JP3639364B2 - Ultrasonic diagnostic equipment - Google Patents

Description

【数２】

上記（１）式は画像信号を表す座標領域から周波数領域へ、また上記（２）式は周波数領域から座標領域へ変換できることを意味する。
【００２６】
このようにフーリエ変換は、座標領域と周波数領域とを相互に変換できる性質を持つ。本実施の形態では、このフーリエ変換の性質を利用してデータの補間を行っている。
【００２７】
図３は、フーリエ変換により補間する方法を説明した図である。実際に使用する信号は、上記（１）式で表される２次元の信号であるが、ここでは、まず、説明を簡単にするため１次元の信号を用いて説明する。
【００２８】
図３（ａ）は、時間τａでサンプリングした１次元の信号である。同図（ｂ）は、同図（ａ）の信号をフーリエ変換した結果、得られた周波数スペクトルである。同図（ｃ）は同図（ａ）の信号より更に短い時間τｂでサンプリングした信号である。同図（ｄ）は同図（ｃ）の信号をフーリエ変換した結果得られた周波数スペクトルである。
【００２９】
一般に、図３（ａ）の時間τａでサンプリングした信号をフーリエ変換すると、サンプリング時間間隔τａの逆数、すなわち１／τａに等しい周波数帯域Ｆａの周波数スペクトル（同図（ｂ））が得られる。このような、信号のサンプリング時間間隔とフーリエ変換後の周波数スペクトルの周波数帯域との関係（１／τ＝ｆ）から、周波数スペクトルの周波数帯域を変更することで、信号のサンプル間を補間することができる。
【００３０】
例えば、信号のサンプル数を２倍にするには、サンプリング間隔を１／２にすることにより実現できる。このとき、周波数スペクトルの周波数帯域は、上記サンプリング時間と周波数帯域の関係を用いると、１／（τａ／２）＝２／τａ＝２Ｆａ（１／τａ＝Ｆａ）より、２倍となる。これより、周波数スペクトルの周波数帯域を２倍にし、逆フーリエ変換すると、信号のサンプル間が補間され、サンプル数が２倍になった信号が得られる。
【００３１】
上記説明より、図３（ｂ）に示す周波数スペクトルを２倍にするため、同図（ｄ）の周波数スペクトルのＡの部分にゼロ値の周波数スペクトルを追加する。図３（ｄ）に示す周波数帯域を２倍にした周波数スペクトルを逆フーリエ変換すると、図３（ｃ）に示すように、同図（ａ）のサンプル点を補間し、サンプル数が２倍になった信号が得られる。
【００３２】
換言すれば、ゼロ値を周波数スペクトルに追加することにより、図３（ｂ）に示す標本化定理を満足しているサンプリング周波数Ｆａが、更に周波数帯域の広がったサンプリング周波数Ｆｂとなるため、標本化間隔が細かくなり、更に滑らかな補間が可能になるともいえる。
【００３３】
そして、これらの補間した信号のサンプル数を処理前のサンプル数と同じ数だけ使用することにより、拡大表示が実現できる。例えば、図３（ａ）の元の信号と同図（ｃ）の２倍に補間した信号の両方について、Ｓ0からＳ3の４サンプルの区間に注目する。このサンプル区間長は、図３（ａ）ではＳａであり、同図（ｃ）ではＳｂである。両データを比較すると、図３（ｃ）のサンプル区間長Ｓｂの方が同図（ａ）のサンプル区間長Ｓａより短くなっている。このように、サンプル数を同じにして表示用データとすることにより、補間したデータで信号を拡大できる。ただし、表示用データは、必ずしも処理前のサンプル数と同じにする必要はなく、任意の領域のサンプル数で行ってもよい。
【００３４】
以上、簡単のために１次元の離散フーリエ変換を例にとって説明した。以下に実際の２次元の超音波断層像の拡大処理に際して、周波数空間での補間処理につき述べる。
【００３５】
図４に本実施例における２次元超音波断層像の拡大補間処理の概要を示す。図４（ａ）は、画像メモリ４に格納された超音波断層像Ａ（Ｒ）を示している。ここで、Ａは各画素の保持している超音波の反射強度であり、Ｒは画像メモリ用アドレス生成回路６及びデータフローセレクタ７により設定された画像メモリ４中の処理領域を表し、ｘ軸方向にｄｘ間隔で並んだＮｘ個の画素、ｙ軸方向にｄｙ間隔で並んだＮｙ個の画素を有する矩形領域からなる。なお、処理領域Ｒの原点を（ｘ0，ｙ0）とした。これら、（ｘ0，ｙ0），Ｎｘ，Ｎｙを設定すると、画像メモリ４の超音波断層像の任意の部分を拡大処理することができる。また、処理領域Ｒは、次式で記述できる。
【００３６】
【数３】
Ｒ＝｛（ｘ，ｙ）；ｘ＝ｘ0＋ｍｄｘ，ｙ＝ｙ0＋ｎｄｙ，ｘ0，ｙ0は前記原画像メモリの任意の画素アドレス、０≦ｍ＜Ｎｘ，０≦ｎ＜Ｎｙ，ｍ，ｎは整数、Ｎｘ，Ｎｙは自然数｝ …（３）
こうして画像メモリ４から抽出された処理領域Ｒ上の超音波断層像Ａ（Ｒ）を、入力バッファメモリ２１に蓄積し、２次元離散フーリエ変換の中で最もよく用いられる２次元ＦＦＴ回路２２でフーリエ変換して得られる原スペクトルＳを図４（ｂ）に示す。２次元ＦＦＴ回路２２で得られる原スペクトルＳの定義される周波数領域Ｕは、処理領域Ｒの画素数Ｎｘ×Ｎｙと同じ数の離散的な要素からなり、次式で与えられる。
【００３７】
【数４】
Ｕ＝｛（ｆｘ，ｆｙ）；ｆｘ，ｆｙはそれぞれｘ軸，ｙ軸についての空間周波数で、ｆｘ＝−（Ｎｘ／２−ｉ）／（Ｎｘｄｘ），ｆｙ＝−（Ｎｙ／２−ｊ）／（Ｎｙｄｙ），０≦ｉ＜Ｎｘ，０≦ｊ＜Ｎｙ，ｉ，ｊは整数｝ …（４）
上記の（４）式から明らかなように、Ｕは−１／２ｄｘ≦ｆｘ＜＋１／２ｄｘ，−１／２ｄｙ≦ｆｙ＜＋１／２ｄｙの矩形範囲となる。
【００３８】
次に、原スペクトルＳは、次式のように周波数領域Ｕ上で定義されたスペクトルｓ（ｆｘ，ｆｙ）の集合である。
【００３９】
【数５】
Ｓ＝｛ｓ（ｆｘ，ｆｙ）；（ｆｘ，ｆｙ）∈Ｕ｝ …（５）
そして、（５）式で与えられる原スペクトルＳはゼロ値挿入回路２３に入力される。ゼロ値入力回路２３は、始めに拡大率設定器１２にて設定された拡大率Ｍを受けて上記周波数領域Ｕを含み、Ｕが縦横共にＭ倍に拡大された拡張周波数領域Ｖを次式のように設定する。
【００４０】
【数６】
Ｖ＝｛（ｆｘ，ｆｙ）；ｆｘ＝−Ｍ（Ｎｘ／２−ｉ）／（Ｎｘｄｘ），ｆｙ＝−Ｍ（Ｎｙ／２−ｊ）／（Ｎｙｄｙ），０≦ｉ＜ＭＮｘ，０≦ｊ＜ＭＮｙ，ｉ，ｊは整数｝ …（６）
次に、拡張周波数領域Ｖ中、周波数領域Ｕの部分には原スペクトルデータｓ（ｆｘ，ｆｙ）をそのまま用い、差領域Ｖ−Ｕの部分にはゼロ値を挿入する。こうして得られたスペクトルを拡張スペクトルＴと呼ぶことにする。拡張スペクトルＴの概略形状を図４（ｃ）に示す。Ｔの各要素をｔ（ｆｘ，ｆｙ）と書けば、拡張スペクトルＴを、
【数７】
Ｔ＝｛ｔ（ｆｘ，ｆｙ）；ｔ（ｆｘ，ｆｙ）＝ｓ（ｆｘ，ｆｙ）（（ｆｘ，ｆｙ）∈Ｕのとき），ｔ（ｆｘ，ｆｙ）＝０（（ｆｘ，ｆｙ）∈Ｖ−Ｕのとき）｝ …（７）
と表すことができる。
【００４１】
こうして得られた拡張スペクトルＴは、２次元逆ＦＦＴ回路２４に送られ実空間のデータに変換される。
【００４２】
ここで、上に述べた原理により拡大及び補間処理が行われる。拡張スペクトルＴはＭＮｘ×ＭＮｙ個のデータからなるので、出力バッファメモリ２５には拡大率設定器１２からの拡大率Ｍに応じて処理領域Ｒを縦横共にＭ倍にしたデータの格納領域である拡大処理領域Ｑが準備される。
【００４３】
そして、上記２次元逆ＦＦＴ回路２４により逆フーリエ変換された補間拡大像Ａ（Ｑ）が、上記出力バッファ２５に送られる。補間拡大像の概略形状を図４（ｄ）に示す。
【００４４】
得られた表示用データは、表示モード切り換え器１３により拡大画像が表示されるようコントロール回路５で設定されたビデオミキサ９を通して、ＤＳＣ１０に転送され、モニタ１１に拡大されたＢモード像が表示される。
【００４５】
（効果）
以上のように、本実施の形態の超音波診断装置１では、拡大処理回路８において、ゼロ値挿入回路２３が２次元ＦＦＴ回路２２により得られた画像データの周波数スペクトルｓにゼロ値を挿入し、周波数帯域を広げ、この周波数スペクトルを２次元逆ＦＦＴ回路２４で逆フーリエ変換することにより、サンプリング間隔の細かい画像データが出力され、モニタ１１でのＢモード像の表示時に、このサンプル数を処理前のサンプル数と同じにすることで、滑らかに補間され、拡大しても良好なＢモード像を得ることができる。
【００４６】
図５及び図６は本発明の第２の実施の形態に係わり、図５は拡大処理回路の構成を示すブロック図、図６は図５の線形補間回路の作用を説明する説明図である。
【００４７】
第２の実施の形態は、第１の実施の形態とほとんど同じであるので、異なる点のみ説明し、同一の構成には同じ符号をつけ説明は省略する。
【００４８】
（構成）
第２の実施の形態は、第１の実施の形態の拡大処理回路８に線形補間拡大回路を設け、線形補間拡大回路とフーリエ変換による補間拡大回路とを用途によって切り換えられるようにした点が、第１の実施の形態と大きく異なる。
【００４９】
このため、本実施の形態の拡大処理回路８ａは、図５に示すように、第１の実施の形態で説明したフーリエ変換によるフーリエ変換補間拡大回路３０を構成する２次元ＦＦＴ回路２２、ゼロ挿入回路２３及び２次元逆ＦＦＴ回路２４の他に、後述する線形補間拡大処理を行う線形補間拡大処理回路３１と、入力バッファメモリ２１に記憶された画像データをコントロール回路５の制御により２次元ＦＦＴ回路２２あるいは線形補間拡大処理回路３１に選択的に出力する拡大処理セレクタ３２と、２次元逆ＦＦＴ回路２４の出力あるいは線形補間拡大処理回路３１の出力をコントロール回路５の制御により選択的に出力バッファメモリ２５に出力する出力データセレクタ３３とを備えて構成される。なお、コントロール回路５には、線形補間拡大回路３１あるいはフーリエ変換補間拡大回路３０の選択を設定する拡大処理設定器３４が接続されている。
【００５０】
その他の構成は第１の実施の形態と同じである。
【００５１】
（作用）
本実施の形態においては、画像メモリ４に各空間での反射体の強度分布を画像化したディジタルデータを格納し、画像メモリ用アドレス生成回路６で画像メモリ４に格納されている画像データを読み出し、データフローセレクタ７で選択された拡大処理回路８ａに画像データを転送するまでの過程は、第１実施の形態と同じである。
【００５２】
拡大処理回路８ａにおいて、転送された画像データは、入力バッファメモリ２１に格納される。そして、予め拡大処理設定器３４で設定した線形補間処理、フーリエ変換による補間拡大処理のどちらかを拡大処理セレクタ３２で選択し、画像データを転送し処理を行う。
【００５３】
拡大処理設定器３４で線形補間拡大処理を選択した場合、画像データは拡大処理セレクタ３２で選択された線形補間拡大処理回路３１に転送され、ここで拡大処理される。処理した結果は、出力データセレクタ３３を通り、出力バッファメモリ２５に格納される。
【００５４】
また、拡大処理設定器３４でフーリエ変換による補間拡大処理を選択した場合、画像データは拡大処理セレクタ３２で選択されたフーリエ変換補間拡大回路３０に転送され、フーリエ変換補間拡大回路３０における２次元ＦＦＴ回路２２、ゼロ挿入回路２３及び２次元逆ＦＦＴ回路２４によってフーリエ変換による拡大処理が行われる。処理した結果は出力データセレクタ３３を通り、出力バッファメモリ２５に格納される。
