JP4118535B2 - X-ray inspection equipment - Google Patents

Description

………（数１）
（数１）より、f_dの値は、f_g、ｄ、Ｄの値を変えることによって任意に設定することができる。図５には、f_dの値を検出器のナイキスト周波数f_qより僅かに低く設定した例が示される。周波数f_dにおける周波数強度５０２は、f_dがナイキスト周波数f_qに近づくに従って小さくなる。これは平面型Ｘ線検出器４のボケにより、Ｘ線グリッド４の周期的な構造情報（以下、グリッド縞とする）が目立たなくなることを意味する。ここで、ナイキスト周波数f_qと投影グリッド密度f_dとの差を、ビート周波数f_bとして、次式で定義する。
【００４８】
【数２】

………（数２）
すなわち、Ｘ線透過像中に含まれるグリッド縞を目立たなくするためには、f_bをできるだけ小さくすればよい。一方、散乱Ｘ線の周波数成分５０１には低周波成分の情報のみが含まれる。
【００４９】
図６は、検出されたＸ線透過像に対する間引きサンプリングを説明するための図である。また、図７は、間引きサンプリング時におけるナイキスト周波数と直接Ｘ線成分および散乱Ｘ線成分の空間周波数分布との関係を説明するための図である。 図４（Ａ）に説明したように、Ｘ線透過像は平面型Ｘ線検出器４の全画素を用いて検出する。すなわち、画素位置(i、j)において検出された信号量をf(i、j)とすると、Ｘ線透過像はf(i、j)（i、j＝0〜N-1）と表すことができる。ただし、Nを平面型Ｘ線検出器４のi、j方向の画素数とした。本実施例１では、N＝1024である。間引きサンプリングは、図６（Ａ）および（Ｃ）に示されるように、f(i、j)中から、i方向に１画素おきに画素値を抽出して画像を作成する方法である。このとき、画素の抽出位置によって、以下の２種類の抽出画像を作成することができる。
【００５０】
【数３】

………（数３）
【００５１】
【数４】

………（数４）
ｆ₁(m、n)およびf₂(m、n)のm方向のナイキスト周波数f'_q（＝１/（４Δｄ））は、f(i、j)のナイキスト周波数f_q（＝１/（２Δｄ））の半分となる。すなわち、f'_q＝f_q/２である。図７に示されるように、間引きサンプリング時の直接Ｘ線成分５００は、f'_qを中心に線対称に折れ込んでエリアシングを発生する。図７中、７０２は直接Ｘ線成分５００のエリアシングによる折り返し成分であり、また、７０１は散乱Ｘ線成分５０１のエリアシングによる折り返し成分である。
【００５２】
この結果、投影グリッド密度f_dの成分５０２が、周波数f_bのビート成分７０２となって低周波数側に現れ、干渉縞を発生する。従って、抽出画像f₁(m、n)およびf₂(m、n)のm方向のプロファイルには干渉縞が含まれる（図６（Ｂ）および（Ｄ）参照）。
【００５３】
上述のように、散乱線補正を行うためには上記干渉縞の振幅を計測すればよい。しかし、抽出画像f₁(m、n)およびf₂(m、n)には、それぞれ干渉縞信号の他に、被検体６の構造情報も含まれているため、干渉縞成分の振幅を正確に抽出するのは難しい。
【００５４】
干渉縞成分の振幅を正確に抽出するためには、f₁(m、n)およびf₂(m、n)の差分画像s(m、n)を次式によって作成する。
【００５５】
【数５】

………（数５）
（数３）および（数４）に示されるように、f₁(m、n)およびf₂(m、n)のf(i、j)に対するサンプリング位置は、i方向に互いに１画素ずつずれている。このため、f₁(m、n)およびf₂(m、n)中に含まれる干渉縞は、m方向に互いに１８０度位相がずれる。従って、差分画像s(m、n)は、図６（Ｅ）に示されるような干渉縞成分のみが含まれるため、s(m、n)から容易に干渉縞の振幅を求めることができる。
【００５６】
図４〜図６を用いた説明においては、投影グリッド密度f_dをナイキスト周波数f_qよりビート周波数f_b分だけ低い値に設定した（以下、非エリアシング設定と呼ぶ）。このとき、Ｘ線透過像f(i、j)中に現れるグリッド縞の周波数はf_d＝(f_q−f_b)であり、抽出画像f₁(m、n)およびf₂(m、n)中に現れる干渉縞の周波数はf_bであった。上述のように、グリッド縞を目立たなくするためにはf_bをできるだけ小さくする必要がある。しかし、このとき同時に干渉縞の周波数f_bが小さくなり、その位置分解能が低下してしまう。このように、f_bは大きくし過ぎても、小さくし過ぎてもいけない。通常、f_bは、２〜５［本/ｃｍ］程度に設定するのが望ましい。
【００５７】
図８は、投影グリッド密度の別の設定方法を説明するための図である。非エリアシング設定においては、投影グリッド密度f_dをナイキスト周波数f_qよりビート周波数f_b分だけ低い値に設定した。これに対し、以下では投影グリッド密度f_dをナイキスト周波数f_qよりビート周波数f_b分だけ高い値に設定する方法（以下、エリアシング設定とする）について説明する。エリアシング設定時には、全画素サンプリングにおいてもエリアシングが発生する。このとき、投影グリッド密度f_dの周波数成分５０２は、エリアシングによって周波数（f_q−f_d）の成分７０２となって観察される（図８（Ａ）参照）。
【００５８】
周波数f_q−f_bという値は、非エリアシング設定の場合と同じ値であるが、エリアシング設定の場合、f_qを中心とする折り返し成分としてグリッド縞が観察されているため、図５の場合よりスペクトル強度が小さくなる。すなわちエリアシング設定にすることで、非エリアシング設定の場合よりもグリッド縞を目立たなくすることができる。
【００５９】
一方、間引きサンプリング時にも、投影グリッド密度f_dにおける周波数成分５０２はエリアシングによって周波数f_bの干渉縞成分として観察される。ただし、本干渉縞成分のスペクトル強度も、非エリアシング設定の場合に比べて小さくなるため、干渉縞振幅の測定精度が低下するという問題もある。以上のように、投影グリッド密度の設定はグリッド縞の抑制を優先するか、干渉縞振幅の測定精度の向上を優先するかによって、エリアシング設定か非エリアシング設定かを選択できる。
【００６０】
なお、本実施例１ではグリッド密度f_gを７０［本/ｃｍ］、Ｘ線発生点−グリッド距離ｄを４０［ｃｍ］、Ｘ線発生点−検出器距離Ｄを１００［ｃｍ］としたので、数１より投影グリッド密度f_dは２８［本/ｃｍ］となる。一方、画素間隔Δｄが２００［μｍ］であるため、ナイキスト周波数f_qは２５［本/ｃｍ］である。従って、投影グリッド密度はエリアシング設定されており、そのビート周波数は３［本/ｃｍ］である。なお、投影グリッド密度f_dを非エリアシング設定するためには、例えば上記設定においてｄを３１.４［ｃｍ］とすればよい。このとき投影グリッド密度f_dは２２［本/ｃｍ］となり、同じくビート周波数３［本/ｃｍ］の干渉縞が発生する。
【００６１】
図９は、干渉縞画像に基づき干渉縞振幅の空間分布を導出する方法を説明するための図である。（数５）により得られた差分画像s(m、n)（m＝0〜N/2-1、n＝0〜N-1）は、干渉縞成分のみを含む干渉縞画像９０である（図９中の（ａ））。干渉縞振幅分布演算手段１１０４は、干渉縞画像９０に基づいて干渉縞の振幅分布画像９５を求める手段であり、以下の演算手順からなる。
【００６２】
まず、干渉縞ピーク検出手段９１は、干渉縞画像９０のm方向のプロファイルに対して干渉縞９９のピーク位置（極大値および極小値）を検出する（図９中の（ｂ））。上記m方向のピーク検出は、全てのn（＝0〜N-1）の位置に対して行われる。次にピーク間画素補間手段９２は、m方向に隣接する極大値同士および極小値同士を直線で結び、干渉縞９９の上限包絡線９６および下限包絡線９７とする（図９中の（ｃ））。
【００６３】
また、上記補間値は、画素位置m（＝0〜N/2-1）上の点のみならず、m方向に隣接する画素の中間位置においても求める。画素の中間位置において補間値を求めることで、間引きサンプリング後にN/２個となっていたm方向の画素数を再びNに戻すことができる。上記補間は全てのn（＝0〜N-1）の位置に対して行われる。n方向の位置は全サンプリング画像のj方向の位置と一致するため、補間終了時には画素マトリックスを再び(i、j)（i、j＝0〜N-1）に戻すことができる。上限包絡線９６と下限包絡線９７の差分９８は、干渉縞９９の振幅に相当する。
【００６４】
従って、次に振幅演算手段９３は全ての位置(i、j)（i、j＝0〜N-1）において差分９８を求め、各位置における干渉縞振幅分布とする（図９中の（ｄ））。ただし、上記干渉縞分布は上記直線補間によって生じた高周波成分が含まれるため、２次元ＬＰＦ（Low Pass Filter）９４を用いて高周波成分を除去し、最終的な振幅分布画像A(i、j) を得る（図９中の（ｅ））。２次元ＬＰＦ９４の例としては、既知の移動平均処理（移動平均をとる画素領域：２×２画素〜５×５画素程度）等が挙げられる。
【００６５】
図１０は、間引きサンプリングにおける別のサンプリング方法を説明するための図である。図６においては、間引きサンプリングとしてi方向に１画素おきに画素を抽出する方法を示したが、抽出の間隔を一般にk画素おき（1≦k）として拡張してもよい。一例として、図１０には、２画素おきに抽出する例を示す。この場合、図１０（Ａ）〜（Ｃ）に示すような３種類の抽出方法が存在し、サンプリング位置がi方向に１画素シフトする毎に、間引きサンプリングによって発生する干渉縞の位相シフトが１２０度変化する。
【００６６】
すなわち、図１０（Ａ）と（Ｂ）の間、および図１０（Ｂ）と（Ｃ）の間における干渉縞の位相シフトは１２０度となる。また、図１０（Ａ）と（Ｃ）の間における干渉縞の位相シフトは２４０度である。いずれの場合も、両画像の差分画像を作成すると信号中の被写体の構造情報が取り除いて干渉縞のみを抽出することができる。
【００６７】
また、差分をとる画像間の位相シフト量に応じて抽出される干渉縞の位相が変化するが、その振幅分布は変化しない。すなわち、上記図１０（Ａ）〜（Ｃ）の、どの組み合わせに対して差分画像を作成しても、後述する図１１における干渉縞振幅分布演算手段１１０４によって最終的に得られる振幅分布画像A(i、j)には大差が生じない。ただし、差分によって被写体の構造情報を精度よく除去するには、差分画像間の画素抽出位置がなるべく近い方がよい。すなわち、画素抽出位置が隣接する２つの抽出画像を用いて差分を行うのが望ましい。一般に、k画素おきに間引きサンプリングを行ったとすると、画素抽出位置が隣接する２つの抽出画像間の位相シフト量qは、次式で表される。
【００６８】
【数６】

………（数６）
差分画像中にはＸ線の量子ノイズが含まれており、干渉縞振幅の測定精度を劣化することがある。このため、間引きサンプリングを行う前に、元画像f(i、j)（i、j＝0〜N-1）に対して移動平均処理を行って量子ノイズを軽減してもよい。i、j方向に移動平均をとる画素数を、それぞれ、2I+1、2J+1（I、J＝0、1、2、…）とすると、移動平均後の画像f_m(i、j)は、次式で計算される。
【００６９】
【数７】

………（数７）
ただし、i方向の移動平均を行うと高周波成分が低下して、間引き画像中の干渉縞強度が低下する。このため、i方向の移動平均は行わないのが一般的である。一方、j方向には通常I＝5〜10程度の範囲で移動平均をとる。
【００７０】
上記移動平均処理、間引きサンプリング処理、振幅分布画像導出処理は、Ｘ線グリッド３のスリット方向がj方向に垂直であるものとして説明した。これに対し、Ｘ線グリッド３のスリット方向がi方向に配置されている場合には、上記全ての処理においてi方向とj方向を入れ替えればよい。また、Ｘ線グリッド３として格子状のスリットを有するものを使用した場合、上記全ての処理をi方向およびj方向の両方について行えばよい。
【００７１】
以上、間引きサンプリングを用いて干渉縞の振幅分布画像を示したが、以下では、振幅分布画像を用いて散乱線補正を行う方法について説明する。
【００７２】
Ｘ線透過像中のある画素値f(i、j)（I＝0〜N-1）に含まれる直接線成分量をf_d(i、j)とすると、f_d(i、j)は同じ位置における干渉縞の振幅A(i、j)に比例する。従って、干渉縞の振幅と直接線成分量の比率（キャリブレーション比）を予め求めておくことで、直接線分布画像f_d(i、j)を求めることができる。
【００７３】
キャリブレーション比は、エア画像を用いて求めることができる。一般に、キャリブレーション比は平面型Ｘ線検出器４上の検出位置によって多少異なるため、その分布をキャリブレーション画像として求める。いま、キャリブレーション画像をr(i、j)で表すと、r(i、j)は次式で求められる。
【００７４】
【数８】

