JP4344038B2 - PET equipment - Google Patents

Description

なる式で表される。ここで、Ｄは検出器リングの内径、ｄは光子検出器の軸方向の幅、ε(δ)はダイレクトプレーンに対して傾斜する投影の相対的な検出感度、δ_maxは２次元同時計数情報の蓄積における最大リング差である。
【００２２】
一方、３次元同時計数情報の蓄積における軸方向視野角をφ_3Dとし、全ての光子検出器の個数Ｎに対するスライスセプタ２０の後方に存在する光子検出器の個数ｎの比をｓ（＝ｎ／Ｎ）とする。このとき、３次元同時計数情報および２次元同時計数情報それぞれの検出感度の比Ｒは、
【００２３】
【数２】

なる近似式で表される。ここで、ｗはスライスセプタ２０の径方向の幅である。なお、この式の導出に際しては、光子の吸収および散乱の効果を無視した。例えば、φ_2D＝１°、φ_3D＝１０°、ｓ＝１／８、ｄ＝６ｍｍ、ｗ＝１ｍｍとすると、Ｒ＝３６となる。
【００２４】
図２は、本実施形態に係るＰＥＴ装置の全体構成を概念的に説明するブロック図である。セプタ回転駆動部３０は、中心軸を中心にしてスライスセプタ３０を回転駆動させるものであり、回転位置検出センサ４０は、スライスセプタ２０の回転位置を検出するものである。測定空間１に被検体２をおいて行う１回の測定の期間中、スライスセプタ２０はセプタ回転駆動部３０により駆動されて回転しており、また、スライスセプタ２０の回転位置は回転位置検出センサ４０により常に把握されている。そして、１対の光子検出器が光子対を同時計数したときに、その１対の光子検出器のうち少なくとも一方がスライスセプタ２０の後方に位置するか否かが判定される。この判定は、回転位置検出センサ４０により検出されたスライスセプタ２０の回転位置に基づいてなされる。
【００２５】
もし、一方の光子検出器がスライスセプタ２０の後方に位置すると判定されれば、その１対の光子検出器が検出した同時計数情報は２次元同時計数情報であると判断され、その２次元同時計数情報は２次元投影データ蓄積部５１に蓄積される。一方、そうでなければ、その１対の光子検出器が検出した同時計数情報は３次元同時計数情報であると判断され、その３次元同時計数情報は３次元投影データ蓄積部５２に蓄積される。このようにして、２次元同時計数情報および３次元同時計数情報それぞれは互いに個別に蓄積されてヒストグラムが作成される。以下では、２次元同時計数情報のヒストグラムを２次元投影データと呼び、また、３次元同時計数情報のヒストグラムを３次元投影データと呼ぶ。
【００２６】
データ処理部６０は、２次元投影データおよび３次元投影データに基づいて、被検体２内における光子対の発生頻度の空間分布を示す画像を再構成する。また、データ処理部６０は、検出器感度補正、吸収補正および散乱補正をも行う。画像表示部７０は、データ処理部６０により再構成された画像を表示する。
【００２７】
次に、ブランク測定およびトランスミッション測定について説明する。図３は、校正線源を用いたブランク測定およびトランスミッション測定の説明図である。図４は、校正線源の説明図である。これらの図は、図１（ａ）と同様に、中心軸に平行な方向に見たときの図を示している。
【００２８】
校正線源３は、例えば⁶⁸Ｇｅの棒状のものであって、スライスセプタ２０の測定空間１側の中央付近に接して中心軸に平行に配置される。また、校正線源３を挟んで２つのシールド３Ａおよび３Ｂが設けられている。スライスセプタ２０ならびにシールド３Ａおよび３Ｂにより、検出部１０により３次元同時計数情報が検出されることなく、２次元同時計数情報のみが検出される。このようにすることにより、散乱線の寄与が減少され、校正線源３の近くの光子検出器の計数率の異常な上昇が防止される。また、スライスセプタ撤去可能型３次元ＰＥＴ装置のために開発された吸収補正法を適用することができる。
【００２９】
測定空間１に被検体２を置くことなく、校正線源３とともにスライスセプタ２０を回転させてブランク測定を行う。そして、このようにして２次元投影データ蓄積部２０に蓄積された２次元投影データがブランクデータであり、このブランクデータに基づいて検出器感度補正が行われる。また、測定空間１に被検体２を置いて、校正線源３とともにスライスセプタ２０を回転させてトランスミッション測定を行う。そして、このようにして２次元投影データ蓄積部２０に蓄積された２次元投影データがトランスミッションデータであり、このトランスミッションデータに基づいて吸収補正が行われる。なお、ＲＩ線源が投入された被検体２を測定空間１に置き、校正線源３とともにスライスセプタ２０を回転させて、エミッション測定とトランスミッション測定とを同時に行ってもよい。
【００３０】
また、検出器感度補正に際しては、「間接的感度校正法」を用いるのが好適である。図５および図６それぞれは、検出器感度補正の説明図である。校正線源３から発生した光子対が飛行する直線すなわち同時計数線Ｌに対する検出感度は、その光子対を同時計数する１対の光子検出器Ｄ_iおよびＤ_jそれぞれの検出効率と種々の幾何学的因子ε_ijとの積として推定される（図５参照）。ここで、幾何学的因子ε_ijは、検出器リング差δや、測定空間の中心点から同時計数線Ｌまでの距離等が考慮された因子である。これらの因子の中でも光子検出器それぞれの検出効率は、時間的に安定しておらず変動するので、定期的に校正を行う必要がある。
【００３１】
本実施形態に係るＰＥＴ装置は、棒状の校正線源３を用いた測定により、スライスセプタ２０によりコリメートされた光子検出器Ｄ_jおよびコリメートされない光子検出器Ｄ_iにより光子対を同時計数するので、各光子検出器がコリメートされたときの検出効率と、各光子検出器がコリメートされないときの検出効率とを、１回のブランク測定から「fun sum法」を用いて同時に求めることができる（図６参照）。すなわち、コリメートされた光子検出器Ｄ_iを通る多数の同時計数線に関する同時計数情報の平均値に基づいて光子検出器Ｄ_iの検出効率を求める。同様にして、コリメートされた光子検出器Ｄ_jの検出効率を求める。２次元投影データの校正は、光子検出器がコリメートされたときの検出効率に基づいて行われ、一方、３次元投影データの校正は、光子検出器がコリメートされないときの検出効率に基づいて行われる。
【００３２】
次に、散乱補正および画像再構成について説明する。一般に、散乱線のレスポンス（点状線源または棒状線源に対する投影データにおける散乱同時計数の分布）は、２次元投影データと３次元投影データとでは大きく異なっている。すなわち、２次元投影データにおける散乱線は、主に検出部１０の内側すなわち測定空間１内または極近傍での散乱に起因し、中心軸に平行に置かれた棒状線源３に対する散乱レスポンスは、指数関数によってよく近似される。一方、３次元投影データにおける散乱線は、主に測定空間１から離れた場所での散乱に起因する。３次元投影データにおける散乱線の散乱レスポンスは、高い空間周波数成分を殆ど含まず、極低い空間周波数成分を主に含み、ガウス関数または放物線関数によってよく近似される。そこで、本実施形態に係るＰＥＴ装置では、以下に説明するように、２次元投影データを利用することにより、３次元投影データにおける散乱線の寄与を正確に補正（散乱補正）して画像再構成する。
【００３３】
以下に説明する方法は、「差分法（difference method）」呼ばれるものである。この方法では、或るダイレクトプレーンに対して、２次元同時計数情報の蓄積から３次元同時計数情報の蓄積に変更したときの散乱成分の増加分が、３次元投影データと２次元投影データとの差から推定できると仮定する。すなわち、この散乱成分の増加分Ｓ'(ｒ,θ)は、
【００３４】
【数３】