【００５５】
これらの処理により、出力バッファメモリ２５に格納された拡大画像データは、第１実施の形態と同様に、ビデオミキサ９に転送され、ＤＳＣ１０によりモニタ１１に表示される。
【００５６】
ここで、線形補間拡大処理回路３１における線形補間拡大処理とフーリエ変換補間拡大回路３０におけるフーリエ変換による補間拡大処理の違いについて説明する。両拡大処理は、補間して増加したデータから補間前のデータ数のみ取り出すことにより、拡大処理を行っており、拡大して得られる画像の違いは補間方法にある。このため、ここでの説明は補間方法の違いに限定する。
【００５７】
さらに、画像データは、図１１に示したように２次元のデータであるが、ここでは、図６に示すように、１次元の信号で説明する。図６は、線形補間とフーリエ変換による補間を示しており、図６（ａ）が１次元の線形補間を、同図（ｂ）がフーリエ変換による補間を示している。図６（ａ）、同図（ｂ）とも曲線Ａは元の信号を、黒い点は曲線Ａをサンプリングして得られたサンプル点を、白ぬきの点は補間処理により得られたサンプル点を示す。
【００５８】
図６（ａ）に示す線形補間の場合、補間して得られたサンプル点Ｓaが元の信号である曲線Ａ上にないが、同図（ｂ）に示すフーリエ変換による補間の場合、補間して得られたサンプル点Ｓbは元の信号の曲線Ａ上に存在する。このように、線形補間の場合、図６（ａ）に示すように求めるサンプル点Ｓaの両側にあるサンプル点ａ1とａ2の２点が画素値の平均をとるため直線的な補間となり、補間に粗さがある。
【００５９】
一方、フーリエ変換による補間の場合、図６（ｂ）に示すように、補間したサンプル点Ｓbが曲線Ａ上に存在するように、サンプル点ｂ1とｂ2の２点間を滑らかに結ぶため、良好な補間が得られる。
【００６０】
したがって、フリーズ後に拡大して良好な画像を得るには、フーリエ変換による補間の方が線形補間に比べ適している。しかし、フーリエ変換による補間処理は時間がかかるため、リアルタイムに処理することはできない。反面、線形補間は、単純に２点間の画素値の平均により補間しているため、リアルタイムに処理することができる。
【００６１】
（効果）
このように、本実施の形態では、第１の実施の効果に加え、リアルタイムに拡大処理する場合、線形補間拡大処理回路３１における線形補間拡大処理を行い、フリーズ後に良好な拡大画像を得たい場合、フーリエ変換補間拡大回路３０におけるフーリエ変換による拡大処理を行うことができ、用途に応じて拡大処理方法を切り換えることができる。
【００６２】
図７及び図８は本発明の第３の実施の形態に係わり、図７は超音波診断装置の構成を示す構成図、図８は図７の超音波診断装置の作用を説明する説明図である。
【００６３】
第３の実施の形態は、第１の実施の形態とほとんど同じであるので、異なる点のみ説明し、同一の構成には同じ符号をつけ説明は省略する。
【００６４】
（構成）
本実施の形態では、第１実施の形態の超音波診断装置１において、画像の拡大領域を設定する機能を追加したものである。
【００６５】
このため、図７に示すように、本実施の形態の超音波診断装置４１は、第１実施の形態の表示モード切り換え操作器１３（図１参照）の代わりに拡大領域設定器４２を設けて構成される。
【００６６】
その他の構成は第１の実施の形態と同じである。
【００６７】
（作用）
本実施の形態においては、画像メモリ４に各空間での反射体の強度分布を画像化したディジタルデータが格納されるまでは、第１実施の形態と同じである。
【００６８】
そして、拡大領域設定器４２で設定した拡大領域により、コントロール回路５を通して、画像メモリ用アドレス生成器６で画像メモリ４から拡大領域に相当する画像データを取り出す。取り出したデータは、第１の実施の形態と同様にデータフローセレクタ７に接続された拡大処理回路８に転送され、フーリエ変換による補間拡大処理が行われ、ビデオミキサ９で拡大前の画像データ上に指定した拡大領域に重ね合わせＤＳＣ１０でモニタ１１に表示する。
【００６９】
得られた拡大処理結果の表示を図８に示す。図８（ａ）が拡大前の表示を示し、図８（ｂ）は同図（ａ）のＦの部分を拡大し、同図（ａ）の画像と重ねた結果を示しており、同図（ｂ）本実施の形態で得られる結果となる。ここで、図８（ｂ）に表示されている拡大表示領域Ｇは、拡大前と同じデータ数を表示領域としているため、実際には同図（ａ）のαの範囲のみ表示されている。
【００７０】
（効果）
このように、本実施の形態では、第１の実施の効果に加え、得られた良好な拡大画像を拡大前の画像上に指定した領域に表示することができるので、拡大している部位を容易に確認することができる。
【００７１】
なお、上記説明において、拡大領域の指定は、拡大前の画像の表示上にマウスカーソルを表示し、このカーソルをマウスで操作して指定してもよい。また、指定した拡大領域をマウスで移動させてもよい。このように、拡大前の画像上に指定した拡大領域をマウスで移動させて表示させることにより、超音波画像を虫眼鏡で見るような拡大表示を実現する。
【００７２】
また、上記説明で表示用データは処理前のサンプル数と同じにしていたが、必ずしもこれに限定される必要はなく、任意のサンプル数を表示用データとしてもよい。
【００７３】
さらに、上記拡大処理回路８の代わりに第２実施の形態で説明した拡大処理回路８ａを用い、線形補間拡大処理とフーリエ補間拡大処理とを切り換えて、リアルタイムに拡大前の画像データと拡大後の画像データの重ね合わせ表示を実現してもよい。
【００７４】
図９及び図１０は本発明の第４の実施の形態に係わり、図９は超音波診断装置の構成を示す構成図、図１０は図９のＤＳＰ演算部による拡大処理の流れを示すフローチャートである。
【００７５】
第４の実施の形態は、第１の実施の形態とほとんど同じであるので、異なる点のみ説明し、同一の構成には同じ符号をつけ説明は省略する。
【００７６】
（構成）
第１の実施の形態の超音波診断装置１が画像の拡大処理をハードウェアで実現しているのに対し、本実施の形態では、高速信号処理プロセッサ（ＤＳＰ）に拡大処理アルゴリズムをプログラミングし、ソフトウェアで実現していることが大きく異なる。
【００７７】
このため、図９に示すように、本実施の形態の超音波診断装置５１では、第１の実施の形態の拡大処理回路８をＤＳＰ演算部５２に置き代えて構成している。ＤＳＰ演算部５２の内部構成は、入力データバッファ２１、出力バッファメモリ用アドレス生成回路２６及び出力データバッファ２５の他に、ソフトウェアで拡大処理を行うＤＳＰ５３と、拡大処理アルゴリズムを格納しているプログラムメモリ５４とから構成される。
【００７８】
その他の構成は第１の実施の形態と同じである。
【００７９】
（作用）
本実施の形態においては、画像メモリ４に各空間での反射体の強度分布を画像化したディジタルデータが格納されるまでは、第１実施の形態と同じである。
【００８０】
そして、この画像メモリ４に格納されているディジタルデータをＤＳＰ演算部５２内の入力バッファメモリ２１に転送する。
【００８１】
ＤＳＰ演算部５２では、予めプログラムメモリ５４からＤＳＰ５３にプログラムをロードする。ＤＳＰ５３では、ロードしたプログラムを実行し入力バッファメモリ２１のデータの画像の拡大処理を行う。
【００８２】
ここで、ＤＳＰ演算部５２の処理手順を、図１０に示すフローチャートに沿って説明する。
【００８３】
ＤＳＰ演算部５２ではプログラムを実行して、まずステップＳ１でＤＳＰ５３を初期化し、ステップＳ２で入力バッファメモリ２１からデータをＤＳＰ５３に読み込む。次のステップＳ３からステップＳ５までは、フーリエ変換を用いた補間拡大処理の部分である。
【００８４】
ステップＳ３では、２次元ＦＦＴ処理を行い、周波数スペクトルを求める。このステップを１次元のデータで考えると、第１の実施の形態で説明した図３（ａ）の信号から図３（ｂ）の周波数スペクトルを求める処理に相当する。
【００８５】
そして、ステップＳ４では、周波数スペクトルにゼロ値のスペクトル追加し、周波数帯域を広げる処理を行う。このステップを１次元のデータで考えると、第１の実施の形態で説明した図３（ｂ）の周波数スペクトルに対して、図３（ｄ）の周波数スペクトルのＡの部分にゼロ値を挿入し、周波数スペクトルの周波数帯域を広げる処理に相当する。
【００８６】
次のステップＳ５では、周波数帯域の広がった周波数スペクトルを逆フーリエ変換して、サンプリング間隔を細かくした信号を求める。このステップを１次元のデータで考えると、第１の実施の形態で説明した図３（ｄ）に示す周波数帯域の広がった周波数スペクトルを逆フーリエ変換して、図３（ｃ）に示す信号に変換する処理に相当する。
【００８７】
そして、ステップＳ６では、補間されたデータを出力バッファメモリ２５に処理前のサンプル数のみ格納する。このステップも、第１の実施の形態で説明した図３に示す１次元のデータで考えると、図３（ｃ）に示すデータ区間Ｓｂを転送する処理に相当する。
【００８８】
そして、拡大された画像データは、第１の実施の形態と同様に、表示モード切り換え操作器１３からビデオミキサ９で拡大表示モードに切り換え、ＤＳＣ１０へ画像データを転送し、モニタ１１に拡大されたＢモード像を表示する。
【００８９】
（効果）
このように、本実施の形態では、第１の実施の効果に加え、拡大処理アルゴリズムをプログラム化し、拡大処理回路自体をＤＳＰ演算部５２に置き換えることによりアルゴリズムの変更が容易となり、仕様に応じた柔軟な対応が可能となる。
【００９０】
なお、本実施の形態における説明では、拡大処理用のプロセッサをＤＳＰとしているが、これはマイクロコンピュータあるいはＣＰＵでもよい。
【００９１】
［付記］
（付記項１） 超音波断層像を直交するｘ軸及びｙ軸に沿ってそれぞれ一定間隔ｄｘ及びｄｙ毎に２次元的に配列された画素の集合として格納する原画像メモリと、
前記超音波断層像の処理領域Ｒ＝｛（ｘ，ｙ）；ｘ＝ｘ0＋ｍｄｘ，ｙ＝ｙ0＋ｎｄｙ，ｘ0，ｙ0は前記原画像メモリの任意の画素アドレス、０≦ｍ＜Ｎｘ，０≦ｎ＜Ｎｙ，ｍ，ｎは整数、Ｎｘ，Ｎｙは自然数｝を設定する空間原領域設定手段と、
前記処理領域Ｒに含まれるＮｘ×Ｎｙ個の画素データを周波数空間上に２次元離散フーリエ変換して、周波数領域Ｕ＝｛（ｆｘ，ｆｙ）；ｆｘ，ｆｙはそれぞれｘ軸、ｙ軸についての空間周波数で、ｆｘ＝−（Ｎｘ／２−ｉ）／（Ｎｘｄｘ），ｆｙ＝−（Ｎｙ／２−ｊ）／（Ｎｙｄｙ），０≦ｉ＜Ｎｘ，０≦ｊ＜Ｎｙ，ｉ，ｊは整数｝上のＮｘ×Ｎｙ個のスペクトルデータｓ（ｆｘ，ｆｙ）からなる原スペクトルＳ＝｛ｓ（ｆｘ，ｆｙ）；（ｆｘ，ｆｙ）∈Ｕ｝を求めるフーリエ変換手段と、
前記空間原領域設定手段にて設定される前記処理領域Ｒ上の前記超音波断層像を拡大処理するための拡大率Ｍを入力する拡大率入力手段と、
前記周波数領域Ｕを含む拡張周波数領域Ｖ＝｛（ｆｘ，ｆｙ）；ｆｘ＝−Ｍ（Ｎｘ／２−ｉ）／（Ｎｘｄｘ），ｆｙ＝−Ｍ（Ｎｙ／２−ｊ）／（Ｎｙｄｙ），０≦ｉ＜ＭＮｘ，０≦ｊ＜ＭＮｙ，ｉ，ｊは整数｝を設定し、差領域であるＶ−Ｕ上の全ての点に値０を挿入して、前記原スペクトルＳから拡張スペクトルＴ＝｛ｔ（ｆｘ，ｆｙ）；ｔ（ｆｘ，ｆｙ）＝ｓ（ｆｘ，ｆｙ）（（ｆｘ，ｆｙ）∈Ｕのとき）、ｔ（ｆｘ，ｆｙ）＝０（（ｆｘ，ｆｙ）∈Ｖ−Ｕのとき）｝を生成するゼロ値挿入手段と、