………（数８）
ただし、f_air(i、j)は被検体６を配置しない状態で計測したエア画像であり、A_air(i、j)はf_air(i、j)に対して求めた振幅分布画像である。
【００７５】
このとき、任意のＸ線透過像f(i、j)に対して、その直接線成分量を、次式で概算することができる．
【００７６】
【数９】

………（数９）
ここで、（数９）の計算結果を直接線成分の概算値としたのは、f_d'(i、j)中には高周波成分が欠如しているためである。なお、上記高周波成分の欠如は振幅分布画像A(i、j)中の高周波成分の欠如に起因し、A(i、j)中の高周波成分の欠如はピーク間画素補間手段９２による補間処理に起因する。直接線成分概算値f_d'(i、j)から正確な直接線分布画像f_d(i、j)を得るためには、まず散乱線分布画像f_s(i、j)を求め、次に元画像f(i、j)からf_s(i、j)を減算すればよい。散乱線分布画像は、次式で求まる。
【００７７】
【数１０】

………（数１０）
ただし、g(i、j)は２次元ＬＰＦ(Low Pass Filter)であり、演算**は２次元畳み込み積分を表すものとする。（数１０）において、右辺｛ ｝内の計算結果は散乱線成分と、直接線の高周波成分を含む。このため、ＬＰＦを用いて高周波成分（直接線成分）を取り除き、散乱線成分のみを抽出する。ＬＰＦとしては通常のデジタルフィルタ、もしくは移動平均処理などを用いる。散乱線分布画像f_s(i、j)が求まったら、直接線分布画像f_d(i、j)は、次式で求めることができる。
【００７８】
【数１１】

