conventional image with full wall filtering. This indicates that, for the clutter-suppressed image, the cut-off frequency of wall filtering can be reduced to preserve low-velocity flow without sacrificing the SCR.

The typical CF and PD images without and with clutter suppression are also compared in Fig. 7, respectively. The conventional images are process by the full wall filtering and the clutter-suppressed images are processed by the quarter wall filtering. In Fig. 7(a), it is noticeable that the estimated flow velocity in the CF images gradually increases from the boundary to the center of the flow region, indicating the laminar flow has been established in the channel. For the conventional CF images, residual clutter artifacts remain visible since even the full wall filtering cannot completely remove the clutter signal in the background. For the clutter- suppressed CF image, on the other hand, the clutter artifacts have been mostly eliminated by the quarter wall filtering. Note that the quarter wall filtering can detect more low-velocity flow near the boundary of the channel due to its lower cut-off frequency. Therefore, more boundary pixels have become detectable and the estimated flow region increases from the conventional case to the clutter-suppressed case. The PD images in Fig. 7(b) also show that the clutter artifacts decrease markedly with clutter suppression. In order to quantitatively measure the sensitivity of Doppler detection, the color-pixel- density (CPD) was provided. The CPD is the ratio of estimated flow area in the Doppler image to the actual cross- section area of the flow channel. The flow area is calculated by counting the number of color pixels in the box shown in the left panels of Fig. 7 to eliminate the interference from the residual clutter artifacts. For the clutter-suppressed image, it is shown that the decrease of cut-off frequency of wall filtering provides a higher CPD due to the preservation of low-velocity flow pixels. Note that a higher CPD in this study represents that the Doppler detection is more sensitive so that the estimation of blood flow is more close to the actual blood flow in area.

Fig. 3 Measured radio-frequency echo spectra from the wire phantom.

Fig. 4 Measured Doppler spectra from the flow region of contrast microbubbles.

Fig. 5 B-mode power images without and with clutter suppression for various wall filtering. From left to right, the cut-off frequency of wall filtering is zero (0), quarter (1/4), half (1/2) and full (1), respectively.

Fig. 6 Estimated SCR values for the images in Fig. 5.

Fig. 7 Estimated (a) CF images and (b) PD images without and with clutter suppression. The clutter-suppressed image shows reduced clutter artifacts and a higher CPD value.

V. C

In this study, a dual-frequency transmit waveform is utilized to suppress tissue clutter magnitude for improving the detection of low-velocity flow in harmonic Doppler imaging.

For a swept-scan system, this is of particular importance because the image lines are acquired continuously during the lateral scanning of the transducer and thus the corresponding tissue clutter signal spectrally broadens in the Doppler domain and interferes with the low-velocity flow signal. Consequently, a tradeoff exists between the detectability of low-velocity flow and the achievable scanning speed of the imaging system. To alleviate the interferences from the tissue clutter, the method of 3f

transmit phasing is applied to reduce the magnitude of tissue clutter signal directly in the radio-frequency domain.

Results indicate that, compared to the conventional CF and PD images, the clutter-suppressed counterparts show apparent reduction of clutter artifacts. Clutter suppression also makes the decrease of cut-off frequency of wall filtering possible while the SCR value remains superior. Therefore, the clutter- suppressed Doppler images can detect more low-velocity flow signal at no cost of the imaging frame rate.

R

[1] C. Passmann, H. Ermert, “A 100-MHz ultrasound imaging system for dermatologic and ophthalmologic diagnostics”, IEEE Trans. Ultrason.

Ferroelect. Freq. Control, vol. 43, no. 4, pp. 545-552, 1996.

[2] L. Sun, W. D. Richard, J. M. Cannata, C. C. Feng, J. A. Johnson, J. T.

Yen, K. K. Shung, “A high-frame rate high-frequency ultrasonic system for cardiac imaging in mice”, IEEE Trans. Ultrason.

Ferroelect. Freq. Control, vol. 54, no. 8, pp. 1648-1654, 2007.

[3] D. E. Kruse, R. H. Silverman, R. J. Fornaris, D. J. Coleman, K. W.

Ferrara, “A swept-scanning mode for estimation of blood velocity in the microvasculature”, IEEE Trans. Ultrason. Ferroelect. Freq.

Control, vol. 45, no. 6, pp. 1437-1443, 1998.

[4] A. Needles, D. E. Goertz, A. M. Cheung, F. S. Foster, “Interframe clutter filtering for high frequency flow imaging”, Ultrasound Med.

Biol., vol. 33, no. 4, pp. 591-600, 2007.

[5] A. Bouakaz, M. Versluis, J. Borsboom, N. de Jong, “Radial modulation of microbubbles for ultrasound contrast imaging”, IEEE Trans. Ultrason. Ferroelect. Freq. Control, vol. 54, no. 11, pp. 2283–

2290, 2007.

[6] D. E. Goertz, M. E. Frijlink, D. Tempel, V. Bhagwandas, A. Gisolf, R.