前記超音波断層像の前記処理領域ＲをＭ倍に拡大した拡大処理領域Ｑを設定する空間拡大領域設定手段と、
前記ゼロ値挿入手段にて生成した前記拡張スペクトルＴを２次元離散逆フーリエ変換して前記拡大処理領域Ｑに格納する逆フーリエ変換手段と
を備え、
前記フーリエ変換手段が前記超音波断層像を前記周波数空間上に２次元離散フーリエ変換し前記原スペクトルＳを算出し、前記ゼロ値挿入手段が、前記フーリエ変換手段が算出した前記原スペクトルＳにより前記拡張スペクトルＴを生成することで拡張補間し、
空間拡大領域設定手段が前記超音波断層像の前記処理領域ＲをＭ倍に拡大した前記拡大処理領域Ｑを設定し、前記逆フーリエ変換手段が前記ゼロ値挿入手段にて生成した前記拡張スペクトルＴを２次元離散逆フーリエ変換して前記拡大処理領域Ｑに格納することで拡大処理する
ことを特徴とする超音波診断装置。
【００９２】
（付記項２） 付記項１に記載の超音波診断装置において、
拡大された画素間を線形補間する線形補間手段と、
前記周波数空間上の拡張補間と前記線形補間を選択する選択手段と、
を有することを特徴とする超音波診断装置。
【００９３】
付記項２では、フーリエ変換による補間拡大処理と線形補間による拡大処理とを用途に応じて切り換え、リアルタイムの拡大処理とフリーズ後の良好な拡大処理の結果が得られる超音波診断装置を提供することを目的としている。
【００９４】
付記項２の超音波診断装置によれば、線形補間による拡大処理とフーリエ変換による拡大処理の２種類を切り換えて使用することにより、リアルタイムの拡大処理と、画像のフリーズ後に良好な拡大画像を必要とする拡大処理の両方を実現でき、用途に応じて選択することができる。
【００９５】
（付記項３） 付記項１に記載の超音波診断装置において、
前記拡大処理領域Ｑを前記原画像メモリに格納されている超音波断層像の領域Ｒ上に設定する
ことを特徴とする超音波診断装置。
【００９６】
付記項３では、フーリエ変換による補間拡大処理より得られた拡大画像を、拡大前の画像上に指定した領域に重ね合わせて表示する超音波診断装置を提供することを目的としている。
【００９７】
付記項３の超音波診断装置によれば、得られた良好な拡大画像を拡大前の画像上に指定した領域に表示することにより、拡大している部位を容易に確認することができる。
【００９８】
（付記項４） 超音波断層像を直交するｘ軸及びｙ軸に沿ってそれぞれ一定間隔ｄｘ及びｄｙ毎に２次元的に配列された画素の集合として格納する原画像メモリと、
前記超音波断層像の処理領域Ｒ＝｛（ｘ，ｙ）；ｘ＝ｘ0＋ｍｄｘ，ｙ＝ｙ0＋ｎｄｙ，ｘ0，ｙ0は前記原画像メモリの任意の画素アドレス、０≦ｍ＜Ｎｘ，０≦ｎ＜Ｎｙ，ｍ，ｎは整数、Ｎｘ，Ｎｙは自然数｝を設定する空間原領域設定手段と、
前記空間原領域設定手段にて設定される前記処理領域Ｒ上の超音波断層像を拡大処理するための拡大率Ｍを入力する拡大率入力手段と、
前記超音波断層像の処理領域ＲをＭ倍に拡大した拡大処理領域Ｑを設定する空間拡大領域設定手段と、
前記超音波断層像を周波数空間上で拡張補間する信号処理プロセッサと
を備え、
信号処理プロセッサは、
前記処理領域に含まれるＮｘ×Ｎｙ個の画素データを２次元離散フーリエ変換して、周波数領域Ｕ＝｛（ｆｘ，ｆｙ）；ｆｘ，ｆｙはそれぞれｘ軸、ｙ軸についての空間周波数で、ｆｘ＝−（Ｎｘ／２−ｉ）／（Ｎｘｄｘ），ｆｙ＝−（Ｎｙ／２−ｊ）／（Ｎｙｄｙ），０≦ｉ＜Ｎｘ，０≦ｊ＜Ｎｙ，ｉ，ｊは整数｝上のＮｘ×Ｎｙ個のスペクトルデータｓ（ｆｘ，ｆｙ）からなる原スペクトルＳ＝｛ｓ（ｆｘ，ｆｙ）；（ｆｘ，ｆｙ）∈Ｕ｝を求めるフーリエ変換動作と、
前記周波数領域Ｕを含む拡張周波数領域Ｖ＝｛（ｆｘ，ｆｙ）；ｆｘ＝−Ｍ（Ｎｘ／２−ｉ）／（Ｎｘｄｘ），ｆｙ＝−Ｍ（Ｎｙ／２−ｊ）／（Ｎｙｄｙ），０≦ｉ＜ＭＮｘ，０≦ｊ＜ＭＮｙ，ｉ，ｊは整数｝を設定し、差領域であるＶ−Ｕ上の全ての点に値０を挿入して、拡張スペクトルＴ＝｛ｔ（ｆｘ，ｆｙ）；ｔ（ｆｘ，ｆｙ）＝ｓ（ｆｘ，ｆｙ）（（ｆｘ，ｆｙ）∈Ｕのとき）、ｔ（ｆｘ，ｆｙ）＝０（（ｆｘ，ｆｙ）∈Ｖ−Ｕのとき）｝を生成するゼロ値挿入動作と、
前記ゼロ値挿入手段にて生成した拡張スペクトルＴを２次元離散逆フーリエ変換して前記拡大処理領域Ｑに格納する逆フーリエ変換動作と
を行う
ことを特徴とする超音波診断装置。
【００９９】
付記項４では、拡大処理部に、ＤＳＰなどの高速演算処理プロセッサを用いることにより、プログラムの変更でアルゴリズムの変更を実現でき、仕様変更に柔軟に対応できる超音波診断装置を提供することを目的としている。
【０１００】
付記項４の超音波診断装置によれば、拡大処理アルゴリズムをプログラム化し、拡大処理回路自体をＤＳＰに置き換えることにより、アルゴリズムを容易に変更できるため、仕様に応じた柔軟な対応が可能となる。
【０１０１】
【発明の効果】
以上説明したように本発明の超音波診断装置によれば、超音波断層像の拡大処理において、超音波断層像を滑らかに補間し、拡大時において良好なＢモード像を得ることができると共に、線形補間による拡大処理とフーリエ変換による補間拡大処理との２種類を切り換えて使用することにより、リアルタイムの拡大処理と、画像のフリーズ後に良好な拡大画像を必要とする拡大処理の両方を実現でき、用途に応じて選択することができ、また、得られた良好な拡大画像を、拡大前の画像上に指定した領域に表示することにより、拡大している部位を容易に確認することができるという効果がある。
【図面の簡単な説明】
【図１】本発明の第１の実施の形態に係る超音波診断装置の構成を示す構成図
【図２】図１の拡大処理回路の構成を示すブロック図
【図３】図２の拡大処理回路の作用を説明する第１の説明図
【図４】図２の拡大処理回路の作用を説明する第２の説明図
【図５】本発明の第２の実施の形態に係る拡大処理回路の構成を示すブロック図
【図６】図５の線形補間回路の作用を説明する説明図
【図７】本発明の第３の実施の形態に係る超音波診断装置の構成を示す構成図
【図８】図７の超音波診断装置の作用を説明する説明図
【図９】本発明の第４の実施の形態に係る超音波診断装置の構成を示す構成図
【図１０】図９のＤＳＰ演算部による拡大処理の流れを示すフローチャート
【図１１】従来の生体断層像の拡大方法を説明する説明図
【図１２】図１１の拡大方法により得られた拡大画像の一表示例を示す図
【符号の説明】
１、４１、５１…超音波診断装置
２…超音波プローブ
３…Ａ／Ｄ変換器
４…画像メモリ
５…コントロール回路
６…画像メモリ用アドレス回路
７…データフローセレクタ
８、８ａ…拡大処理回路
９…ビデオミキサ
１０…ＤＳＣ
１１…モニタ
１２…拡大率設定
１３…表示モード切り換え制御器
２１…入力バッファメモリ
２２…２次元ＦＦＴ回路
２３…ゼロ挿入回路
２４…２次元逆ＦＦＴ回路
２５…出力バッファメモリ
２６…出力バッファメモリ用アドレス生成回路
３０…フーリエ変換補間拡大回路
３１…線形補間拡大処理回路
３２…拡大処理セレクタ
３３…出力データセレクタ
３４…線形補間拡大回路
４２…拡大領域設定器
５２…ＤＳＰ演算部
５３…ＤＳＰ
５４…プログラムメモリ[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an ultrasonic diagnostic apparatus, and more particularly to an ultrasonic diagnostic apparatus that is characterized by a magnification process of a biological tomographic image.
[0002]
[Prior art]
There is an ultrasonic diagnostic apparatus that emits an ultrasonic pulse into a living body, receives a reflected wave from tissue in the living body, and processes the received signal to obtain a tomographic image of the living body. In recent years, functions have been added so that the ultrasonic diagnostic apparatus can cope with various diagnoses.
[0003]
In particular, the enlarged display of biological tomograms is effective for detecting minute abnormal sites in a subject and properly capturing minute movements of organs, so it is added as a function and is often used for diagnosis. It has become.
[0004]
FIG. 11 shows a conventional enlargement method in terms of pixel arrangement. FIG. 11A shows data before enlargement, and FIGS. 11B and 11C show data after enlargement. Here, the color coding of the samples in FIGS. 11A to 11C represents the following state. Black indicates that the sample has a pixel value, white indicates that there is no sample pixel value, and the black and white half indicates that the pixel value is half that of the black sample.
[0005]
One conventional enlargement method is a method in which the same pixels are simply arranged side by side. For example, when the magnification is doubled by this method, the sample P11 in FIG. 11A is arranged side by side, below, and obliquely below as shown in FIG. 11B. Do this for all samples and increase the number of samples. Then, only the portion of the range A having the same number of samples as in FIG. 11A is displayed and enlarged.