………（数１１）
以上、キャリブレーション画像を用いて散乱線補正を行う方法について説明した。なおキャリブレーション画像中におけるキャリブレーション比の変化が比較的小さい場合はr(i、j)＝roとして近似してもよい。この場合、roには、例えばr(i、j)の平均値等を用いる。また、r(i、j)もしくはroの値は、Ｘ線管１の管電圧によって多少変化する。従って、予め数種類の管電圧に対してr(i、j)またはroを測定しておき、撮影または透視時の管電圧に応じてこれらを選択してもよい。
【００７９】
図１１は、キャリブレーション画像を求める手順を説明するためのブロック図である。キャリブレーション画像の作成は、まずエア画像の撮影から行う。エア画像の撮影によって、画像メモリ１００にエア画像f_air(i、j)が確保される。次に、移動平均演算手段１１００を用いてi、j方向の移動平均を行う。なお、移動平均の計算には（数７）を用いる。
【００８０】
次に、間引き画像作成手段１１０１および１１０２を用いて、抽出画像f₁(m、n)およびf₂(m、n)をそれぞれ作成する。なお、抽出画像の作成には、（数３）および（数４）を用いる。次に、減算器１１０３により抽出画像f₁(m、n)およびf₂(m、n)の差分がとられ、差分画像s(m、n)が作成される。次に、干渉縞振幅分布演算手段１１０４によって、エア画像に対する振幅分布画像A_air(i、j)が作成される。なお、干渉縞振幅分布演算手段１１０４による演算については、図９を用いて詳細を説明したので、ここでは説明を省略する。
【００８１】
次に、除算器１１０５により振幅分布画像A_air(i、j)とf_air(i、j)の比が計算され、キャリブレーション画像r(i、j)が作成される。最後にキャリブレーション画像r(i、 j)は、コンソール１０から入力される管電圧の情報と共にキャリブレーション画像保存メモリ１０２に保存される。
【００８２】
図１２は、キャリブレーション画像に基づいて散乱線分布画像を求める手順を説明するためのブロック図である。まずＸ線透視像または撮影像f(i、j)が、画像メモリ１００に確保される。引き続き行われる移動平均演算手段１１００〜干渉縞振幅分布演算手段１１０４に至るまでの過程は、図１１に示した上記キャリブレーション画像の作成手段と同一のため、説明を省略する。上記過程によって振幅分布画像A(i、 j)が作成される。
【００８３】
次に、積算器１２００は、振幅分布画像A(i、j)とキャリブレーション画像保存メモリ１０２に保存されたキャリブレーション画像r(i、j)を積算して、直接線成分概算画像f'_d(i、j)を作成する。なお、本積算は上記（数９）の演算に相当する。また上記積算においては、元画像f(i、j)を作成した時の管電圧に最も近い管電圧で作成されたキャリブレーション画像r(i、j)が選択される。次に、減算器１２０１は元画像f(i、j)から直接線成分概算画像f'_d(i、j)を減算し、引き続きＬＰＦ１２０２は上記減算画像の高周波成分を遮断して散乱線成分画像f_s(i、j)を作成する。なお、上記減算器１２０１およびＬＰＦ１２０２による一連の作業は（数１０）に相当する。最後に、散乱線成分画像f_s(i、j)は散乱線画像保存メモリ１０３に保存される。
【００８４】
以上示したように、本実施例１では間引きサンプリングによって作成した２枚の抽出画像に基づいて干渉縞の振幅分布を計測し、散乱線補正を行う。１枚のＸ線透過像から正確に干渉縞を得ることができるため、高速かつ高精度の散乱線補正を行うことができる。
【００８５】
（実施例２）
図１３は、本発明の第２の実施例に係るＸ線検査装置の構成を説明するための図である。
【００８６】
以下、実施例１に係るＸ線検査装置との相違点のみについて説明すると、実施例１に係るＸ線検査装置ではＸ線管１、コリメータ２、およびＸ線グリッド３が被検体６の上面に配置されていた（以下、オーバーチューブ構成とする）のに対し、本実施例２に係るＸ線検査装置では、Ｘ線管１、コリメータ２、およびＸ線グリッド３が全て寝台天板７の下面に配置される（以下、アンダーチューブ構成とする）。
【００８７】
オーバーチューブ構成の場合、Ｘ線グリッド３と被検体６との間の距離が狭いため被検体６が誤ってＸ線グリッド３に接触する可能性が高いという問題があったが、アンダーチューブ構成にすることで、Ｘ線グリッド３が被検体６に誤って接触してしまう危険を回避することができる。
【００８８】
（実施例３）
図１４は、本発明の第３の実施例に係るＸ線検査装置の構成を説明するための図である。本実施例３では、間引きサンプリングを用いずに干渉縞振幅分布を計測する。以下、実施例１に係るＸ線検査装置との相違点のみを説明する。
【００８９】
本実施例３に係るＸ線検査装置はグリッド位置制御器１４００を有する。グリッド位置制御器１４００は、図示しない位置変動機構を用いてＸ線グリッド３の位置をｘ軸方向に高速に変動することができる。Ｘ線グリッド３のグリッド比は１２：１、グリッド密度は１１７.５［本/ｃｍ］、鉛箔の厚さは４０［μｍ］である。なお、上記位置変動機構は圧電材料等を用いて実現する。
【００９０】
Ｘ線撮影時には、時間間隔Δｔを隔ててパルス状のＸ線が２回放射される。また、上記２回のパルスＸ線の放射に同期して、平面型Ｘ線検出器４が２枚のＸ線撮影像を検出する。一方、Ｘ線透視時には、時間間隔Δｔ毎にパルスＸ線が継続的に放射される。また、上記パルスＸ線の放射に同期して平面型Ｘ線検出器４によるＸ線透視像の検出が継続的に行われる。上記撮影および透視におけるＸ線パルス間隔Δｔの代表例は３３.３［ｍｓ］であるが、６６.６［ｍｓ］とすることも可能である。
【００９１】
グリッド位置制御器１４００は、撮影時および透視時における上記パルスＸ線の放射に同期してＸ線グリッド３のｘ軸方向の位置を変動する。なお、このときの変動量については後述する。
【００９２】
図１５は、本発明の実施例３において投影グリッド密度と平面型Ｘ線検出器４のナイキスト周波数との関係を説明するための図である。本実施例３に係るＸ線検査装置においては、Ｘ線透過像そのものに干渉縞を発生させる。このため、投影グリッド密度f_dをナイキスト周波数f_qの２倍の値に近づけるように設定する。いま、２f_qとf_dとの差をf'_bとすると、Ｘ線透過像中にはエリアシングによって周波数f'_bの干渉縞成分７０２が発生する。
【００９３】
なお、本実施例３においては、Ｘ線グリッド３のグリッド密度を１１７.５［本/ｃｍ］、Ｘ線発生点−グリッド距離ｄを４０［ｃｍ］、Ｘ線発生点−検出器距離Ｄを１００［ｃｍ］とした。従って、（数１）より、投影グリッド密度f_dは４７［本/ｃｍ］となる。一方、画素間隔Δｄが２００［μｍ］であるため、ナイキスト周波数f_qは２５［本/ｃｍ］である。以上より、Ｘ線透過像中に発生する干渉縞の周波数f'_bは、３［本/ｃｍ］となる。
【００９４】
図１６は、本発明の実施例３においてＸ線グリッドの微小変動を説明するための図である。上述ように、本実施例３に係るＸ線検査装置においては、Ｘ線パルス間隔Δｔの間にＸ線グリッド３をｘ軸方向に微小移動させる。そして、その移動量はＸ線グリッド３を構成するＸ線吸収材料２１の周期Δｈによって規定される。
【００９５】
まず、撮影時においてはＸ線パルスが２回照射され、２枚の撮影画像が計測される。このときグリッド位置制御器１４００は、２回の撮影間でＸ線グリッド３の位置がｘ軸方向にΔｈ/２だけシフトするように、Ｘ線グリッド３の位置を制御する。すなわち、１回目の撮影時におけるＸ線グリッド３の位置を図１６（Ａ）とすると、２回目の撮影時におけるＸ線グリッドの位置は図１６（Ｂ）である。
【００９６】
一方、透視時においては、Ｘ線パルスが連続的に発生される。このためグリッド位置制御器１４００は、Ｘ線パルスを発生するたびにＸ線グリッド３の位置が図１６（Ａ）→（Ｂ）→（Ａ）→（Ｂ）→…となるようにＸ線グリッドを振動させる。
【００９７】
上記Ｘ線グリッド３の位置変動に伴い、Ｘ線透過像中に発生する干渉縞の位相がｘ方向に１８０度シフトする。従って、位相が１８０度異なる干渉縞を有する２枚のＸ線透過像を得ることができる。これら２枚のＸ線透過像の差分画像用いて干渉縞を抽出し散乱線補正を行う方法については、実施例１において説明した方法と同一であるため説明を省略する。なお、Ｘ線透過像中の干渉縞成分を取り除くためには、上記２枚のＸ線透過像の加算画像を作成すればよい。本加算によりＸ線透過像中に含まれる干渉縞を相殺し、取り除くことができる。散乱線補正は、上記加算画像に対して行われる。
【００９８】
（実施例４）
図１７は、本発明の第４の実施例に係るＸ線検査装置の構成を説明するための図である。本実施例４は、実施例１で説明した間引きサンプリングに基づく散乱線補正方法を、Ｘ線ＣＴに適用したものである。このため、Ｘ線透過像取得後の散乱線補正処理は実施例１で示した方法と同一であるため省略し、装置構成および撮影方法における相違点のみを説明する。
【００９９】
本実施例４に係るＸ線検査装置は、Ｘ線管１、Ｘ線フィルタ１２、コリメータ２、Ｘ線グリッド３、Ｘ線固体検出器４、被検体６、寝台天板７、回転板１７０１、ガントリ１７００、モニタ９、コンソール１０、画像メモリ１００、画像処理手段１０１、キャリブレーション画像保存メモリ１０２、散乱線画像保存メモリ１０３、画像再構成手段１７０２等から構成される。なお、上記各装置および機構は公知のものを用いる。
【０１００】
以下では、Ｘ線管１、Ｘ線フィルタ１２、コリメータ２、Ｘ線グリッド３、およびＸ線検出器４で構成される系を撮影系と呼ぶ。撮影系は回転板１７０１に固定され、図示しない既知の駆動モータによって回転する。また、以下では、回転板１７０１の回転軸をＺ軸とする。また、回転中心Ｏを原点とする水平および垂直方向の座標軸をそれぞれＸ軸、Ｙ軸とする。回転板１７０１全体はガントリ１７００によって支持されている。
【０１０１】
図１７において、Ｘ線発生点２０と回転板１７０１の回転中心Ｏとの距離は６９［ｃｍ］、Ｘ線発生点２０とＸ線グリッド３の入力面の距離は３２［ｃｍ］、Ｘ線発生点２０とＸ線検出器４の入力面との距離は１０７［ｃｍ］、Ｘ線検出器４の回転中心Ｏを中心とする有効視野は直径４８［ｃｍ］、開口部１７０３の直径は７０［ｃｍ］である。回転板１７０１の１回転のスキャンに要する時間の代表例は０.６秒である。Ｘ線検出器４はセラミックシンチレータ素子から構成される固体検出器であり、ＸＹ平面方向の素子数は８９６チャンネル、Ｚ軸方向の素子数は６４チャンネルである。前記各素子のＸＹ面方向およびＺ軸方向のサイズは、１［ｍｍ］である。また、各素子はＸ線発生点Ｓから略等距離の円弧上に配置される。回転板１７０１の１回転における撮影枚数の代表例は９００枚であり、回転板１７０１の０.４度の回転毎に１回の撮影が行われる。
【０１０２】
撮影時には、Ｘ線検出器４で撮影された撮影画像が、図示しない既知のスリップリング機構を通して順次画像メモリ１００に保存される。画像メモリ１００に記録されたＸ線撮影像は、実施例１で説明した方法と同一の方法を用いて散乱線補正が行われた後に画像再構成手段１７０２に入力される。画像再構成手段１７０２は、公知の画像再構成アルゴリズムを用いて被検体６のＣＴ断層像を作成し、モニタ９に表示する。
【０１０３】
図１８は、本発明の実施例４に係るＸ線検査装置のＸ線グリッド３およびＸ線検出器４の位置関係を説明するための図である。Ｘ線グリッド３は、Ｘ線発生点２０を中心とする円弧上にＸ線吸収材料およびＸ線透過材料が交互に配置され、スリットを形成する。また、上記スリットは、Ｚ軸に平行な方向に配置される。本Ｘ線グリッド３のＸ線吸収材料には鉛、Ｘ線透過材料に紙を使用しているが、これらの材料に限定されるものでなく、例えばタングステンとアルミニウム等でこれらを代用してもよい。Ｘ線グリッド３のグリッド比は１４：１、グリッド密度は２７［本/ｃｍ］、鉛箔の厚さは１００［μｍ］である。また、Ｘ線グリッド３は焦点を有しており、その焦点距離は４０［ｃｍ］である。Ｘ線検出器４の画素は回転面方向の位置をi、回転軸（Ｚ軸）方向の位置をjとするマトリクス状に配置されている。Ｘ線グリッド３のスリットはj方向に投影され、i軸方向の干渉縞を発生する。
【０１０４】
本実施例４では、グリッド密度fgを７０［本/ｃｍ］、Ｘ線発生点−グリッド距離ｄを３２［ｃｍ］、Ｘ線発生点−検出器距離Ｄを１０７［ｃｍ］としたので、（数１）より投影グリッド密度f_dは８［本/ｃｍ］となる。一方、画素間隔Δｄが１［ｍｍ］であるため、ナイキスト周波数f_qは５［本/ｃｍ］である。従って、投影グリッド密度はエリアシング設定されており、そのビート周波数は３［本/ｃｍ］である。本干渉縞成分波形を用いて、実施例１で説明した間引きサンプリングを用いて散乱線補正を行うことができる。
【０１０５】
以上、実施例１〜４に基づいて本発明を具体的に説明した。従来の散乱線補正方法であるＸ線グリッド法では、グリッド比（格子比）の大きなＸ線グリッドを製作するのが物理的に困難であるため、散乱Ｘ線が完全に除去されることはなかった。一例として、グリッド密度６０［本/ｃｍ］、グリッド比１０：１、鉛箔の厚さ５０［μｍ］のＸ線グリッドを用いて、厚さ３０［ｃｍ］の水槽のＸ線撮影を行った場合（管電圧１２０［ｋＶ］）、検出されるＸ線信号に占める散乱Ｘ線成分の割合（散乱線成分比）は約５０％であった。
【０１０６】
これに対して、本発明を用いれば散乱線成分比を約５％程度に抑えることができるため、Ｘ線透過像の画質を大幅に改善することができる。また、従来のＸ線グリッド法ではＸ線グリッドをＸ線検出器と被検体の間に配置するため、被検体の無効被曝が発生するという問題があった。これに対して、本方法ではＸ線グリッドをＸ線源と被検体の間に配置するので上記無効被曝を解消し、上記Ｘ線グリッド法に比べて被検体の被曝量を３０〜４０％減少することが可能である。
【０１０７】
本発明は、上記実施例１〜４のみに限定されるものではなく、その要旨を逸脱しない範囲において種々変更しうることはいうまでもない。
【０１０８】
例えば、本実施例ではＸ線源としてＸ線管を使用したが、放射光などでこれを代用してもよい。また、Ｘ線以外の放射線、可視光線、紫外線、赤外線等に本方法を応用してもよい。さらには、本発明は、Ｘ線ＣＴ装置、Ｘ線透視撮影装置の他に、Ｘ線非破壊検査装置、Ｘ線手荷物検査装置等に適用することも可能である。
【０１０９】
最後に、本発明に基づくＸ線検査方法を以下に示す。
（１）本発明のＸ線検査方法は、Ｘ線を発生するＸ線源と検査対象の間にＸ線グリッドを配置して、前記検査対象に前記Ｘ線を照射して前記検査対象の透過像を２次元Ｘ線検出器により検出する工程と、所定の個数の画素の間隔をおいたサンプリング周波数で、前記Ｘ線グリッドのＸ線吸収体グリッドが配列する方向で前記透過像の画像から画素を抽出し、所定の角度だけ異なる位相をもつ干渉縞がそれぞれ出現する２つの抽出画像を求める工程と、前記２つの抽出画像の間の画素値の差分から干渉縞の振幅の分布を表わす振幅分布画像を求める工程と、前記振幅分布画像を用いて前記透過像の画像に含まれる散乱Ｘ線成分の分布を表わす散乱Ｘ線分布画像を求める工程と、前記透過像の画像の画素値から前記散乱Ｘ線分布画像の画素値を減算して、前記透過像の画像から前記散乱Ｘ線成分を除去する工程と有する。
（２）（１）Ｘ線検査方法において、前記透過像の画像の画素値から、前記透過像の画像に含まれる直接Ｘ線成分の分布を推定するために予め求めらている推定画像を記憶するメモリから読み出した前記推定画像の画素値と前記振幅分布画像の画素値との積を減算して、前記透過像の画像に含まれる散乱Ｘ線成分の分布を表わす散乱Ｘ線分布画像を求める工程と、前記透過像の画像の画素値から前記散乱Ｘ線分布画像の画素値を減算する工程とを有し、前記透過像の画像から前記散乱Ｘ線成分を除去する。
（３）（２）のＸ線検査方法において、前記Ｘ線源としてＸ線管が使用され、前記透過像が検出された時の前記Ｘ線管の管電圧、前記Ｘ線グリッドの組合せに対応する前記推定画像を前記メモリから読み出して、前記散乱Ｘ線成分を除去する。
（４）（２）のＸ線検査方法において、前記所定の個数の画素の間隔をおいた前記サンプリング周波数で、前記Ｘ線吸収体グリッドが配列する方向で、前記検査対象が置かれない状態で検出されたエア画像から画素を抽出し、前記所定の角度だけ異なる位相をもつ干渉縞がそれぞれ出現する２つのエア抽出画像を求める工程と、前記２つのエア抽出画像の間の画素値の差分から干渉縞の振幅の分布を表わすエア振幅分布画像を求める工程と、前記エア画像の前記Ｘ線グリッドを通過した直接Ｘ線成分の分布を表わすエア分布画像の画素値の前記エア振幅分布画像の画素値に対する比を画素値としてもつ前記推定画像を求める工程とを有し、前記推定画像を前記メモリに記憶する。
（５）（２）のＸ線検査方法において、前記Ｘ線源としてＸ線管を使用し、１又は複数の管電圧と、１又は複数の散乱Ｘ線除去率の異なる前記Ｘ線グリッドとの組合せにより、前記検査対象が置かれない状態で検出された１又は複数のエア画像に対して、前記所定の個数の画素の間隔をおいた前記サンプリング周波数で、前記Ｘ線吸収体グリッドが配列する方向で、前記エア画像から画素を抽出し、前記所定の角度だけ異なる位相をもつ干渉縞がそれぞれ出現する２つのエア抽出画像を求める工程と、前記２つのエア抽出画像の間の画素値の差分から干渉縞の振幅の分布を表わすエア振幅分布画像を求める工程と、前記エア画像の前記Ｘ線グリッドを通過した直接Ｘ線成分の分布を表わすエア分布画像の画素値の前記エア振幅分布画像の画素値に対する比を画素値としてもつ前記推定画像を求める工程とを有し、前記各エア画像に対して行ない、前記各エア画像に対する前記推定画像をメモリに記憶する。
（６）（２）のＸ線検査方法において、前記所定の個数が、１、２、３の何れかであることを特徴とするＸ線検査方法。
（７）（２）のＸ線検査方法において、前記所定の角度が、９０°、１２０°、１８０°、２４０°、２７０°の何れかの近傍の角度である。
（８）（２）のＸ線検査方法において、予め求められている同一の前記推定画像を用いて、複数の前記透過像の画像から前記散乱Ｘ線成分を除去する。
（９）（２）のＸ線検査方法において、前記２次元Ｘ線検出器の検出面に於ける前記Ｘ線吸収体グリッドの配列の空間周波数と前記サンプリング周波数との差が、前記サンプリング周波数よりも小さい。
（１０）（２）のＸ線検査方法において、前記２次元Ｘ線検出器の検出面に於ける前記Ｘ線吸収体グリッドの配列の空間周波数と前記Ｘ線検出器のナイキスト周波数との差が、前記ナイキスト周波数の半分の値よりも小さくなるように、前記空間周波数を設定、又は／及び、前記Ｘ線グリッドを配置する位置を設定する。
（１１）（２）のＸ線検査方法において、前記Ｘ線吸収体グリッドが配列する方向で、前記透過像の画素の隣接する複数の画素の画素値を加算して得られる加算画像から、前記所定の個数の画素の間隔をおいた前記サンプリング周波数で画素を抽出して前記抽出画像を求める工程を有する。
（１２）本発明のＸ線検査方法は、Ｘ線を発生するＸ線源と検査対象の間にＸ線グリッドを配置して、前記検査対象に前記Ｘ線を照射して前記検査対象の透過像を２次元Ｘ線検出器により検出する工程と、前記Ｘ線グリッドのＸ線吸収体グリッドが配列する方向で前記透過像の画像から画素を抽出し、所定の角度だけ異なる位相をもつ干渉縞がそれぞれ出現する２つの抽出画像を求める工程と、前記２つの抽出画像から求めた干渉縞の振幅の分布を表わす振幅分布画像を用いて、前記透過像の画像に含まれる散乱Ｘ線成分の分布を表わす散乱Ｘ線分布画像を求める工程と、前記透過像の画像の画素値から前記散乱Ｘ線分布画像の画素値を減算して、前記透過像の画像から前記散乱Ｘ線成分を除去する工程とを有する。
（１３）本発明のＸ線検査方法は、Ｘ線を発生するＸ線源と検査対象の間にＸ線グリッドを配置して、前記検査対象に前記Ｘ線を照射して前記検査対象の透過像を２次元Ｘ線検出器により検出する工程と、ｎを整数として、前記２次元Ｘ線検出器の前記空間サンプリング周波数の（１/ｎ）の空間サンプリング周波数で、前記透過像の画像の画素データをサンプリングして所定の逆転する位相をもつ干渉縞がそれぞれ出現する２つの抽出画像を求める工程と、前記２つの抽出画像の間の画素値の差分により干渉縞の振幅の分布を表わす振幅分布画像を求める工程と、前記振幅分布画像を用いて、前記透過像の画像に含まれる散乱Ｘ線成分の分布を表わす散乱Ｘ線分布画像を求める工程と、前記透過像の画像の画素値から前記散乱Ｘ線分布画像の画素値を減算して、前記透過像の画像から前記散乱Ｘ線成分を除去する。
（１４）（１３）のＸ線検査方法において、ｎは２、３、４の何れかである。
（１５）本発明のＸ線検査方法は、Ｘ線を発生するＸ線源と検査対象との間に配置されたＸ線グリッドを通して前記検査対象に前記Ｘ線を照射する工程と、前記Ｘ線照射によって得られる前記検査対象の透過像をＸ線検出器により検出する工程と、前記検査対象の透過像中に発生する干渉縞の振幅に基づいて、前記透過像に含まれる散乱Ｘ線成分を抽出し、除去する工程とを有することを特徴とする。
【０１１０】
なお、以上のＸ線検査方法は、検査対象を人体とする医用のＸ線検査方法、人体を除く一般の検査対象を対象とする手荷物検査方法等の非破壊検査方法に適用可能であることは言うまでもない。
【０１１１】
【発明の効果】
本発明によれば、Ｘ線透過像中に含まれる散乱線成分を干渉縞の振幅に基づき直接抽出した後に除去するので、高精度の散乱線補正を行うことができる。また、Ｘ線グリッドをＸ線源と被検体の間に配置するため、被検体の無効被曝が解消され、被曝線量を抑えることができる。
【０１１２】
以上より、少ない被曝線量で高画質のＸ線撮影像、Ｘ線透視像、Ｘ線ＣＴ像等を得ることができる。
【図面の簡単な説明】
【図１】本発明の第１の実施例に係るＸ線検査装置の構成を説明するための図。
【図２】散乱Ｘ線除去の原理を説明するための図。
【図３】Ｘ線グリッドおよび平面型Ｘ線検出器の位置関係を説明するための図。
【図４】平面型Ｘ線検出器によるＸ線透過像の検出において、検出画素位置を説明するための図。
【図５】直接Ｘ線成分および散乱Ｘ線成分の空間周波数分布と平面型Ｘ線検出器のナイキスト周波数との関係を説明するための図。
【図６】検出されたＸ線透過像に対する間引きサンプリングを説明するための図。
【図７】間引きサンプリング時におけるナイキスト周波数と直接Ｘ線成分および散乱Ｘ線成分の空間周波数分布との関係を説明するための図。
【図８】投影グリッド密度の別の設定方法を説明するための図。
【図９】干渉縞画像に基づき干渉縞振幅の空間分布を導出する方法を説明するための図。
【図１０】間引きサンプリングにおける別のサンプリング方法を説明するための図。
【図１１】キャリブレーション画像を求める手順を説明するためのブロック図。
【図１２】キャリブレーション画像に基づいて散乱線分布画像を求める手順を説明するためのブロック図。
【図１３】本発明の第２の実施例に係るＸ線検査装置の構成を説明するための図。
【図１４】本発明の第３の実施例に係るＸ線検査装置の構成を説明するための図。
【図１５】本発明の第３の実施例において投影グリッド密度と平面型Ｘ線検出器のナイキスト周波数との関係を説明するための図。
【図１６】本発明の第３の実施例においてＸ線グリッドの微小変動を説明するための図。
【図１７】本発明の第４の実施例に係るＸ線検査装置の構成を説明するための図。
【図１８】本発明の第４の実施例に係るＸ線検査装置のＸ線グリッドおよびＸ線検出器の位置関係を説明するための図。
【符号の説明】
１…Ｘ線管、２…コリメータ、３…Ｘ線グリッド、４…平面型Ｘ線検出器、５…支柱、６…被検体、７…寝台天板、８…寝台、９…モニタ、１０…コンソール、１１…臨床コンソール、１００…画像メモリ、１０１…画像処理手段、１０２…キャリブレーション画像保存メモリ、１０３…散乱線画像保存メモリ、１７００…ガントリ、１７０１…回転板、１７０２…画像再構成手段、１７０３…開口部。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a radiation inspection apparatus using radiation such as X-rays and γ-rays, and more particularly, to a non-destructive inspection and medical X-ray inspection apparatus.
[0002]
[Prior art]
Conventionally, methods and apparatuses for removing various scattered X-ray components have been reported. For example, an X-ray grid having the same or a fraction of a grid frequency as the detection pixel pitch of the X-ray detector is formed on the X-ray detector surface, and the scattered radiation component is incident obliquely on the detector surface. Has been reported (Prior Art 1: "JP-A-9-75332").
[0003]
Also, a method of removing scattered X-ray components using a deconvolution filter based on a point spread function of scattered X-rays measured in advance has been reported (Prior Art 2: “M. Honda, et al. Med. Phys., 20 (1), 56-69 (1993) ").
[0004]
In addition, an apparatus that directly measures scattered X-ray components and removes scattered X-ray components by scanning an X-ray slit arranged in front of the subject has been reported (Prior Art 3: “K. Doi, et al. al., Radiology, 161, 513-518 (1986) ").
[0005]
In addition, a method for obtaining a high-quality reduced image without moire has been reported when wavelet transform is performed on original image data having a periodic structure to obtain a reduced image (Prior Art 4: Japanese Patent Laid-Open No. Hei 10-). No. 031737 ").
[0006]
Furthermore, a method has been reported in which an X-ray grid is arranged between an X-ray tube and a subject to correct the radiation angle dependence of X-ray quality and dose (conventional technique 5: “Japanese Patent Laid-Open No. 2000-245731”. Gazette ").
[0007]
[Problems to be solved by the invention]
The scattered X-ray component occupies 50% or more of the signal component of the detected X-ray transmission image in some cases, and lowers the contrast of the non-destructive inspection apparatus and medical X-ray inspection apparatus.
[0008]
In the prior art 1, since the X-ray transmissive material constituting the X-ray grid blocks not only scattered X-rays but also X-rays directly, the amount of signal detected by the X-ray detector is reduced. Had. In order to increase the detection signal amount, it is necessary to increase the X-ray dose incident on the subject, and as a result, there is a problem that the X-ray exposure amount of the subject increases. In addition, there is a problem that scattered X-rays having a small incident angle to the detector cannot be removed, and that it is difficult to form an X-ray grid on the front surface of the detector.
[0009]
In the prior art 2, the scattered X-ray correction is performed by measuring the point spread function of the scattered X-ray in advance for a uniform subject with a constant tube voltage and a uniform object. The scattered X-ray correction is performed on the assumption that the composition of the subject is uniform. However, in actual photographing, the composition and thickness of the subject change variously, and the tube voltage also changes variously. Therefore, the scattered X-ray component cannot be completely removed, and there is a problem that correction accuracy is poor.
[0010]
In the prior art 3, since imaging is performed while scanning the X-ray slit, there is a problem in that imaging for a short time and X-ray fluoroscopy cannot be performed.
[0011]
Conventional technique 4 obtains a reduced image for high-quality observation, but does not mention removal of scattered X-ray components contained in the original image.
[0012]
In Prior Art 5, since the radiation angle dependency of the X-rays emitted from the rotating anode X-ray tube is reduced, a uniform and high-quality X-ray image can be obtained. There is no mention.
[0013]
An object of the present invention is to provide an X-ray inspection method and an X-ray inspection apparatus capable of obtaining an X-ray image obtained by removing scattered X-ray components from an X-ray transmission image image to be inspected detected by a two-dimensional X-ray detector. Is to provide.
[0014]
Another object of the present invention is to provide an X-ray inspection method and an X-ray inspection apparatus capable of reducing the X-ray exposure amount to be inspected and simultaneously obtaining a high-quality X-ray image.
[0015]
[Means for Solving the Problems]
In order to achieve the above object, in the present invention, an X-ray grid is arranged between an X-ray source and an inspection object, and an X-ray transmission image of the inspection object is detected by a two-dimensional X-ray detector. The X-rays incident on the inspection object are slit-shaped X-rays that have passed through the X-ray grid. The X-ray transmitted through the inspection object includes a slit-like high-frequency component (direct X-ray component) and a non-slit-like low-frequency component (scattered X-ray component) scattered by the inspection object. When an X-ray transmission image to be inspected is detected by a two-dimensional X-ray detector, direct X-ray components generate interference fringes between the two-dimensional X-ray detector and the X-ray grid, but scattered X-ray components are interference fringes. Does not occur.
[0016]
Now, if the position of the X-ray grid is changed by a minute distance in a direction substantially perpendicular to the direction connecting the focal point of the X-ray source and the center of the detection surface of the two-dimensional X-ray detector, the phase of the interference fringe changes. . At this time, the amplitude of the interference fringes directly reflects the X-ray component amount. That is, the signal component that changes with the change in the position of the X-ray grid is a direct line component, and the signal component that does not change corresponds to a scattered X-ray component.
[0017]
As described above, the scattered X-ray component distribution image (hereinafter, scattered X-ray distribution image) included in the X-ray transmission image to be inspected can be obtained from the amplitude of the interference fringes. Further, an X-ray distribution image can be directly obtained by subtracting the scattered X-ray distribution image from the X-ray transmission image. The operation for obtaining the direct X-ray distribution image is hereinafter referred to as scattered X-ray correction.
[0018]