Krams, N. de Jong, A. F. W. van der Steen, “Subharmonic contrast intravascular ultrasound for vasa vasorum imaging”, Ultrasound Med.

Biol., vol. 33, no. 12, pp. 1859–1872, 2007.

[7] S. Krishnan, J. D. Hamilton, M. O’Donnell, “Suppression of propagating second harmonic in ultrasound contrast imaging”, IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 45, no. 3, pp. 704–

711, 1998.

[8] C. C. Shen, Y. C. Wang, Y. C. Hsieh, “Third harmonic transmit phasing for tissue harmonic generation”, IEEE Trans. Ultrason.

Ferroelectr. Freq. Control, vol. 54, no. 7, pp. 1370–1381, 2007.

[9] S. C. Lu, J. A. Ho, C. K. Yeh, “Echogenic liposomes in high-frequency ultrasound imaging”, IEEE Ultrason. Symp., art. no. 4410127, pp.

2203-2206, 2007.

[10] C. Kasai, K. Namekawa, A. Koyano, R. Omoto, “Real-time two- dimensional blood flow imaging using an autocorrelation technique”, IEEE Trans. Sonics Ultrason., pp. 458-464, 1985.

國科會補助計畫衍生研發成果推廣資料表

日期:2012/08/03

國科會補助計畫

計畫名稱: 最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 計畫主持人: 沈哲州

計畫編號: 100-2320-B-011-001- 學門領域: 醫學工程

無研發成果推廣資料

100 年度專題研究計畫研究成果彙整表

計畫主持人：沈哲州 計畫編號：100-2320-B-011-001-

計畫名稱：最佳基頻發射相位法於諧波影像之組織背景抑制與微氣泡對比劑造影 量化

成果項目

實際已達 成數（被接

受或已發 表）

預期總達成 數(含實際 已達成數)

本計畫 實際貢 獻百分

比

單位

備註（質 化 說 明：如 數 個 計 畫 共 同 成 果 、 成 果 列 為 該 期 刊 之 封 面 故 事 ...等）

期刊論文 0 0 100%

研 究 報 告 / 技 術 報

告 0 0 100%

研討會論文 1 1 100%

論文著作 篇

專書 0 0 100%

申請中件數 0 0 100%

專利 已獲得件數 0 0 100% 件

件數 0 0 100% 件

技術移轉

權利金 0 0 100% 千元

碩士生 0 0 100%

博士生 0 0 100%

博士後研究員 0 0 100%

國內

參與計畫人力

（本國籍）

專任助理 0 0 100%

人次

期刊論文 1 2 50%

NSC

99-2320-B-011-002-MY3 共同成果

研 究 報 告 / 技 術 報

告 0 0 100%

研討會論文 1 1 100%

論文著作 篇

專書 0 0 100% 章/本

申請中件數 0 0 100%

專利 已獲得件數 0 0 100% 件

件數 0 0 100% 件

技術移轉

權利金 0 0 100% 千元

碩士生 0 0 100%

博士生 0 0 100%

博士後研究員 0 0 100%

國外

參與計畫人力

（外國籍）

專任助理 0 0 100%

人次

其他成果

(

無

成果項目 量化 名稱或內容性質簡述

測驗工具(含質性與量性) 0

課程/模組 0

電腦及網路系統或工具 0

教材 0

舉辦之活動/競賽 0

研討會/工作坊 0

電子報、網站 0

科 教 處 計 畫 加 填 項

目 計畫成果推廣之參與（閱聽）人數 0

國科會補助專題研究計畫成果報告自評表

請就研究內容與原計畫相符程度、達成預期目標情況、研究成果之學術或應用價 值（簡要敘述成果所代表之意義、價值、影響或進一步發展之可能性） 、是否適 合在學術期刊發表或申請專利、主要發現或其他有關價值等，作一綜合評估。

1. 請就研究內容與原計畫相符程度、達成預期目標情況作一綜合評估

■達成目標

□未達成目標（請說明，以 100 字為限）

□實驗失敗

□因故實驗中斷

□其他原因 說明：

2. 研究成果在學術期刊發表或申請專利等情形：

論文：■已發表 □未發表之文稿 □撰寫中 □無 專利：□已獲得 □申請中 ■無

技轉：□已技轉 □洽談中 ■無 其他：（以 100 字為限）

3. 請依學術成就、技術創新、社會影響等方面，評估研究成果之學術或應用價 值（簡要敘述成果所代表之意義、價值、影響或進一步發展之可能性）（以 500 字為限）

本計劃之研究動機在於探討超音波非線性諧波成像中發射相位對於組織區域與對比劑區 域之間影像對比度的重要性 我們的研究成果顯示基頻信號發射相位的最佳化調整可以達 成組織區域的非線性組織諧波與線性溢漏諧波之間的失相相消，以進一步改善諧波影像的 對比劑偵測品質。我們也進一步顯示這樣的信號發射相位最佳化不一定只限於配備有任意 波形產生功能的高階超音波機才能使用，只要透過二元碼轉換技術也可以讓一般僅具備二 元脈衝產生功能的中低階系統達成相位最佳化以提昇影像品質。