[0006]
As another method, there is a method of enlarging by linear interpolation between samples. For example, in the case of enlarging to 2 times by this method, first, an average of both pixel values is obtained and linearly interpolated between the sample P11 and the sample P12 in FIG. Obtain sample SP12. Then, the sample P11 and the sample P21 are interpolated in the same manner to obtain the sample SP21 in FIG. The above interpolation processing is performed between all samples, and the number of samples is increased as shown in FIG. Only the portion of the range B having the same number of samples as that in FIG. 11A is displayed and enlarged.
[0007]
On the other hand, in the conventional ultrasonic diagnostic apparatus, as described above, the method of simply expanding the same pixels shown in FIG. 11B and the method of expanding by linear interpolation between samples shown in FIG. Using either of these, real-time enlargement processing is realized.
[0008]
Here, FIG. 12 shows an enlargement process result of the conventional ultrasonic diagnostic apparatus. FIG. 12A shows an image before enlargement, and FIG. 12B shows an enlarged image. As shown in FIG. 12B, the range surrounded by C in FIG. 12A is enlarged, and the result is displayed on the entire screen.
[0009]
[Problems to be solved by the invention]
However, when the same enlargement display is performed by simply arranging the same pixels as shown in FIG. 11B by the above-described conventional enlargement method, the pixels themselves have become larger because the same pixel values are arranged. There is a problem that the enlarged image appears as a mosaic image.
[0010]
In addition, when the enlarged display is performed by interpolating between the pixels as shown in FIG. 11C, the result of linear interpolation with the average of the two pixel values is used as the pixel value of the sample. There is also.
[0011]
The present invention has been made in view of the above circumstances, and an ultrasonic diagnosis capable of smoothly interpolating an ultrasonic tomographic image and obtaining a good B-mode image at the time of enlargement in the enlargement processing of the ultrasonic tomographic image. The object is to provide a device.
[0012]
[Means for Solving the Problems]
The ultrasonic diagnostic apparatus of the present invention stores an ultrasonic tomographic image as a set of pixels two-dimensionally arranged at regular intervals dx and dy along the orthogonal x-axis and y-axis, respectively. The ultrasonic tomographic image processing region R = {(x, y); x = x 0 + mdx, y = y 0 + ndy, x 0, y 0 are arbitrary pixel addresses of the original image memory, 0 ≦ m <Nx, 0 ≦ n <Ny , M, n are integers, Nx, Ny are natural numbers}, and Nx × Ny pixel data included in the processing region R is subjected to two-dimensional discrete Fourier transform on the frequency space, Frequency region U = {(fx, fy); fx and fy are spatial frequencies about the x-axis and y-axis, respectively, fx = − (Nx / 2−i) / (Nxdx), fy = − (Ny / 2− j) / (Nydy), 0 ≦ i <Nx, 0 ≦ j <Ny, i Fourier transform means for obtaining an original spectrum S = {s (fx, fy); (fx, fy) εU} consisting of Nx × Ny pieces of spectral data s (fx, fy) on an integer}, and the space An enlargement factor input means for inputting an enlargement factor M for enlarging the ultrasonic tomogram on the processing region R set by the original region setting means, and an extended frequency region V = {including the frequency region U (Fx, fy); fx = −M (Nx / 2−i) / (Nxdx), fy = −M (Ny / 2−j) / (Nydy), 0 ≦ i <MNx, 0 ≦ j <MNy, i, j is an integer}, and the value 0 is inserted into all points on VU which is the difference region, and the extended spectrum T = {t (fx, fy); t (fx , Fy) = s (fx, fy) (when (fx, fy) εU), t (fx, fy) = ((Fx, fy) when ∈ V-U) and the zero value insertion means for generating a}, the processing region R of the ultrasound tomographic image to set the enlargement processing area Q which expanded M-fold At the same time, the enlargement processing region Q is set on the region R of the ultrasonic tomographic image stored in the original image memory. Spatial expansion region setting means, inverse Fourier transform means for storing the expanded spectrum T generated by the zero value insertion means in the expansion processing region Q by performing two-dimensional discrete inverse Fourier transform, and linear expansion between the expanded pixels Linear interpolation means for interpolation; and By Fourier transform means and zero value insertion means An extended interpolation on the frequency space and a selection means for selecting the linear interpolation are provided.