In order to obtain the interference fringe amplitude, it is necessary to separate the measurement image into an interference fringe component image (hereinafter referred to as an interference fringe image) and an image of other components (hereinafter referred to as a non-interference fringe image). In order to obtain an interference fringe image, it is only necessary to detect two X-ray transmission images in which interference fringes having phases that differ by approximately 180 ° appear and create a differential image of the two X-ray transmission images. On the other hand, in order to obtain a non-interference fringe image, it is only necessary to create an addition image of two X-ray transmission images whose phases are approximately 180 ° different from each other. However, in order to cause an interference fringe having a phase difference of approximately 180 ° between two X-ray transmission images, it is necessary to accurately control a minute distance change in the position of the X-ray grid.
[0019]
On the other hand, two X-ray transmission images having a phase difference of approximately 180 ° can be created by the following method. In the following description, the grid density of the X-ray grid is a value in which the direct X-ray component is less likely to cause interference fringes generated between the two-dimensional X-ray detector and the X-ray grid. The direction in which the slits in the X-ray grid are arranged is called a slit arrangement direction.
[0020]
An image extracted from the X-ray transmission image by thinning out every other pixel in the slit arrangement direction (hereinafter referred to as an extracted image) is created. By extracting every other pixel, interference fringes are generated in the extracted image. Two extracted images are created. However, the two extracted images are created by shifting the pixel extraction positions by one pixel in the slit arrangement direction. At this time, interference fringes appearing in the two extracted images have phases that differ by approximately 180 °. Therefore, an interference fringe image can be created by taking the difference between the two extracted images. Further, the scattered X-ray correction can be performed based on the interference fringe image.
[0021]
In the scattered X-ray correction, the X-ray component amount is directly obtained based on the amplitude of the interference fringes. The amplitude of the interference fringes is directly proportional to the amount of X-ray components, and the ratio (hereinafter referred to as amplitude ratio) is obtained in advance by measurement. The amplitude ratio is obtained using an air image obtained in a state where the inspection object is not placed. The air image is one or more air images detected by a combination of one or more tube voltages of the X-ray tube and one or more X-ray grids having different scattered X-ray removal rates. An interference fringe image is created for each air image using the same method as described above. At this time, the amplitude ratio is obtained directly as the ratio of the X-ray component amount to the interference fringe amplitude. However, the direct X-ray component amount can be obtained as the signal amount of the air image itself.
[0022]
The amplitude ratio changes according to the position in the air image. Therefore, the amplitude ratio distribution is created as a calibration image and stored in the memory. To create a calibration image, first, an interference fringe image is created for the air image. Next, an interference fringe amplitude distribution image is created based on the interference fringe image. Finally, an image created by dividing each pixel value of the air image by the pixel value of the amplitude distribution image is taken as a calibration image.
[0023]
The procedure for correcting scattered radiation for an arbitrary X-ray transmission image using a calibration image is as follows. First, an interference fringe image is created for an X-ray transmission image (original image). Next, an interference fringe amplitude distribution image is created based on the interference fringe image. Further, a calibration image is read from the memory, and a product of the calibration image and the amplitude distribution image is subtracted from the X-ray transmission image (original image) to obtain a scattered X-ray distribution image. Finally, the scattered X-ray distribution image is subtracted from the X-ray transmission image (original image) to directly create an X-ray distribution image.
[0024]
When the calibration image is read from the memory, the calibration image corresponding to the combination of the tube voltage and the X-ray grid when the X-ray transmission image is detected is read from the memory.
[0025]
When substantially the same site to be inspected is continuously detected, the scattered X-ray components are removed from the images of the plurality of X-ray transmission images using the same scattered X-ray distribution image at a specific time interval. be able to.
[0026]
As the X-ray grid used in the above description, a parallel grid or a focused grid can be used, and two parallel grids or a cross grid in which focused grids are arranged to overlap each other can be used.
[0027]
As described above, in the present invention, the scattered X-ray correction can be executed with high accuracy by a simple calculation process.
[0028]
DETAILED DESCRIPTION OF THE INVENTION
(Example 1)
FIG. 1 is a diagram for explaining the configuration of an X-ray inspection apparatus according to a first embodiment of the present invention. In FIG. 1, the left-right direction in the plane of the paper is the y axis, the vertical direction in the plane of the paper is the z axis, and the direction perpendicular to the plane of the paper is the x axis.
[0029]
The X-ray inspection apparatus according to the first embodiment includes an X-ray tube 1, a collimator 2, an X-ray grid 3, a planar X-ray detector 4, a column 5, a subject (inspection object) 6, a bed top 7, and a bed. 8, monitor 9, console 10, clinical console 11, image memory 100, image processing means 101, calibration image storage memory 102, scattered radiation image storage memory 103, and the like. In addition, each said apparatus and mechanism use a well-known thing. Hereinafter, a system constituted by the X-ray tube 1, the collimator 2, the X-ray grid 3, and the planar X-ray detector 4 is referred to as an imaging system.
[0030]
The photographing system is fixed to the column 5. The support column 5 can be inclined with respect to the bed top plate 7 by an inclination mechanism (not shown). However, the inclination direction is a direction parallel to the yz plane. By tilting the column 5, the X-ray irradiation direction on the subject 6 can be freely changed. The bed 8 can move the position of the bed top 7 and the subject 6 placed on the bed top 7 in the x, y, and z directions by a known moving mechanism. Further, the position of the X-ray grid 3 in the x, y, and z directions with respect to the column 5 can be adjusted by a position adjusting device (not shown).
[0031]
In FIG. 1, the distance between the X-ray generation point of the X-ray tube 1 and the input surface of the flat X-ray detector 4 is 100 cm. Further, the X-ray input surface of the flat X-ray detector 4 is a square having a side of 204.8 mm. The number of pixels of the planar X-ray detector 4 is 1024 × 1024 pixels, and the size of one side of each pixel is 200 μm. The typical frame rate of the planar X-ray detector 4 at the time of X-ray fluoroscopy is 30 [frames / second], but may be 7.5 or 15 [frames / second]. The planar X-ray detector 4 is a known direct X-ray detector or indirect X-ray detector. Representative examples of the direct type detector and the indirect type detector include “M. Choquette, et al., SPIE Vol. 3977, 128-136 (2000)” (hereinafter referred to as Reference 1) and “Tom JC Bruijns”, respectively. , et al., SPIE Vol.3977, 117-127 (2000) ”(hereinafter, document 2).
[0032]
The distance between the X-ray generation point of the X-ray tube 1 and the input surface of the X-ray grid 3 is 40 cm. The X-ray grid 3 is a known grid, and a large number of slits are formed by alternately arranging an X-ray absorbing material and an X-ray transmitting material. The slit direction of the X-ray grid 3 is formed in only one direction, and the X-ray grid 3 is arranged on the support column 5 so that the slit direction is parallel to the y axis. The present X-ray grid 3 uses lead as an X-ray absorbing material and paper as an X-ray transmitting material, but is not limited to these materials, and for example, tungsten and aluminum may be used instead. . The grid ratio (lattice ratio) is 12: 1, the grid density is 70 [lines / cm], and the thickness of the lead foil is 50 [μm]. The X-ray grid 3 has a focal point, and the focal length is 40 [cm].
[0033]
Next, the operation of the X-ray inspection apparatus according to the first embodiment will be described. The X-ray inspection apparatus acquires an X-ray fluoroscopic image and an X-ray image of the subject 6, removes scattered radiation components contained in these images, and displays them on the monitor 9. The procedure from the X-ray irradiation during X-ray fluoroscopy and X-ray imaging to the display of the scattered radiation correction image is as follows.
[0034]
First, the examiner (operator) instructs the start of fluoroscopy or radiography using the console 10 or the clinical console 11, and emits X-rays from the X-ray tube 1. Next, the X-ray transmitted through the subject 6 is detected by the flat X-ray detector 4 and immediately recorded in the image memory 100. Further, the image processing means 101 extracts only the scattered radiation component in the X-ray transmission image recorded in the image memory 100 and stores it in the scattered radiation image storage memory 103. Finally, the subtractor 104 subtracts the scattered radiation component image stored in the scattered radiation image storage memory 103 from the X-ray transmission image stored in the image memory 100 to create a scattered radiation corrected image and displays it on the monitor 9. . At the time of fluoroscopy, the above series of operations is repeated until the examiner gives an instruction to end fluoroscopy.
[0035]
At this time, the scattered ray component image may be created every time an X-ray fluoroscopic image is input to the image memory 100, but may be created every several frames (usually 2 to 30 frames). When a scattered ray component image is created every several frames, the scattered ray correction is approximately performed using the same scattered ray component image until the next scattered ray component image is created. Hereinafter, such a recent correction method at the time of fluoroscopy is referred to as thinning correction. When the movement of the subject 6 is not so large with respect to the frame rate of the flat X-ray detector 4, thinning correction can be performed. By adopting thinning correction, the processing speed required for the image processing means 101 can be reduced, so that the apparatus configuration can be simplified.
[0036]
For the extraction of the scattered radiation component, a calibration image stored in the calibration image storage memory 102 is used. Details of the scattered radiation component extraction calculation by the image processing means 101 and the calibration image will be described later.
[0037]
FIG. 2 is a diagram for explaining the principle of removing scattered X-rays. 2A is a diagram showing the relationship between the direct X-ray transmitted through the subject 6 and the planar X-ray detector 4, and FIG. 2B is scattered inside the subject 6. It is a figure showing the relationship between the scattered X-rays and the planar X-ray detector 4.
[0038]
First, in FIG. 2A, the uniform X-ray 200 generated from the X-ray generation point 20 is partially blocked by the X-ray absorbing material 21 inside the X-ray grid 3. Therefore, the X-rays that have passed through the X-ray grid 3 become slit-like X-rays 201. Hereinafter, such a slit-shaped X-ray 201 is referred to as a modulated X-ray. The modulated X-ray 201 is a substantially rectangular X-ray beam, and itself has a high-frequency signal component. For this reason, X-rays (direct X-rays) that have passed through the inside of the subject 6 cause interference with the pixels 22 of the planar X-ray detector 4 to generate interference fringes (moire). On the other hand, in FIG. 2B, the X-ray (scattered X-ray) 202 scattered inside the subject 6 loses a high-frequency signal component due to the scattering, and thus does not generate interference fringes with the pixel 22. .
[0039]
From the above, it can be seen that the interference fringes directly reflect only X-rays, and the amplitude of the interference fringes is directly proportional to the amount of X-ray components. Therefore, by obtaining the amplitude of the interference fringes, the detection signal can be separated into the direct line component and the scattered ray component to perform the scattered ray correction. However, in order to perform scattered ray correction, it is necessary to solve the following two problems. That is, obtaining the ratio of the interference fringe amplitude and the direct X-ray component (Problem 1) and obtaining the interference fringe amplitude (Problem 2).
[0040]
Regarding the problem 1, the ratio of the interference fringe amplitude to the direct line component (hereinafter referred to as a calibration ratio) is obtained using an X-ray transmission image (hereinafter referred to as an air image) detected without the subject 6 being placed. Can be sought. That is, since the signal included in the air image is only the direct line component, the calibration ratio can be easily obtained if the amplitude of the interference fringes can be obtained.
[0041]
Regarding the problem 2, for example, the amplitude of the interference fringes can be obtained by observing the phase change of the interference fringes while moving the position of the X-ray grid 3 in the horizontal direction. However, the above method involves technical difficulties that the X-ray grid 3 is moved at high speed and with high accuracy. For this reason, in the first embodiment, a method of measuring the amplitude of interference fringes by thinning sampling of pixels is adopted. Details of the thinning sampling will be described later.
[0042]
FIG. 3 is a diagram for explaining the positional relationship between the X-ray grid 3 and the planar X-ray detector 4. The distance from the X-ray generation point 20 to the input surface of the X-ray grid 3 is represented by d, and the distance from the X-ray generation point 20 to the input surface of the flat X-ray detector 4 is represented by D. In Example 1, the value of d is 40 [cm], and the value of D is 100 [cm]. The slit direction of the X-ray grid 3 is arranged to be parallel to the y-axis direction.
[0043]
The X-ray detector 4 is arranged so that the pixel array is parallel to the x and y directions. The positions of the pixel array in the x and y directions are i and j directions, respectively, and the positions are represented as (i, j) (i, j = 0, 1,..., 1023). In the following description, an acrylic phantom 30 is used as the subject 6 for simplicity. However, the acrylic phantom 30 is composed of an acrylic plate and a lead bar 31. The lead bar 31 is arranged at the center of the acrylic plate so that its longitudinal direction is parallel to the y direction.
[0044]
FIG. 4 is a diagram for explaining a detection pixel position in the detection of an X-ray transmission image by the planar X-ray detector 4. As shown in FIG. 4A, the pixel interval in the i and j directions of the X-ray detector 4 is denoted by Δd. However, in Example 1, Δd is 200 [μm]. During X-ray fluoroscopy and X-ray imaging, X-rays are detected at all pixel positions (i, j). An example of the profile of the detected image is shown in FIG. However, FIG. 4B shows a profile of the acrylic phantom 30 in the i direction. At this time, if the specification and spatial arrangement of the X-ray grid 3 are appropriately set based on a method to be described later, as shown in FIG. be able to. The signal shown in FIG. 4B is obtained by simultaneously detecting a direct X-ray component as shown in FIG. 4C and a scattered X-ray component as shown in FIG. 4D. The scattered radiation correction process corresponds to a process of extracting the direct X-ray component shown in FIG. 4C from the detection signal shown in FIG.
[0045]
FIG. 5 is a diagram for explaining the relationship between the spatial frequency distribution of the direct X-ray component and the scattered X-ray component and the Nyquist frequency of the planar X-ray detector 4. The direct X-ray frequency component 500 includes the frequency characteristics of the modulated X-ray 201, the frequency characteristics of the direct X-ray transmitted through the subject 6, and the frequency response information of the planar X-ray detector 4. Usually, the frequency response of the planar X-ray detector 4 decreases as the frequency increases. The frequency characteristic of the modulated X-ray 201 represents the structure information of the X-ray grid 3 and contains a lot of information regarding the period of the grid.
[0046]
Now the grid density is f _g Assuming [lines / cm], the projected grid density f projected onto the surface of the planar X-ray detector 4 _d Is expressed by the following equation.
[0047]
[Expression 1]