[0013]
In the ultrasonic diagnostic apparatus of the present invention, the Fourier transform unit calculates the original spectrum S by performing a two-dimensional discrete Fourier transform on the ultrasonic tomogram on the frequency space, and the zero value insertion unit is the Fourier transform unit. The extended spectrum T is generated from the calculated original spectrum S and subjected to extended interpolation, and the space enlargement region setting means sets an enlargement processing region Q obtained by enlarging the processing region R of the ultrasonic tomographic image M times. In addition, the enlargement processing region Q is set on the region R of the ultrasonic tomographic image stored in the original image memory, Ultrasound tomographic image enlarging processing is performed by the inverse Fourier transforming means performing two-dimensional discrete inverse Fourier transform on the extended spectrum T generated by the zero value inserting means and storing it in the enlarging processing region Q for enlarging processing. , The ultrasonic tomographic image is smoothly interpolated, and a good B-mode image can be obtained at the time of enlargement.
Further, the selection means switches between interpolation enlargement processing by Fourier transform and enlargement processing by linear interpolation according to the application, and obtains a result of real-time enlargement processing and good enlargement processing after freezing. In addition, the enlarged image obtained by the interpolation enlargement process by Fourier transform is displayed superimposed on the designated area on the image before enlargement. .
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0015]
1 to 4 relate to a first embodiment of the present invention, FIG. 1 is a block diagram showing a configuration of an ultrasonic diagnostic apparatus, FIG. 2 is a block diagram showing a configuration of an enlargement processing circuit of FIG. FIG. 4 is a first explanatory diagram for explaining the operation of the enlargement processing circuit of FIG. 2, and FIG. 4 is a second explanatory diagram for explaining the operation of the enlargement processing circuit of FIG.
[0016]
(Constitution)
As shown in FIG. 1, the ultrasonic diagnostic apparatus 1 according to the present embodiment inputs an analog ultrasonic signal from an ultrasonic probe 2 that emits an ultrasonic pulse into the living body and receives a reflected wave from the tissue in the living body. An A / D converter 3 that converts the digital signal into a digital signal, and an image memory 4 that stores an ultrasonic signal converted into a digital signal by the A / D converter 3 as image data of an ultrasonic tomographic image, The image memory 4 stores the image data based on the address from the image memory address circuit 6 controlled by the control circuit 5.
[0017]
The image data stored in the image memory 4 is read based on the address from the image memory address circuit 6 and selectively enlarged by the data flow selector 7 controlled by the control circuit 5. 9 is output.
[0018]
The video mixer 9 is controlled by the control circuit 5 and generates display data for an image in a predetermined display mode from the image data via the data flow selector 7 and processed image data subjected to enlargement processing described later by the enlargement processing circuit 8. And output to a DSC (digital scan converter) 10. The DSC 10 converts the display data generated by the video mixer 9 into, for example, a standard television signal, and displays the processed image on the monitor 11.
[0019]
The control circuit 5 is connected with an enlargement ratio setting device 12 and a display mode switching controller 13. The enlargement ratio setting device 12 sets an enlargement ratio at the time of enlargement processing in the enlargement processing circuit 8, and switches the display mode. The controller 13 designates a display mode (for example, B mode) in the video mixer 9.
[0020]
As shown in FIG. 2, the enlargement processing circuit 8 includes an input buffer memory 21 that temporarily stores image data via the data flow selector 7, and a two-dimensional fast Fourier transform for the image data stored in the input buffer memory 21. A two-dimensional FFT circuit 22 that performs conversion, a zero insertion circuit 23 that performs zero insertion processing to be described later on the data that has been two-dimensional fast Fourier transformed by the two-dimensional FFT circuit 22, and zero insertion processing that is performed by the zero insertion circuit 23 The two-dimensional inverse FFT circuit 24 that performs two-dimensional inverse fast Fourier transform on the transmitted data, and the processed image data that has been subjected to a series of enlargement processing by two-dimensional inverse fast Fourier transform by the two-dimensional inverse FFT circuit 24 are temporarily stored. And an output buffer memory 25 for outputting to the video mixer 9. The output buffer memory 25 is controlled by the control circuit 5. It is adapted to perform storing and reading of the processing image data based on the address from the output buffer memory address generating circuit 26 to be controlled.
[0021]
(Function)
In the ultrasonic diagnostic apparatus 1 according to the present embodiment configured as described above, first, an analog ultrasonic signal received by the ultrasonic probe 2 is converted into a digital signal by the A / D converter 3, and the image memory 4 is stored. The image memory 4 stores image data, which is digital data obtained by imaging the intensity distribution of the reflector in each space, for example, for one screen or more with the horizontal axis as the transducer direction and the vertical axis as the distance direction. Store things. Then, the image data is extracted from the image memory 4 by the image memory address generation circuit 6 through the control circuit 5 and transferred to the enlargement processing circuit 8 connected to the data flow selector 7.
[0022]
In the enlargement processing circuit 8, first, the image data is stored in the input buffer memory 21. The stored data is Fourier transformed by the two-dimensional FFT circuit 22 and outputs a two-dimensional frequency spectrum.
[0023]
The frequency spectrum of the zero component is added to the obtained frequency spectrum according to the enlargement ratio set by the enlargement ratio setting unit 12 by the zero insertion circuit 23. The frequency spectrum to which the zero component is added is converted from the frequency spectrum to image data by the two-dimensional inverse FFT circuit 24. Then, the converted data is stored in the output buffer memory 25 by the same number as the number of samples before processing. The output buffer address generation circuit 26 takes out image data from the output buffer memory 25 in accordance with the enlargement ratio set in advance by the enlargement ratio setting unit 12, thereby interpolating between the pixels through these series of processing, and enlarging. doing.
[0024]
Here, the details of the process of interpolation and enlargement by the above method will be described. In general, the image signal f (x, y) and the function F (ξ, η) of the frequency spectrum obtained by the Fourier transform thereof have the following relationship.
[0025]
[Expression 1]

[Expression 2]

The above equation (1) means that the coordinate region representing the image signal can be converted into the frequency region, and the above equation (2) means that the frequency region can be converted into the coordinate region.
[0026]
Thus, the Fourier transform has the property of being able to mutually transform the coordinate domain and the frequency domain. In this embodiment, data is interpolated using the property of the Fourier transform.
[0027]
FIG. 3 is a diagram illustrating a method of interpolation by Fourier transform. The signal actually used is a two-dimensional signal represented by the above equation (1). Here, for the sake of simplicity, the description will be given first using a one-dimensional signal.
[0028]
FIG. 3A shows a one-dimensional signal sampled at time τa. FIG. 4B is a frequency spectrum obtained as a result of Fourier transform of the signal of FIG. FIG. 10C shows a signal sampled at a shorter time τb than the signal shown in FIG. FIG. 4D is a frequency spectrum obtained as a result of Fourier transform of the signal in FIG.
[0029]
In general, when the signal sampled at time τa in FIG. 3A is Fourier transformed, a frequency spectrum of the frequency band Fa equal to 1 / τa, that is, the inverse of the sampling time interval τa (FIG. 3B) is obtained. Interpolating between signal samples by changing the frequency band of the frequency spectrum from the relationship between the sampling time interval of the signal and the frequency band of the frequency spectrum after Fourier transform (1 / τ = f). Can do.
[0030]
For example, doubling the number of signal samples can be realized by halving the sampling interval. At this time, the frequency band of the frequency spectrum is doubled from 1 / (τa / 2) = 2 / τa = 2Fa (1 / τa = Fa) using the relationship between the sampling time and the frequency band. As a result, when the frequency band of the frequency spectrum is doubled and inverse Fourier transform is performed, the signal samples are interpolated to obtain a signal with the number of samples doubled.
[0031]
From the above description, in order to double the frequency spectrum shown in FIG. 3B, a zero-value frequency spectrum is added to the portion A of the frequency spectrum shown in FIG. When the frequency spectrum obtained by doubling the frequency band shown in FIG. 3 (d) is subjected to inverse Fourier transform, as shown in FIG. 3 (c), the sample points in FIG. 3 (a) are interpolated to double the number of samples. The resulting signal is obtained.
[0032]
In other words, by adding a zero value to the frequency spectrum, the sampling frequency Fa that satisfies the sampling theorem shown in FIG. 3B becomes the sampling frequency Fb with a wider frequency band. It can be said that the interval becomes finer and smoother interpolation becomes possible.
[0033]
By using the same number of samples of these interpolated signals as the number of samples before processing, enlarged display can be realized. For example, attention is paid to a 4-sample interval from S0 to S3 for both the original signal in FIG. 3A and the signal interpolated twice as in FIG. The sample section length is Sa in FIG. 3A and Sb in FIG. Comparing the two data, the sample section length Sb in FIG. 3C is shorter than the sample section length Sa in FIG. Thus, by using the same number of samples as display data, the signal can be expanded with the interpolated data. However, the display data is not necessarily the same as the number of samples before processing, and may be the number of samples in an arbitrary region.
[0034]
In the above, for the sake of simplicity, the one-dimensional discrete Fourier transform has been described as an example. In the following, interpolation processing in the frequency space will be described when enlarging the actual two-dimensional ultrasonic tomographic image.