......... (Equation 1)
From (Equation 1), f _d The value of f _g , D, and D can be arbitrarily set by changing the values. In FIG. _d The value of the detector Nyquist frequency f _q An example of setting it slightly lower is shown. Frequency f _d The frequency intensity 502 at _d Nyquist frequency f _q It gets smaller as it approaches. This means that the periodic structure information of the X-ray grid 4 (hereinafter referred to as grid stripes) becomes inconspicuous due to the blur of the planar X-ray detector 4. Where Nyquist frequency f _q And projected grid density f _d And the beat frequency f _b Is defined by the following equation.
[0048]
[Expression 2]

......... (Equation 2)
That is, in order to make the grid stripes included in the X-ray transmission image inconspicuous, f _b Should be as small as possible. On the other hand, the frequency component 501 of scattered X-rays includes only low frequency component information.
[0049]
FIG. 6 is a diagram for explaining thinning sampling for the detected X-ray transmission image. FIG. 7 is a diagram for explaining the relationship between the Nyquist frequency and the spatial frequency distribution of the direct X-ray component and scattered X-ray component at the time of thinning sampling. As described with reference to FIG. 4A, the X-ray transmission image is detected using all the pixels of the planar X-ray detector 4. That is, if the signal amount detected at the pixel position (i, j) is f (i, j), the X-ray transmission image is expressed as f (i, j) (i, j = 0 to N-1). Can do. However, N is the number of pixels in the i and j directions of the planar X-ray detector 4. In the first embodiment, N = 1024. As shown in FIGS. 6A and 6C, thinning sampling is a method of creating an image by extracting pixel values from f (i, j) every other pixel in the i direction. At this time, the following two types of extracted images can be created depending on the pixel extraction position.
[0050]
[Equation 3]

......... (Equation 3)
[0051]
[Expression 4]

......... (Equation 4)
f ₁ (m, n) and f ₂ Nyquist frequency f ′ in m direction of (m, n) _q (= 1 / (4Δd)) is the Nyquist frequency f of f (i, j) _q It becomes half of (= 1 / (2Δd)). That is, f ' _q = F _q / 2. As shown in FIG. 7, the direct X-ray component 500 at the time of thinning sampling is f ′ _q Aliasing occurs by folding the line symmetrically around the center. In FIG. 7, 702 is a folding component by direct aliasing of the X-ray component 500, and 701 is a folding component by aliasing of the scattered X-ray component 501.
[0052]
This results in a projected grid density f _d Component 502 of frequency f _b Appearing on the low frequency side as a beat component 702, generating interference fringes. Therefore, the extracted image f ₁ (m, n) and f ₂ The m-direction profile of (m, n) includes interference fringes (see FIGS. 6B and 6D).
[0053]
As described above, the amplitude of the interference fringes may be measured in order to perform scattered ray correction. However, the extracted image f ₁ (m, n) and f ₂ Since (m, n) includes structure information of the subject 6 in addition to the interference fringe signal, it is difficult to accurately extract the amplitude of the interference fringe component.
[0054]
To accurately extract the amplitude of the interference fringe component, f ₁ (m, n) and f ₂ A difference image s (m, n) of (m, n) is created by the following equation.
[0055]
[Equation 5]

......... (Equation 5)
As shown in (Equation 3) and (Equation 4), f ₁ (m, n) and f ₂ Sampling positions of (m, n) with respect to f (i, j) are shifted by one pixel from each other in the i direction. For this reason, f ₁ (m, n) and f ₂ The interference fringes included in (m, n) are 180 degrees out of phase with each other in the m direction. Accordingly, since the difference image s (m, n) includes only interference fringe components as shown in FIG. 6E, the amplitude of the interference fringes can be easily obtained from s (m, n).
[0056]
In the description using FIGS. 4 to 6, the projected grid density f _d The Nyquist frequency f _q More beat frequency f _b The value was set to a value lower by the amount (hereinafter referred to as non-aliasing setting). At this time, the frequency of the grid stripes appearing in the X-ray transmission image f (i, j) is f _d = (F _q −f _b ) And the extracted image f ₁ (m, n) and f ₂ The frequency of interference fringes appearing in (m, n) is f _b Met. As mentioned above, to make grid stripes inconspicuous, f _b Must be as small as possible. However, at this time, the fringe frequency f _b Becomes smaller, and its position resolution decreases. Thus, f _b Don't be too big or too small. Usually f _b Is preferably set to about 2 to 5 [lines / cm].
[0057]
FIG. 8 is a diagram for explaining another method of setting the projected grid density. For non-aliasing settings, the projected grid density f _d The Nyquist frequency f _q More beat frequency f _b The value was set to a value lower by a minute. In contrast, the projected grid density f _d The Nyquist frequency f _q More beat frequency f _b A method of setting a value higher by the amount (hereinafter referred to as aliasing setting) will be described. When aliasing is set, aliasing also occurs in all pixel sampling. At this time, the projected grid density f _d The frequency component 502 of the _q −f _d ) Component 702 (see FIG. 8A).
[0058]
Frequency f _q −f _b Is the same value as in the non-aliasing setting, but in the aliasing setting, f _q Since the grid stripes are observed as the folded component centered on, the spectral intensity is smaller than in the case of FIG. That is, by setting the aliasing setting, the grid stripes can be made inconspicuous as compared with the case of the non-aliasing setting.
[0059]
On the other hand, even during thinning sampling, the projected grid density f _d The frequency component 502 in FIG. _b Are observed as interference fringe components. However, since the spectrum intensity of the interference fringe component is also smaller than that in the non-aliasing setting, there is a problem that the measurement accuracy of the interference fringe amplitude is lowered. As described above, the setting of the projection grid density can be selected between the aliasing setting and the non-aliasing setting depending on whether priority is given to suppression of grid fringes or improvement of measurement accuracy of interference fringe amplitude.
[0060]
In the first embodiment, the grid density f _g Is 70 [lines / cm], the X-ray generation point-grid distance d is 40 [cm], and the X-ray generation point-detector distance D is 100 [cm]. _d Is 28 [lines / cm]. On the other hand, since the pixel interval Δd is 200 [μm], the Nyquist frequency f _q Is 25 [lines / cm]. Accordingly, the projection grid density is set to aliasing, and the beat frequency is 3 [lines / cm]. The projected grid density f _d Can be set to non-aliasing, for example, d can be set to 31.4 [cm] in the above setting. Projected grid density f _d Is 22 [lines / cm], and interference fringes with a beat frequency of 3 [lines / cm] are generated.
[0061]
FIG. 9 is a diagram for explaining a method of deriving a spatial distribution of interference fringe amplitude based on the interference fringe image. The difference image s (m, n) (m = 0 to N / 2-1, n = 0 to N-1) obtained by (Equation 5) is an interference fringe image 90 including only interference fringe components ( (A) in FIG. The interference fringe amplitude distribution calculation means 1104 is a means for obtaining the interference fringe amplitude distribution image 95 based on the interference fringe image 90, and comprises the following calculation procedure.
[0062]
First, the interference fringe peak detecting means 91 detects the peak position (maximum value and minimum value) of the interference fringe 99 with respect to the m-direction profile of the interference fringe image 90 ((b) in FIG. 9). The peak detection in the m direction is performed for all n (= 0 to N-1) positions. Next, the peak-to-peak pixel interpolation unit 92 connects the local maximum values and the local minimum values adjacent in the m direction with a straight line to obtain an upper limit envelope 96 and a lower limit envelope 97 of the interference fringe 99 ((c) in FIG. 9). ).
[0063]
The interpolation value is obtained not only at a point on the pixel position m (= 0 to N / 2-1) but also at an intermediate position between pixels adjacent in the m direction. By obtaining the interpolation value at the intermediate position of the pixels, the number of pixels in the m direction, which has been N / 2 after thinning sampling, can be returned to N again. The above interpolation is performed for all n (= 0 to N-1) positions. Since the position in the n direction coincides with the position in the j direction of all the sampled images, the pixel matrix can be returned to (i, j) (i, j = 0 to N−1) at the end of interpolation. A difference 98 between the upper limit envelope 96 and the lower limit envelope 97 corresponds to the amplitude of the interference fringe 99.
[0064]
Therefore, next, the amplitude calculation means 93 obtains the difference 98 at all the positions (i, j) (i, j = 0 to N−1), and obtains the interference fringe amplitude distribution at each position ((d in FIG. 9). )). However, since the interference fringe distribution includes a high-frequency component generated by the linear interpolation, a high-frequency component is removed using a two-dimensional LPF (Low Pass Filter) 94 to obtain a final amplitude distribution image A (i, j). ((E) in FIG. 9) is obtained. As an example of the two-dimensional LPF 94, a known moving average process (pixel area for which a moving average is taken: about 2 × 2 pixels to 5 × 5 pixels) or the like can be given.
[0065]
FIG. 10 is a diagram for explaining another sampling method in thinning sampling. Although FIG. 6 shows a method of extracting pixels every other pixel in the i direction as thinning sampling, the extraction interval may generally be extended every k pixels (1 ≦ k). As an example, FIG. 10 shows an example of extracting every two pixels. In this case, there are three kinds of extraction methods as shown in FIGS. 10A to 10C, and every time the sampling position is shifted by one pixel in the i direction, the phase shift of the interference fringes generated by the thinning sampling is 120. Change degrees.
[0066]
That is, the phase shift of the interference fringes between FIGS. 10A and 10B and between FIGS. 10B and 10C is 120 degrees. Further, the phase shift of the interference fringes between FIGS. 10A and 10C is 240 degrees. In either case, when a difference image between the two images is created, the structure information of the subject in the signal can be removed and only the interference fringes can be extracted.
[0067]
Moreover, although the phase of the interference fringe extracted changes according to the phase shift amount between the images which take the difference, the amplitude distribution does not change. That is, no matter which combination of FIGS. 10A to 10C is created, the amplitude distribution image A (finally obtained by the interference fringe amplitude distribution calculating means 1104 in FIG. There is no big difference between i and j). However, in order to accurately remove the structure information of the subject by the difference, the pixel extraction positions between the difference images should be as close as possible. That is, it is desirable to perform the difference using two extracted images with adjacent pixel extraction positions. In general, if thinning sampling is performed every k pixels, the phase shift amount q between two extracted images whose pixel extraction positions are adjacent is expressed by the following equation.
[0068]
[Formula 6]

......... (Equation 6)
The differential image contains X-ray quantum noise, which may degrade the measurement accuracy of the interference fringe amplitude. For this reason, before performing thinning sampling, a moving average process may be performed on the original image f (i, j) (i, j = 0 to N−1) to reduce quantum noise. If the number of pixels taking a moving average in the i and j directions is 2I + 1 and 2J + 1 (I, J = 0, 1, 2,...), respectively, the image f after the moving average _m (i, j) is calculated by the following equation.
[0069]
[Expression 7]

......... (Equation 7)
However, when the moving average in the i direction is performed, the high frequency component is reduced, and the interference fringe intensity in the thinned image is reduced. For this reason, the moving average in the i direction is generally not performed. On the other hand, in the j direction, a moving average is normally taken in a range of about I = 5 to 10.
[0070]
The moving average process, the thinning sampling process, and the amplitude distribution image derivation process have been described on the assumption that the slit direction of the X-ray grid 3 is perpendicular to the j direction. On the other hand, when the slit direction of the X-ray grid 3 is arranged in the i direction, the i direction and the j direction may be switched in all the above processes. When the X-ray grid 3 having a lattice-like slit is used, all the above processes may be performed in both the i direction and the j direction.
[0071]
As described above, the amplitude distribution image of the interference fringes is shown by using the thinning sampling. Hereinafter, a method for correcting the scattered radiation using the amplitude distribution image will be described.
[0072]
The direct line component amount included in a certain pixel value f (i, j) (I = 0 to N−1) in the X-ray transmission image is expressed as f. _d (i, j) then f _d (i, j) is proportional to the amplitude A (i, j) of the interference fringes at the same position. Therefore, the direct line distribution image f can be obtained by obtaining the ratio (calibration ratio) of the interference fringe amplitude and the direct line component amount in advance. _d (i, j) can be obtained.
[0073]
The calibration ratio can be obtained using an air image. In general, since the calibration ratio differs somewhat depending on the detection position on the flat X-ray detector 4, the distribution is obtained as a calibration image. Now, if the calibration image is represented by r (i, j), r (i, j) is obtained by the following equation.
[0074]
[Equation 8]

......... (Equation 8)
Where f _air (i, j) are air images measured without the subject 6 being placed, _air (i, j) is f _air It is an amplitude distribution image obtained for (i, j).
[0075]
At this time, with respect to an arbitrary X-ray transmission image f (i, j), the direct ray component amount can be estimated by the following equation.
[0076]
[Equation 9]

......... (Equation 9)
Here, the calculation result of (Equation 9) is used as an approximate value of the direct line component. _d This is because '(i, j) lacks high-frequency components. The lack of the high-frequency component is caused by the lack of the high-frequency component in the amplitude distribution image A (i, j). The lack of the high-frequency component in A (i, j) to cause. Approximate direct line component f _d 'Accurate direct line distribution image f from (i, j) _d To obtain (i, j), first, the scattered radiation distribution image f _s find (i, j), then f from the original image f (i, j) _s (i, j) may be subtracted. The scattered radiation distribution image is obtained by the following equation.
[0077]
[Expression 10]

......... (Equation 10)
However, g (i, j) is a two-dimensional LPF (Low Pass Filter), and an operation ** represents a two-dimensional convolution integral. In (Equation 10), the calculation result in the right side {} includes a scattered ray component and a high frequency component of a direct ray. For this reason, high-frequency components (direct line components) are removed using LPF, and only scattered radiation components are extracted. As the LPF, a normal digital filter or a moving average process is used. Scattered ray distribution image f _s Once (i, j) is obtained, direct line distribution image f _d (i, j) can be obtained by the following equation.
[0078]
## EQU11 ##