[0035]
FIG. 4 shows an outline of enlargement interpolation processing of a two-dimensional ultrasonic tomographic image in the present embodiment. FIG. 4A shows an ultrasonic tomographic image A (R) stored in the image memory 4. Here, A is the reflection intensity of the ultrasonic wave held by each pixel, R represents the processing area in the image memory 4 set by the image memory address generation circuit 6 and the data flow selector 7, and the x axis It consists of a rectangular region having Nx pixels arranged at dx intervals in the direction and Ny pixels arranged at dy intervals in the y-axis direction. The origin of the processing region R is (x0, y0). When these (x0, y0), Nx, and Ny are set, an arbitrary portion of the ultrasonic tomographic image in the image memory 4 can be enlarged. The processing region R can be described by the following equation.
[0036]
[Equation 3]
R = {(x, y); x = x0 + mdx, y = y0 + ndy, x0, y0 is an arbitrary pixel address of the original image memory, 0 ≦ m <Nx, 0 ≦ n <Ny, m, n is an integer, Nx , Ny is a natural number} (3)
The ultrasonic tomographic image A (R) on the processing region R thus extracted from the image memory 4 is accumulated in the input buffer memory 21 and Fourier-transformed by the two-dimensional FFT circuit 22 most frequently used in the two-dimensional discrete Fourier transform. The original spectrum S obtained by the conversion is shown in FIG. The frequency region U in which the original spectrum S obtained by the two-dimensional FFT circuit 22 is defined is made up of the same number of discrete elements as the number of pixels Nx × Ny in the processing region R, and is given by the following equation.
[0037]
[Expression 4]
U = {(fx, fy); fx and fy are spatial frequencies about the x-axis and y-axis, respectively. Fx = − (Nx / 2−i) / (Nxdx), fy = − (Ny / 2−j) / (Nydy), 0 ≦ i <Nx, 0 ≦ j <Ny, i, j are integers} (4)
As is clear from the above equation (4), U is a rectangular range of −1/2 dx ≦ fx <+1/2 dx and −1/2 dy ≦ fy <+1/2 dy.
[0038]
Next, the original spectrum S is a set of spectra s (fx, fy) defined on the frequency domain U as in the following equation.
[0039]
[Equation 5]
S = {s (fx, fy); (fx, fy) εU} (5)
Then, the original spectrum S given by the equation (5) is input to the zero value insertion circuit 23. The zero value input circuit 23 first receives an enlargement factor M set by the enlargement factor setting unit 12 and includes the frequency region U. An extended frequency region V in which U is enlarged M times in both vertical and horizontal directions is expressed by the following equation. Set as follows.
[0040]
[Formula 6]
V = {(fx, fy); fx = −M (Nx / 2−i) / (Nxdx), fy = −M (Ny / 2−j) / (Nydy), 0 ≦ i <MNx, 0 ≦ j <MNy, i, j are integers} (6)
Next, in the extended frequency domain V, the original spectrum data s (fx, fy) is used as it is in the frequency domain U, and a zero value is inserted in the difference domain V-U. The spectrum thus obtained will be referred to as an extended spectrum T. A schematic shape of the extended spectrum T is shown in FIG. If each element of T is written as t (fx, fy), the extended spectrum T is
[Expression 7]
T = {t (fx, fy); t (fx, fy) = s (fx, fy) (when (fx, fy) ∈U), t (fx, fy) = 0 ((fx, fy) ∈ V-U)} (7)
It can be expressed as.
[0041]
The extended spectrum T obtained in this way is sent to the two-dimensional inverse FFT circuit 24 and converted into real space data.
[0042]
Here, enlargement and interpolation processing is performed according to the principle described above. Since the extended spectrum T is composed of MNx × MNy pieces of data, the output buffer memory 25 is an enlargement that is a data storage area in which the processing area R is multiplied by M both vertically and horizontally in accordance with the enlargement ratio M from the enlargement ratio setting unit 12. A processing area Q is prepared.
[0043]
Then, the interpolated enlarged image A (Q) subjected to inverse Fourier transform by the two-dimensional inverse FFT circuit 24 is sent to the output buffer 25. FIG. 4D shows a schematic shape of the interpolation enlarged image.
[0044]
The obtained display data is transferred to the DSC 10 through the video mixer 9 set by the control circuit 5 so that the enlarged image is displayed by the display mode changer 13, and the enlarged B-mode image is displayed on the monitor 11. The
[0045]
(effect)
As described above, in the ultrasonic diagnostic apparatus 1 according to the present embodiment, in the enlargement processing circuit 8, the zero value insertion circuit 23 inserts a zero value into the frequency spectrum s of the image data obtained by the two-dimensional FFT circuit 22. The frequency band is widened, and the frequency spectrum is subjected to inverse Fourier transform by the two-dimensional inverse FFT circuit 24 to output image data with a fine sampling interval, and this sample number is processed when the monitor 11 displays the B-mode image. By making the number the same as the previous number of samples, smooth interpolation is possible, and a good B-mode image can be obtained even when enlarged.
[0046]
5 and 6 relate to the second embodiment of the present invention, FIG. 5 is a block diagram showing the configuration of the enlargement processing circuit, and FIG. 6 is an explanatory diagram for explaining the operation of the linear interpolation circuit of FIG.
[0047]
Since the second embodiment is almost the same as the first embodiment, only different points will be described, and the same components are denoted by the same reference numerals and description thereof will be omitted.
[0048]
(Constitution)
In the second embodiment, a linear interpolation enlargement circuit is provided in the enlargement processing circuit 8 of the first embodiment, and the linear interpolation enlargement circuit and the interpolation enlargement circuit based on Fourier transform can be switched depending on applications. This is very different from the first embodiment.
[0049]
For this reason, as shown in FIG. 5, the enlargement processing circuit 8a of the present embodiment includes a two-dimensional FFT circuit 22 that constitutes the Fourier transform interpolation enlargement circuit 30 based on the Fourier transform described in the first embodiment, and zero insertion. In addition to the circuit 23 and the two-dimensional inverse FFT circuit 24, a linear interpolation enlargement processing circuit 31 that performs linear interpolation enlargement processing, which will be described later, and a two-dimensional FFT circuit under the control of the control circuit 5 the image data stored in the input buffer memory 22 or the enlargement processing selector 32 that selectively outputs to the linear interpolation enlargement processing circuit 31 and the output of the two-dimensional inverse FFT circuit 24 or the output of the linear interpolation enlargement processing circuit 31 selectively under the control of the control circuit 5 as an output buffer memory 25 and an output data selector 33 that outputs to 25. The control circuit 5 is connected to an enlargement processing setting unit 34 for setting selection of the linear interpolation enlargement circuit 31 or the Fourier transform interpolation enlargement circuit 30.
[0050]
Other configurations are the same as those of the first embodiment.
[0051]
(Function)
In this embodiment, digital data obtained by imaging the intensity distribution of the reflector in each space is stored in the image memory 4, and the image data stored in the image memory 4 is read by the image memory address generation circuit 6. The process until the image data is transferred to the enlargement processing circuit 8a selected by the data flow selector 7 is the same as that in the first embodiment.
[0052]
In the enlargement processing circuit 8a, the transferred image data is stored in the input buffer memory 21. Then, either the linear interpolation process set in advance by the enlargement process setting unit 34 or the interpolation enlargement process by Fourier transform is selected by the enlargement process selector 32, and the image data is transferred and processed.
[0053]
When the linear interpolation enlargement processing is selected by the enlargement processing setting unit 34, the image data is transferred to the linear interpolation enlargement processing circuit 31 selected by the enlargement processing selector 32, and is enlarged here. The processed result passes through the output data selector 33 and is stored in the output buffer memory 25.
[0054]
When the interpolation enlargement process by Fourier transform is selected by the enlargement process setting unit 34, the image data is transferred to the Fourier transform interpolation enlargement circuit 30 selected by the enlargement process selector 32, and the two-dimensional FFT in the Fourier transform interpolation enlargement circuit 30 is performed. The circuit 22, the zero insertion circuit 23, and the two-dimensional inverse FFT circuit 24 perform enlargement processing by Fourier transform. The processed result passes through the output data selector 33 and is stored in the output buffer memory 25.
[0055]
Through these processes, the enlarged image data stored in the output buffer memory 25 is transferred to the video mixer 9 and displayed on the monitor 11 by the DSC 10 as in the first embodiment.
[0056]
Here, the difference between the linear interpolation enlargement processing in the linear interpolation enlargement processing circuit 31 and the interpolation enlargement processing by Fourier transform in the Fourier transform interpolation enlargement circuit 30 will be described. In both enlargement processes, the enlargement process is performed by extracting only the number of data before the interpolation from the data increased by the interpolation, and the difference between the images obtained by the enlargement is in the interpolation method. For this reason, the description here is limited to the difference in the interpolation method.
[0057]
Furthermore, the image data is two-dimensional data as shown in FIG. 11, but here, it will be described with a one-dimensional signal as shown in FIG. 6 shows linear interpolation and interpolation by Fourier transform. FIG. 6A shows one-dimensional linear interpolation, and FIG. 6B shows interpolation by Fourier transform. 6 (a) and 6 (b), curve A is the original signal, black dots are sample points obtained by sampling curve A, and white dots are sample points obtained by interpolation processing. Show.
[0058]
In the case of the linear interpolation shown in FIG. 6A, the sample point Sa obtained by the interpolation is not on the curve A which is the original signal, but in the case of the interpolation by the Fourier transform shown in FIG. The sample point Sb obtained in this manner exists on the curve A of the original signal. In this way, in the case of linear interpolation, since the two sample points a1 and a2 on both sides of the sample point Sa to be obtained as shown in FIG. There is roughness.
[0059]
On the other hand, in the case of interpolation by Fourier transform, the two sample points b1 and b2 are smoothly connected so that the interpolated sample point Sb exists on the curve A as shown in FIG. Interpolation is obtained.