......... (Equation 11)
The method for correcting scattered radiation using the calibration image has been described above. If the change in the calibration ratio in the calibration image is relatively small, it may be approximated as r (i, j) = ro. In this case, for example, an average value of r (i, j) or the like is used for ro. Further, the value of r (i, j) or ro slightly changes depending on the tube voltage of the X-ray tube 1. Therefore, r (i, j) or ro may be measured for several types of tube voltages in advance, and these may be selected according to the tube voltage at the time of photographing or fluoroscopy.
[0079]
FIG. 11 is a block diagram for explaining a procedure for obtaining a calibration image. The calibration image is created by first taking an air image. By taking an air image, the air image f is stored in the image memory 100. _air (i, j) is secured. Next, the moving average calculation means 1100 is used to perform a moving average in the i and j directions. Note that (Equation 7) is used to calculate the moving average.
[0080]
Next, using the thinned image creating means 1101 and 1102, the extracted image f ₁ (m, n) and f ₂ Create (m, n) respectively. In addition, (Formula 3) and (Formula 4) are used for creation of an extraction image. Next, the subtractor 1103 extracts the extracted image f. ₁ (m, n) and f ₂ The difference of (m, n) is taken and a difference image s (m, n) is created. Next, the interference fringe amplitude distribution calculating means 1104 uses the amplitude distribution image A for the air image. _air (i, j) is created. The calculation by the interference fringe amplitude distribution calculation means 1104 has been described in detail with reference to FIG.
[0081]
Next, an amplitude distribution image A is obtained by a divider 1105. _air (i, j) and f _air The ratio of (i, j) is calculated and a calibration image r (i, j) is created. Finally, the calibration image r (i, j) is stored in the calibration image storage memory 102 together with the tube voltage information input from the console 10.
[0082]
FIG. 12 is a block diagram for explaining a procedure for obtaining a scattered radiation distribution image based on a calibration image. First, an X-ray fluoroscopic image or a captured image f (i, j) is secured in the image memory 100. The subsequent processes from moving average calculation means 1100 to interference fringe amplitude distribution calculation means 1104 are the same as the calibration image creation means shown in FIG. The amplitude distribution image A (i, j) is created by the above process.
[0083]
Next, the integrator 1200 integrates the amplitude distribution image A (i, j) and the calibration image r (i, j) stored in the calibration image storage memory 102 to directly calculate the line component approximate image f ′. _d Create (i, j). This integration corresponds to the above calculation (Equation 9). In the above integration, a calibration image r (i, j) created with a tube voltage closest to the tube voltage when the original image f (i, j) was created is selected. Next, the subtractor 1201 directly calculates the line component approximate image f ′ from the original image f (i, j). _d (i, j) is subtracted, and then the LPF 1202 cuts off the high-frequency component of the subtracted image, and the scattered radiation component image f _s Create (i, j). A series of operations by the subtractor 1201 and the LPF 1202 corresponds to (Equation 10). Finally, the scattered radiation component image f _s (i, j) is stored in the scattered radiation image storage memory 103.
[0084]
As described above, in the first embodiment, the amplitude distribution of interference fringes is measured based on two extracted images created by thinning sampling, and scattered ray correction is performed. Since interference fringes can be accurately obtained from a single X-ray transmission image, high-speed and high-accuracy scattered ray correction can be performed.
[0085]
(Example 2)
FIG. 13 is a diagram for explaining the configuration of the X-ray inspection apparatus according to the second embodiment of the present invention.
[0086]
Hereinafter, only differences from the X-ray inspection apparatus according to the first embodiment will be described. In the X-ray inspection apparatus according to the first embodiment, the X-ray tube 1, the collimator 2, and the X-ray grid 3 are on the upper surface of the subject 6. Whereas the X-ray inspection apparatus according to the second embodiment is arranged (hereinafter referred to as an overtube configuration), the X-ray tube 1, the collimator 2, and the X-ray grid 3 are all on the lower surface of the bed top plate 7. (Hereinafter referred to as an undertube configuration).
[0087]
In the case of the overtube configuration, the distance between the X-ray grid 3 and the subject 6 is narrow, and thus there is a high possibility that the subject 6 is in contact with the X-ray grid 3 by mistake. By doing so, it is possible to avoid the danger that the X-ray grid 3 will accidentally contact the subject 6.
[0088]
(Example 3)
FIG. 14 is a diagram for explaining the configuration of the X-ray inspection apparatus according to the third embodiment of the present invention. In the third embodiment, the interference fringe amplitude distribution is measured without using thinning sampling. Only differences from the X-ray inspection apparatus according to the first embodiment will be described below.
[0089]
The X-ray inspection apparatus according to the third embodiment includes a grid position controller 1400. The grid position controller 1400 can change the position of the X-ray grid 3 at high speed in the x-axis direction using a position changing mechanism (not shown). The X-ray grid 3 has a grid ratio of 12: 1, a grid density of 117.5 [lines / cm], and a lead foil thickness of 40 [μm]. The position variation mechanism is realized using a piezoelectric material or the like.
[0090]
At the time of X-ray imaging, pulsed X-rays are emitted twice with a time interval Δt. In addition, the planar X-ray detector 4 detects two X-ray images in synchronism with the two pulse X-ray emission. On the other hand, during X-ray fluoroscopy, pulsed X-rays are continuously emitted every time interval Δt. In addition, detection of an X-ray fluoroscopic image by the planar X-ray detector 4 is continuously performed in synchronization with the emission of the pulse X-ray. A typical example of the X-ray pulse interval Δt in the above imaging and fluoroscopy is 33.3 [ms], but 66.6 [ms] can also be used.
[0091]
The grid position controller 1400 varies the position of the X-ray grid 3 in the x-axis direction in synchronization with the emission of the pulse X-rays at the time of imaging and fluoroscopy. The amount of change at this time will be described later.
[0092]
FIG. 15 is a diagram for explaining the relationship between the projected grid density and the Nyquist frequency of the planar X-ray detector 4 in the third embodiment of the present invention. In the X-ray inspection apparatus according to the third embodiment, interference fringes are generated in the X-ray transmission image itself. Therefore, the projected grid density f _d The Nyquist frequency f _q It is set to be close to twice the value. Now 2f _q And f _d And the difference from f ' _b Then, in the X-ray transmission image, the frequency f ′ is obtained by aliasing. _b Interference fringe component 702 is generated.
[0093]
In the third embodiment, the X-ray grid 3 has a grid density of 117.5 [lines / cm], an X-ray generation point-grid distance d of 40 [cm], and an X-ray generation point-detector distance D. 100 [cm]. Therefore, from (Equation 1), the projected grid density f _d Is 47 [lines / cm]. On the other hand, since the pixel interval Δd is 200 [μm], the Nyquist frequency f _q Is 25 [lines / cm]. From the above, the frequency f ′ of the interference fringes generated in the X-ray transmission image _b Is 3 [lines / cm].
[0094]
FIG. 16 is a diagram for explaining minute fluctuations in the X-ray grid in the third embodiment of the present invention. As described above, in the X-ray inspection apparatus according to the third embodiment, the X-ray grid 3 is slightly moved in the x-axis direction during the X-ray pulse interval Δt. The amount of movement is defined by the period Δh of the X-ray absorbing material 21 constituting the X-ray grid 3.
[0095]
First, at the time of imaging, X-ray pulses are irradiated twice, and two captured images are measured. At this time, the grid position controller 1400 controls the position of the X-ray grid 3 so that the position of the X-ray grid 3 is shifted by Δh / 2 in the x-axis direction between two imaging operations. That is, assuming that the position of the X-ray grid 3 at the first imaging is FIG. 16A, the position of the X-ray grid at the second imaging is FIG. 16B.
[0096]
On the other hand, during fluoroscopy, X-ray pulses are continuously generated. For this reason, the grid position controller 1400 causes the X-ray grid 3 so that the position of the X-ray grid 3 becomes as shown in FIG. 16 (A) → (B) → (A) → (B) →. Vibrate.
[0097]
As the position of the X-ray grid 3 changes, the phase of interference fringes generated in the X-ray transmission image is shifted by 180 degrees in the x direction. Accordingly, two X-ray transmission images having interference fringes that are 180 degrees out of phase can be obtained. The method of extracting interference fringes using the difference image of these two X-ray transmission images and correcting the scattered radiation is the same as the method described in the first embodiment, and thus the description thereof is omitted. In order to remove the interference fringe component in the X-ray transmission image, an addition image of the two X-ray transmission images may be created. By this addition, interference fringes included in the X-ray transmission image can be canceled and removed. Scattered ray correction is performed on the added image.
[0098]
(Example 4)
FIG. 17 is a diagram for explaining the configuration of an X-ray inspection apparatus according to the fourth embodiment of the present invention. In the fourth embodiment, the scattered radiation correction method based on the thinning sampling described in the first embodiment is applied to the X-ray CT. For this reason, the scattered radiation correction processing after acquisition of the X-ray transmission image is the same as the method shown in the first embodiment, and thus will be omitted, and only differences in the apparatus configuration and the imaging method will be described.
[0099]
The X-ray inspection apparatus according to the fourth embodiment includes an X-ray tube 1, an X-ray filter 12, a collimator 2, an X-ray grid 3, an X-ray solid state detector 4, a subject 6, a bed top 7, a rotating plate 1701, A gantry 1700, a monitor 9, a console 10, an image memory 100, an image processing unit 101, a calibration image storage memory 102, a scattered radiation image storage memory 103, an image reconstruction unit 1702, and the like. In addition, each said apparatus and mechanism use a well-known thing.
[0100]
Hereinafter, a system including the X-ray tube 1, the X-ray filter 12, the collimator 2, the X-ray grid 3, and the X-ray detector 4 is referred to as an imaging system. The imaging system is fixed to a rotating plate 1701 and rotated by a known drive motor (not shown). Hereinafter, the rotation axis of the rotating plate 1701 is referred to as a Z-axis. Also, the horizontal and vertical coordinate axes with the rotation center O as the origin are taken as an X axis and a Y axis, respectively. The entire rotating plate 1701 is supported by a gantry 1700.
[0101]
In FIG. 17, the distance between the X-ray generation point 20 and the rotation center O of the rotating plate 1701 is 69 [cm], the distance between the X-ray generation point 20 and the input surface of the X-ray grid 3 is 32 [cm], and the X-ray generation occurs. The distance between the point 20 and the input surface of the X-ray detector 4 is 107 [cm], the effective visual field around the rotation center O of the X-ray detector 4 is 48 [cm], and the diameter of the opening 1703 is 70 [cm]. cm]. A typical example of the time required for one rotation of the rotating plate 1701 is 0.6 seconds. The X-ray detector 4 is a solid state detector composed of ceramic scintillator elements. The number of elements in the XY plane direction is 896 channels, and the number of elements in the Z-axis direction is 64 channels. The size of each element in the XY plane direction and the Z-axis direction is 1 [mm]. Each element is arranged on an arc that is substantially equidistant from the X-ray generation point S. A typical example of the number of shots in one rotation of the rotating plate 1701 is 900, and one shooting is performed every rotation of the rotating plate 1701 by 0.4 degrees.
[0102]
At the time of imaging, captured images captured by the X-ray detector 4 are sequentially stored in the image memory 100 through a known slip ring mechanism (not shown). The X-ray image recorded in the image memory 100 is input to the image reconstruction unit 1702 after being subjected to scattered radiation correction using the same method as described in the first embodiment. The image reconstruction unit 1702 creates a CT tomographic image of the subject 6 using a known image reconstruction algorithm and displays it on the monitor 9.
[0103]
FIG. 18 is a diagram for explaining the positional relationship between the X-ray grid 3 and the X-ray detector 4 of the X-ray inspection apparatus according to Embodiment 4 of the present invention. In the X-ray grid 3, X-ray absorbing material and X-ray transmitting material are alternately arranged on an arc centered on the X-ray generation point 20 to form a slit. The slit is disposed in a direction parallel to the Z axis. The X-ray absorbing material of the present X-ray grid 3 uses lead and paper for the X-ray transmitting material, but is not limited to these materials. For example, tungsten and aluminum may be used instead. Good. The X-ray grid 3 has a grid ratio of 14: 1, a grid density of 27 [lines / cm], and a lead foil thickness of 100 [μm]. The X-ray grid 3 has a focal point, and the focal length is 40 [cm]. The pixels of the X-ray detector 4 are arranged in a matrix having a position in the rotation plane direction i and a position in the rotation axis (Z-axis) direction j. The slits of the X-ray grid 3 are projected in the j direction and generate interference fringes in the i-axis direction.
[0104]
In the fourth embodiment, the grid density fg is 70 [lines / cm], the X-ray generation point-grid distance d is 32 [cm], and the X-ray generation point-detector distance D is 107 [cm]. Projected grid density f from Equation 1) _d Is 8 [lines / cm]. On the other hand, since the pixel interval Δd is 1 [mm], the Nyquist frequency f _q Is 5 [lines / cm]. Accordingly, the projection grid density is set to aliasing, and the beat frequency is 3 [lines / cm]. Using this interference fringe component waveform, the scattered radiation correction can be performed using the thinning sampling described in the first embodiment.
[0105]
The present invention has been specifically described above based on Examples 1 to 4. In the X-ray grid method, which is a conventional scattered radiation correction method, it is physically difficult to manufacture an X-ray grid having a large grid ratio (lattice ratio), and thus scattered X-rays are not completely removed. It was. As an example, an X-ray image of a water tank having a thickness of 30 [cm] was performed using an X-ray grid having a grid density of 60 [lines / cm], a grid ratio of 10: 1, and a lead foil thickness of 50 [μm]. In the case (tube voltage 120 [kV]), the ratio of the scattered X-ray component to the detected X-ray signal (scattered ray component ratio) was about 50%.
[0106]
On the other hand, if the present invention is used, the scattered radiation component ratio can be suppressed to about 5%, so that the image quality of the X-ray transmission image can be greatly improved. Further, the conventional X-ray grid method has a problem in that the subject is subject to invalid exposure because the X-ray grid is disposed between the X-ray detector and the subject. On the other hand, in this method, the X-ray grid is arranged between the X-ray source and the subject, so that the invalid exposure is eliminated, and the subject exposure is reduced by 30 to 40% compared to the X-ray grid method. Is possible.
[0107]
It goes without saying that the present invention is not limited to only the first to fourth embodiments, and various modifications can be made without departing from the scope of the invention.
[0108]
For example, in the present embodiment, an X-ray tube is used as an X-ray source, but this may be substituted by radiated light or the like. Further, the present method may be applied to radiation other than X-rays, visible light, ultraviolet light, infrared light, and the like. Furthermore, the present invention can be applied to an X-ray non-destructive inspection apparatus, an X-ray baggage inspection apparatus, etc. in addition to an X-ray CT apparatus and an X-ray fluoroscopic apparatus.
[0109]
Finally, an X-ray inspection method based on the present invention will be described below.
(1) In the X-ray inspection method of the present invention, an X-ray grid is disposed between an X-ray source that generates X-rays and an inspection object, and the inspection object is irradiated with the X-ray to transmit the inspection object. Pixels from the transmitted image in a direction in which the X-ray absorber grid of the X-ray grid is arranged at a sampling frequency with a predetermined number of pixel intervals and a step of detecting an image with a two-dimensional X-ray detector Extracting two extracted images in which interference fringes having phases different from each other by a predetermined angle appear, and an amplitude distribution representing an amplitude distribution of the interference fringes from a difference in pixel values between the two extracted images A step of obtaining an image, a step of obtaining a scattered X-ray distribution image representing a distribution of a scattered X-ray component included in the image of the transmission image using the amplitude distribution image, and the scattering from a pixel value of the image of the transmission image. Decrease pixel value of X-ray distribution image To have a step of removing the scattered X-ray component from the image of the transmitted image.
(2) (1) In the X-ray inspection method, an estimated image obtained in advance for estimating a direct X-ray component distribution included in the transmission image from a pixel value of the transmission image is stored. A scattered X-ray distribution image representing a distribution of scattered X-ray components included in the transmitted image is obtained by subtracting the product of the pixel value of the estimated image read from the memory to be read and the pixel value of the amplitude distribution image. A step of subtracting a pixel value of the scattered X-ray distribution image from a pixel value of the image of the transmission image, and removing the scattered X-ray component from the image of the transmission image.
(3) In the X-ray inspection method of (2), an X-ray tube is used as the X-ray source, and it corresponds to a combination of the tube voltage of the X-ray tube and the X-ray grid when the transmission image is detected. The estimated image to be read is read from the memory, and the scattered X-ray component is removed.
(4) In the X-ray inspection method of (2), the inspection object is not placed in the direction in which the X-ray absorber grid is arranged at the sampling frequency with the predetermined number of pixels being spaced apart. Extracting pixels from the detected air image, obtaining two air extracted images in which interference fringes having phases different from each other by the predetermined angle appear, and a difference in pixel values between the two air extracted images A step of obtaining an air amplitude distribution image representing an amplitude distribution of interference fringes, and a pixel value of the air amplitude distribution image of a pixel value of an air distribution image representing a distribution of a direct X-ray component that has passed through the X-ray grid of the air image. Obtaining the estimated image having a ratio to a value as a pixel value, and storing the estimated image in the memory.
(5) In the X-ray inspection method of (2), an X-ray tube is used as the X-ray source, and one or more tube voltages and one or more X-ray grids having different scattered X-ray removal rates By the combination, the X-ray absorber grid is arranged at the sampling frequency with the predetermined number of pixels spaced from one or a plurality of air images detected in a state where the inspection object is not placed. Extracting a pixel from the air image in a direction and obtaining two air extracted images in which interference fringes having phases different from each other by the predetermined angle appear, and a pixel value difference between the two air extracted images Obtaining an air amplitude distribution image representing the amplitude distribution of the interference fringes from the air amplitude distribution image, and a pixel value of the air distribution image representing the distribution of the direct X-ray component passing through the X-ray grid of the air image. Pixel And a step of determining the estimated image with a ratio as the pixel value for the performed for each air image, storing said estimated image for said each air image in the memory.
(6) The X-ray inspection method according to (2), wherein the predetermined number is any one of 1, 2, and 3.
(7) In the X-ray inspection method of (2), the predetermined angle is an angle in the vicinity of any of 90 °, 120 °, 180 °, 240 °, and 270 °.
(8) In the X-ray inspection method of (2), the scattered X-ray component is removed from a plurality of images of the transmission image using the same estimated image obtained in advance.
(9) In the X-ray inspection method of (2), the difference between the sampling frequency and the spatial frequency of the array of the X-ray absorber grids on the detection surface of the two-dimensional X-ray detector is greater than the sampling frequency. Is also small.
(10) In the X-ray inspection method of (2), the difference between the spatial frequency of the X-ray absorber grid arrangement on the detection surface of the two-dimensional X-ray detector and the Nyquist frequency of the X-ray detector is The spatial frequency is set and / or the position where the X-ray grid is arranged is set so as to be smaller than half the value of the Nyquist frequency.
(11) In the X-ray inspection method of (2), from the addition image obtained by adding pixel values of a plurality of adjacent pixels of the transmission image in the direction in which the X-ray absorber grid is arranged, Extracting the pixels at the sampling frequency with a predetermined number of pixel intervals therebetween to obtain the extracted image.
(12) In the X-ray inspection method of the present invention, an X-ray grid is arranged between an X-ray source that generates X-rays and an inspection object, and the inspection object is irradiated with the X-ray to transmit the inspection object. A step of detecting an image by a two-dimensional X-ray detector, and extracting fringes from the image of the transmission image in a direction in which the X-ray absorber grid of the X-ray grid is arranged, and interference fringes having phases different by a predetermined angle Using the amplitude distribution image representing the distribution of the amplitude of the interference fringes obtained from the two extracted images and the distribution of the scattered X-ray components included in the transmitted image. And a step of subtracting a pixel value of the scattered X-ray distribution image from a pixel value of the transmitted image to remove the scattered X-ray component from the transmitted image. And have.
(13) In the X-ray inspection method of the present invention, an X-ray grid is disposed between an X-ray source that generates X-rays and an inspection object, and the inspection object is irradiated with the X-ray to transmit the inspection object. A step of detecting an image by a two-dimensional X-ray detector, and pixels of the transmission image at a spatial sampling frequency (1 / n) of the spatial sampling frequency of the two-dimensional X-ray detector, where n is an integer Sampling two data to obtain two extracted images each having an interference fringe having a predetermined reversal phase, and an amplitude distribution representing an amplitude distribution of the interference fringes by a difference in pixel values between the two extracted images A step of obtaining an image, a step of obtaining a scattered X-ray distribution image representing a distribution of a scattered X-ray component included in the image of the transmission image using the amplitude distribution image, and the pixel value of the image of the transmission image Of scattered X-ray distribution image Pixel values are subtracted to remove the scattered X-ray component from the transmitted image.
(14) In the X-ray inspection method of (13), n is 2, 3, or 4.
(15) In the X-ray inspection method of the present invention, the X-ray is irradiated to the inspection object through an X-ray grid disposed between the X-ray source generating the X-ray and the inspection object, and the X-ray Based on the step of detecting a transmission image of the inspection object obtained by irradiation with an X-ray detector and the amplitude of interference fringes generated in the transmission image of the inspection object, the scattered X-ray component contained in the transmission image is detected. And extracting and removing.
[0110]
It should be noted that the above X-ray inspection methods can be applied to non-destructive inspection methods such as medical X-ray inspection methods in which the inspection object is a human body and baggage inspection methods for general inspection objects other than the human body. Needless to say.
[0111]
【The invention's effect】
According to the present invention, since the scattered radiation component contained in the X-ray transmission image is directly extracted based on the interference fringe amplitude and then removed, highly accurate scattered radiation correction can be performed. In addition, since the X-ray grid is disposed between the X-ray source and the subject, invalid exposure of the subject is eliminated, and the exposure dose can be suppressed.
[0112]
As described above, a high-quality X-ray image, X-ray fluoroscopic image, X-ray CT image, etc. can be obtained with a small exposure dose.
[Brief description of the drawings]
FIG. 1 is a diagram for explaining a configuration of an X-ray inspection apparatus according to a first embodiment of the present invention.
FIG. 2 is a view for explaining the principle of removing scattered X-rays.
FIG. 3 is a diagram for explaining the positional relationship between an X-ray grid and a planar X-ray detector.
FIG. 4 is a diagram for explaining a detection pixel position in detection of an X-ray transmission image by a planar X-ray detector.
FIG. 5 is a diagram for explaining the relationship between the spatial frequency distribution of direct X-ray components and scattered X-ray components and the Nyquist frequency of a planar X-ray detector.
FIG. 6 is a diagram for explaining thinning sampling for a detected X-ray transmission image;
FIG. 7 is a diagram for explaining the relationship between the Nyquist frequency and the spatial frequency distribution of direct X-ray components and scattered X-ray components during thinning sampling.
FIG. 8 is a diagram for explaining another method of setting the projected grid density.
FIG. 9 is a diagram for explaining a method of deriving a spatial distribution of interference fringe amplitude based on an interference fringe image.
FIG. 10 is a diagram for explaining another sampling method in thinning sampling.
FIG. 11 is a block diagram for explaining a procedure for obtaining a calibration image;
FIG. 12 is a block diagram for explaining a procedure for obtaining a scattered radiation distribution image based on a calibration image.
FIG. 13 is a diagram for explaining a configuration of an X-ray inspection apparatus according to a second embodiment of the present invention.
FIG. 14 is a diagram for explaining the configuration of an X-ray inspection apparatus according to a third embodiment of the present invention.
FIG. 15 is a diagram for explaining the relationship between the projected grid density and the Nyquist frequency of the planar X-ray detector in the third embodiment of the present invention.
FIG. 16 is a view for explaining minute fluctuations of the X-ray grid in the third embodiment of the present invention.
FIG. 17 is a view for explaining the configuration of an X-ray inspection apparatus according to a fourth embodiment of the present invention.
FIG. 18 is a diagram for explaining the positional relationship between an X-ray grid and an X-ray detector of an X-ray inspection apparatus according to a fourth embodiment of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... X-ray tube, 2 ... Collimator, 3 ... X-ray grid, 4 ... Planar X-ray detector, 5 ... Support | pillar, 6 ... Subject, 7 ... Bed top plate, 8 ... Bed, 9 ... Monitor, 10 ... Console: 11 Clinical console 100 Image memory 101 Image processing means 102 Calibration image storage memory 103 Scattered ray image storage memory 1700 Gantry 1701 Rotary plate 1702 Image reconstruction means 1703: Opening.

Claims (7)

An X-ray source that generates X-rays, a two-dimensional X-ray detector that detects a transmission image from the X-rays that have passed through the inspection object, and an X-ray grid that is disposed between the X-ray source and the inspection object Arithmetic processing means for performing arithmetic processing on the output of the two-dimensional X-ray detector, wherein the arithmetic processing means uses the X-rays of the X-ray grid at a sampling frequency with a predetermined number of pixel intervals. A pixel is extracted from the image of the transmission image in the direction in which the absorber grid is arranged, and an operation for obtaining two extraction images in which interference fringes having phases different from each other by a predetermined angle appear, and between the two extraction images, An operation for obtaining an amplitude distribution image representing the distribution of amplitude of interference fringes from the difference in pixel values, and a scattered X-ray distribution image representing the distribution of scattered X-ray components contained in the transmitted image using the amplitude distribution image. The obtained calculation and the transmission image By subtracting the pixel values of the scattered X-ray distribution image from the pixel values of the image, X-rays inspection apparatus characterized by the image of the transmitted image to perform the operation of removing the scattered X-ray component. An X-ray source that generates X-rays, a two-dimensional X-ray detector that detects a transmission image from the X-rays that have passed through the inspection object, and an X-ray grid that is disposed between the X-ray source and the inspection object Arithmetic processing means for performing arithmetic processing of the output of the two-dimensional X-ray detector, the arithmetic processing means from the image of the transmission image in the direction in which the X-ray absorber grid of the X-ray grid is arranged Extracting pixels and calculating two extracted images in which interference fringes having phases different from each other by a predetermined angle appear, and using an amplitude distribution image representing the distribution of amplitudes of the interference fringes obtained from the two extracted images Calculating the scattered X-ray distribution image representing the distribution of scattered X-ray components included in the transmitted image, and subtracting the pixel value of the scattered X-ray distribution image from the pixel value of the transmitted image; The scattered X-ray component from the image of the transmission image X-ray inspection apparatus and carrying out the operation for removal. An X-ray source for generating X-rays, a two-dimensional X-ray detector for detecting a transmission image from the X-rays transmitted through the inspection object at a preset spatial sampling frequency, and between the X-ray source and the inspection object And an arithmetic processing means for performing arithmetic processing of the output of the two-dimensional X-ray detector, wherein the arithmetic processing means uses n as an integer, and the arithmetic processing means of the two-dimensional X-ray detector An operation of sampling pixel data of the image of the transmission image at a spatial sampling frequency of (1 / n) of the spatial sampling frequency to obtain two extracted images in which interference fringes having predetermined reverse phases appear, An operation for obtaining an amplitude distribution image representing an amplitude distribution of interference fringes based on a pixel value difference between the two extracted images, and using the amplitude distribution image, a scattered X-ray component included in the image of the transmission image Distribution An operation for obtaining a scattered X-ray distribution image, and an operation for subtracting a pixel value of the scattered X-ray distribution image from a pixel value of the image of the transmission image to remove the scattered X-ray component from the image of the transmission image. An X-ray inspection apparatus characterized by   An X-ray source that generates X-rays, an X-ray detector that detects a transmission image from the X-rays that have passed through the inspection object, an X-ray grid disposed between the X-ray source and the inspection object, X-ray characterized by comprising: arithmetic processing means for extracting and removing scattered X-ray components contained in the transmission image based on the amplitude of interference fringes generated in the transmission image of the inspection object Inspection device.   The X-ray inspection apparatus according to claim 4, further comprising an inspection target support unit that supports the inspection target, wherein the inspection target is disposed on an upper surface of the inspection target support unit, and the lower surface of the inspection target support unit An X-ray inspection apparatus, wherein an X-ray source is disposed, and the X-ray grid is disposed between the X-ray source and a lower surface of the inspection object support means.   5. The X-ray inspection apparatus according to claim 4, further comprising an imaging system rotating unit configured to rotate an imaging system including a pair of the X-ray source and the two-dimensional X-ray detector around the inspection target, The X-ray grid is arranged between the X-ray source and the inspection object, and the imaging system rotation means rotates the imaging system while maintaining the position of the X-ray grid with respect to the imaging system substantially constant, An X-ray inspection apparatus, wherein a plurality of X-ray transmission images are taken from a plurality of directions with respect to an inspection object, and an X-ray tomographic image of the inspection object is generated and displayed.   The X-ray inspection apparatus according to claim 4, wherein the X-ray source and the X-ray grid are arranged on an upper surface of the inspection target.

Images