[0060]
Therefore, in order to enlarge and obtain a good image after freezing, interpolation by Fourier transform is more suitable than linear interpolation. However, since interpolation processing by Fourier transform takes time, it cannot be processed in real time. On the other hand, since linear interpolation is simply performed by averaging pixel values between two points, it can be processed in real time.
[0061]
(effect)
As described above, in the present embodiment, in addition to the effect of the first embodiment, when performing the enlargement process in real time, when performing the linear interpolation enlargement process in the linear interpolation enlargement processing circuit 31 and want to obtain a good enlarged image after freezing The enlargement process by the Fourier transform in the Fourier transform interpolation enlargement circuit 30 can be performed, and the enlargement process method can be switched according to the application.
[0062]
7 and 8 relate to the third embodiment of the present invention, FIG. 7 is a block diagram showing the configuration of the ultrasonic diagnostic apparatus, and FIG. 8 is an explanatory diagram for explaining the operation of the ultrasonic diagnostic apparatus of FIG. is there.
[0063]
Since the third embodiment is almost the same as the first embodiment, only different points will be described, and the same components are denoted by the same reference numerals and description thereof will be omitted.
[0064]
(Constitution)
In the present embodiment, a function for setting an enlarged region of an image is added to the ultrasonic diagnostic apparatus 1 of the first embodiment.
[0065]
For this reason, as shown in FIG. 7, the ultrasonic diagnostic apparatus 41 of this embodiment is provided with an enlarged region setting device 42 instead of the display mode switching operation device 13 (see FIG. 1) of the first embodiment. Composed.
[0066]
Other configurations are the same as those of the first embodiment.
[0067]
(Function)
The present embodiment is the same as the first embodiment until digital data in which the intensity distribution of the reflector in each space is imaged is stored in the image memory 4.
[0068]
Then, the image data corresponding to the enlargement area is extracted from the image memory 4 by the image memory address generator 6 through the control circuit 5 using the enlargement area set by the enlargement area setting unit 42. The extracted data is transferred to the enlargement processing circuit 8 connected to the data flow selector 7 in the same manner as in the first embodiment, subjected to interpolation enlargement processing by Fourier transform, and is processed on the image data before enlargement by the video mixer 9. Is displayed on the monitor 11 by the DSC 10.
[0069]
The display of the obtained enlargement processing result is shown in FIG. FIG. 8A shows the display before enlargement, and FIG. 8B shows the result of enlarging the portion F in FIG. 8A and overlaying it on the image of FIG. (B) The result obtained in this embodiment is obtained. Here, since the enlarged display area G displayed in FIG. 8B uses the same number of data as the display area before enlargement, only the range α in FIG. 8A is actually displayed.
[0070]
(effect)
As described above, in this embodiment, in addition to the effect of the first embodiment, the obtained good enlarged image can be displayed in the designated area on the image before enlargement. It can be easily confirmed.
[0071]
In the above description, the enlargement area may be designated by displaying a mouse cursor on the display of the image before enlargement and operating the cursor with the mouse. Further, the designated enlarged area may be moved with the mouse. In this way, by displaying the designated enlargement area on the pre-enlargement image by moving it with the mouse, an enlargement display in which the ultrasonic image is viewed with a magnifying glass is realized.
[0072]
In the above description, the display data is the same as the number of samples before processing, but is not necessarily limited to this, and any number of samples may be used as the display data.
[0073]
Further, the enlargement processing circuit 8a described in the second embodiment is used in place of the enlargement processing circuit 8, and the linear interpolation enlargement process and the Fourier interpolation enlargement process are switched, and the image data before enlargement and the image data after enlargement are switched in real time. Image data overlay display may be realized.
[0074]
9 and 10 relate to the fourth embodiment of the present invention, FIG. 9 is a block diagram showing the configuration of the ultrasonic diagnostic apparatus, and FIG. 10 is a flowchart showing the flow of enlargement processing by the DSP calculation unit of FIG. is there.
[0075]
Since the fourth embodiment is almost the same as the first embodiment, only different points will be described, and the same components are denoted by the same reference numerals and description thereof will be omitted.
[0076]
(Constitution)
In contrast to the ultrasonic diagnostic apparatus 1 of the first embodiment that implements image enlargement processing by hardware, in this embodiment, a high-speed signal processor (DSP) is programmed with an enlargement processing algorithm, What is realized by software is very different.
[0077]
For this reason, as shown in FIG. 9, in the ultrasonic diagnostic apparatus 51 of the present embodiment, the enlargement processing circuit 8 of the first embodiment is replaced with a DSP calculation unit 52. In addition to the input data buffer 21, the output buffer memory address generation circuit 26, and the output data buffer 25, the DSP arithmetic unit 52 has an internal configuration including a DSP 53 for performing an enlargement process by software, and a program memory storing an enlargement process algorithm 54.
[0078]
Other configurations are the same as those of the first embodiment.
[0079]
(Function)
The present embodiment is the same as the first embodiment until digital data in which the intensity distribution of the reflector in each space is imaged is stored in the image memory 4.
[0080]
Then, the digital data stored in the image memory 4 is transferred to the input buffer memory 21 in the DSP calculation unit 52.
[0081]
The DSP calculation unit 52 loads a program from the program memory 54 to the DSP 53 in advance. The DSP 53 executes the loaded program and performs an image enlargement process of the data in the input buffer memory 21.
[0082]
Here, the processing procedure of the DSP calculation unit 52 will be described with reference to the flowchart shown in FIG.
[0083]
The DSP calculation unit 52 executes a program, first initializes the DSP 53 in step S1, and reads data from the input buffer memory 21 into the DSP 53 in step S2. The next steps S3 to S5 are the part of the interpolation enlargement process using Fourier transform.
[0084]
In step S3, a two-dimensional FFT process is performed to obtain a frequency spectrum. Considering this step with one-dimensional data, it corresponds to the processing for obtaining the frequency spectrum of FIG. 3B from the signal of FIG. 3A described in the first embodiment.
[0085]
In step S4, a zero-value spectrum is added to the frequency spectrum, and the frequency band is widened. Considering this step with one-dimensional data, a zero value is inserted into the portion A of the frequency spectrum of FIG. 3D with respect to the frequency spectrum of FIG. 3B described in the first embodiment. This corresponds to the process of expanding the frequency band of the frequency spectrum.
[0086]
In the next step S5, a frequency spectrum having a wide frequency band is subjected to inverse Fourier transform to obtain a signal with a fine sampling interval. Considering this step with one-dimensional data, the frequency spectrum with the expanded frequency band shown in FIG. 3 (d) described in the first embodiment is subjected to inverse Fourier transform to obtain the signal shown in FIG. 3 (c). It corresponds to the process to convert.
[0087]
In step S6, the interpolated data is stored in the output buffer memory 25 only for the number of samples before processing. This step also corresponds to the process of transferring the data section Sb shown in FIG. 3C, considering the one-dimensional data shown in FIG. 3 described in the first embodiment.
[0088]
The enlarged image data is switched to the enlarged display mode by the video mixer 9 from the display mode switching operation unit 13 as in the first embodiment, transferred to the DSC 10, and enlarged to the monitor 11. A B-mode image is displayed.
[0089]
(effect)
As described above, in this embodiment, in addition to the effects of the first embodiment, the enlargement processing algorithm is programmed, and the enlargement processing circuit itself is replaced with the DSP calculation unit 52, so that the algorithm can be easily changed, and according to the specification Flexible response is possible.
[0090]
In the description of the present embodiment, the processor for enlargement processing is a DSP, but this may be a microcomputer or a CPU.
[0091]
[Appendix]
(Additional Item 1) An original image memory for storing an ultrasonic tomographic image as a set of pixels two-dimensionally arranged at regular intervals dx and dy along the orthogonal x-axis and y-axis, respectively.
Processing region R of the ultrasonic tomogram R = {(x, y); x = x 0 + mdx, y = y 0 + ndy, x 0, y 0 are arbitrary pixel addresses of the original image memory, 0 ≦ m <Nx, 0 ≦ n <Ny , M, n are integers, Nx, Ny are natural numbers}
Two-dimensional discrete Fourier transform is performed on the Nx × Ny pixel data included in the processing region R in the frequency space, and the frequency region U = {(fx, fy); fx and fy are about the x axis and the y axis, respectively. At spatial frequency, fx = − (Nx / 2−i) / (Nxdx), fy = − (Ny / 2−j) / (Nydy), 0 ≦ i <Nx, 0 ≦ j <Ny, i, j Fourier transform means for obtaining an original spectrum S = {s (fx, fy); (fx, fy) ∈U} consisting of Nx × Ny pieces of spectral data s (fx, fy) on an integer};
An enlargement factor input means for inputting an enlargement factor M for enlarging the ultrasonic tomogram on the processing region R set by the spatial original region setting means;
Extended frequency region V including the frequency region U = {(fx, fy); fx = −M (Nx / 2−i) / (Nxdx), fy = −M (Ny / 2−j) / (Nydy), 0 ≦ i <MNx, 0 ≦ j <MNy, i, j is an integer}, and the value 0 is inserted into all points on VU which is the difference region, and the extended spectrum T is extracted from the original spectrum S. = {T (fx, fy); t (fx, fy) = s (fx, fy) (when (fx, fy) ∈U), t (fx, fy) = 0 ((fx, fy) ∈V Zero value insertion means for generating
A spatial enlargement region setting means for setting an enlargement processing region Q obtained by enlarging the processing region R of the ultrasonic tomographic image M times;
Inverse Fourier transform means for performing two-dimensional discrete inverse Fourier transform on the extended spectrum T generated by the zero value insertion means and storing it in the expansion processing region Q;
With
The Fourier transform means calculates the original spectrum S by performing a two-dimensional discrete Fourier transform on the ultrasonic tomogram on the frequency space, and the zero value insertion means uses the original spectrum S calculated by the Fourier transform means to calculate the original spectrum S. Extended interpolation by generating extended spectrum T,
The space expansion region setting means sets the expansion processing region Q obtained by magnifying the processing region R of the ultrasonic tomographic image M times, and the inverse Fourier transform unit generates the extended spectrum T generated by the zero value insertion unit. Is subjected to enlargement processing by performing two-dimensional discrete inverse Fourier transform and storing in the enlargement processing region Q
An ultrasonic diagnostic apparatus.
[0092]
(Additional Item 2) In the ultrasonic diagnostic apparatus according to Additional Item 1,
Linear interpolation means for linearly interpolating between enlarged pixels;
Selection means for selecting the extended interpolation on the frequency space and the linear interpolation;
An ultrasonic diagnostic apparatus comprising:
[0093]
Additional Item 2 provides an ultrasonic diagnostic apparatus capable of switching between interpolation enlargement processing by Fourier transform and enlargement processing by linear interpolation according to the application, and obtaining a result of real-time enlargement processing and good enlargement processing after freezing. It is an object.
[0094]
According to the ultrasonic diagnostic apparatus of Supplementary Item 2, by switching between two types of enlargement processing by linear interpolation and enlargement processing by Fourier transform, a real-time enlargement process and a good enlarged image are required after image freeze Both of the enlargement processing can be realized and can be selected according to the application.
[0095]
(Additional Item 3) In the ultrasonic diagnostic apparatus according to Additional Item 1,
The enlargement processing area Q is set on an ultrasonic tomographic image area R stored in the original image memory.
An ultrasonic diagnostic apparatus.
[0096]
The supplementary item 3 is intended to provide an ultrasonic diagnostic apparatus that displays an enlarged image obtained by interpolation enlargement processing using Fourier transform in an overlapped manner on a designated area on an image before enlargement.
[0097]
According to the ultrasonic diagnostic apparatus of the supplementary item 3, an enlarged portion can be easily confirmed by displaying the obtained good enlarged image in the designated area on the image before enlargement.
[0098]
(Additional Item 4) An original image memory for storing an ultrasonic tomogram as a set of pixels two-dimensionally arranged at regular intervals dx and dy along the orthogonal x-axis and y-axis, respectively.
Processing region R of the ultrasonic tomogram R = {(x, y); x = x 0 + mdx, y = y 0 + ndy, x 0, y 0 are arbitrary pixel addresses of the original image memory, 0 ≦ m <Nx, 0 ≦ n <Ny , M, n are integers, Nx, Ny are natural numbers}
An enlargement factor input means for inputting an enlargement factor M for enlarging the ultrasonic tomogram on the processing region R set by the spatial original region setting means;
A spatial enlargement region setting means for setting an enlargement processing region Q obtained by enlarging the processing region R of the ultrasonic tomographic image M times;
A signal processor for extending and interpolating the ultrasonic tomogram in a frequency space;
With
The signal processor
Two-dimensional discrete Fourier transform is performed on Nx × Ny pixel data included in the processing region, and the frequency region U = {(fx, fy); fx and fy are spatial frequencies about the x-axis and the y-axis, respectively. = − (Nx / 2−i) / (Nxdx), fy = − (Ny / 2−j) / (Nydy), 0 ≦ i <Nx, 0 ≦ j <Ny, i, j is an integer} Nx Fourier transform operation for obtaining an original spectrum S = {s (fx, fy); (fx, fy) εU} composed of × Ny pieces of spectral data s (fx, fy);
Extended frequency region V including the frequency region U = {(fx, fy); fx = −M (Nx / 2−i) / (Nxdx), fy = −M (Ny / 2−j) / (Nydy), 0 ≦ i <MNx, 0 ≦ j <MNy, i, j is an integer}, and the value 0 is inserted into all points on VU which is the difference region, and the extended spectrum T = {t (fx , Fy); t (fx, fy) = s (fx, fy) (when (fx, fy) ∈U), t (fx, fy) = 0 (when (fx, fy) ∈V−U) } To generate a zero value;
An inverse Fourier transform operation in which the extended spectrum T generated by the zero value insertion means is subjected to a two-dimensional discrete inverse Fourier transform and stored in the expansion processing region Q;
I do
An ultrasonic diagnostic apparatus.
[0099]
Additional Item 4 aims to provide an ultrasonic diagnostic apparatus that can change an algorithm by changing a program and flexibly cope with a specification change by using a high-speed arithmetic processing processor such as a DSP for an enlargement processing unit. It is said.
[0100]
According to the ultrasonic diagnostic apparatus of the supplementary item 4, since the algorithm can be easily changed by programming the enlargement processing algorithm and replacing the enlargement processing circuit itself with a DSP, a flexible response according to the specification is possible.
[0101]
【The invention's effect】
As described above, according to the ultrasonic diagnostic apparatus of the present invention, Super In the ultrasonic tomographic image enlargement process, the ultrasonic tomographic image is smoothly interpolated to obtain a good B-mode image at the time of enlargement, and two types of enlargement processing by linear interpolation and interpolation enlargement processing by Fourier transform are available. By switching and using, both real-time enlargement processing and enlargement processing that requires a good enlarged image after image freezing can be realized and can be selected according to the application. In addition, by displaying the obtained magnified image in a specified area on the image before magnification, it is possible to easily confirm the magnified part. There is an effect that can.
[Brief description of the drawings]
FIG. 1 is a configuration diagram showing a configuration of an ultrasonic diagnostic apparatus according to a first embodiment of the present invention.
FIG. 2 is a block diagram showing the configuration of the enlargement processing circuit of FIG.
FIG. 3 is a first explanatory diagram illustrating the operation of the enlargement processing circuit of FIG. 2;
4 is a second explanatory diagram illustrating the operation of the enlargement processing circuit of FIG. 2; FIG.
FIG. 5 is a block diagram showing a configuration of an enlargement processing circuit according to a second embodiment of the present invention.
6 is an explanatory diagram for explaining the operation of the linear interpolation circuit of FIG. 5;
FIG. 7 is a configuration diagram showing a configuration of an ultrasonic diagnostic apparatus according to a third embodiment of the present invention.
FIG. 8 is an explanatory diagram for explaining the operation of the ultrasonic diagnostic apparatus of FIG.
FIG. 9 is a configuration diagram showing the configuration of an ultrasonic diagnostic apparatus according to a fourth embodiment of the present invention.
FIG. 10 is a flowchart showing a flow of enlargement processing by the DSP calculation unit of FIG. 9;
FIG. 11 is an explanatory diagram for explaining a conventional method for enlarging a tomographic image of a living body
12 is a view showing a display example of an enlarged image obtained by the enlargement method of FIG. 11;
[Explanation of symbols]
1, 41, 51 ... Ultrasonic diagnostic apparatus
2 ... Ultrasonic probe
3 ... A / D converter
4 ... Image memory
5 ... Control circuit
6 ... Address circuit for image memory
7 Data flow selector
8, 8a ... Expansion processing circuit
9 ... Video mixer
10 ... DSC
11 ... Monitor
12 ... Enlargement ratio setting
13 ... Display mode switching controller
21 ... Input buffer memory
22 ... Two-dimensional FFT circuit
23 ... Zero insertion circuit
24 ... Two-dimensional inverse FFT circuit
25 ... Output buffer memory
26: Output buffer memory address generation circuit
30 ... Fourier transform interpolation expansion circuit
31 ... Linear interpolation enlargement processing circuit
32 ... Enlargement processing selector
33 ... Output data selector
34 ... Linear interpolation expansion circuit
42 ... Enlarged area setting device
52. DSP calculation unit
53 ... DSP
54 ... Program memory

Claims (1)

An original image memory for storing an ultrasonic tomogram as a set of pixels two-dimensionally arranged at regular intervals dx and dy along the orthogonal x-axis and y-axis, respectively;
Processing area R of the ultrasonic tomographic image R = {(x, y); x = x 0 + mdx, y = y 0 + ndy, x 0, y 0 is an arbitrary pixel address of the original image memory, 0 ≦ m <Nx, 0 ≦ n <Ny , M, n are integers, Nx, Ny are natural numbers}
Two-dimensional discrete Fourier transform is performed on the Nx × Ny pixel data included in the processing region R in the frequency space, and the frequency region U = {(fx, fy); fx and fy are about the x axis and the y axis, respectively. At spatial frequency, fx = − (Nx / 2−i) / (Nxdx), fy = − (Ny / 2−j) / (Nydy), 0 ≦ i <Nx, 0 ≦ j <Ny, i, j Fourier transform means for obtaining an original spectrum S = {s (fx, fy); (fx, fy) ∈U} consisting of Nx × Ny pieces of spectral data s (fx, fy) on an integer};
An enlargement factor input means for inputting an enlargement factor M for enlarging the ultrasonic tomogram on the processing region R set by the spatial original region setting means;
Extended frequency region V including the frequency region U = {(fx, fy); fx = −M (Nx / 2−i) / (Nxdx), fy = −M (Ny / 2−j) / (Nydy), 0 ≦ i <MNx, 0 ≦ j <MNy, i, j is an integer}, and the value 0 is inserted into all points on VU which is the difference region, and the extended spectrum T is extracted from the original spectrum S. = {T (fx, fy); t (fx, fy) = s (fx, fy) (when (fx, fy) ∈U), t (fx, fy) = 0 ((fx, fy) ∈V Zero value insertion means for generating
An enlarged processing area Q obtained by enlarging the processing area R of the ultrasonic tomographic image M times is set, and the enlarged processing area Q is set on the ultrasonic tomographic image area R stored in the original image memory. A space expansion area setting means to perform ,
An inverse Fourier transform means for performing a two-dimensional discrete inverse Fourier transform on the extended spectrum T generated by the zero value insertion means and storing it in the expansion processing region Q;
Linear interpolation means for linearly interpolating between enlarged pixels;
Selection means for selecting the extended interpolation on the frequency space by the Fourier transform means, zero value insertion means and the linear interpolation;
An ultrasonic diagnostic apparatus comprising:

Images