なる式で与えられると仮定する。ここで、ｒは投影の位置座標であり、θは投影の方位角であり、ｐ_2D(ｒ,θ)は２次元投影データであり、ｐ_3D(ｒ,θ)は３次元投影データであり、ε(ｒ,θ)は効率補正因子である。
【００３５】
さらに、２次元投影データに含まれる散乱成分の影響を補正するために、３次元投影データの全散乱成分Ｓ(ｒ,θ)は上記(3)式の散乱分布Ｓ'(ｒ,θ)に補正因子ｋ(θ)を乗じて、
【００３６】
【数４】

なる式で表される。補正因子ｋ(θ)は、線源が存在する領域以外の領域における２次元投影データおよび３次元投影データそれぞれの分布を比較することにより決定される。
【００３７】
各々の検出器リング番号がｎ，ｍである２つの検出器リングＲ_n，Ｒ_mの間で得られる傾斜した投影における散乱分布ｐ_n,m(ｒ,θ)は、検出器リング番号が int((n+m)/2) であるダイレクトプレーンの散乱分布と、検出器リング番号が int((n+m)/2)+1 であるダイレクトプレーンの散乱分布とから、直線補間により得られる。ここで、int は、整数化を行うことを表す演算記号である。ここでは、小さい傾斜角の投影の散乱分布は、その中央位置付近のダイレクトプレーンの散乱分布に略等しいと仮定している。
【００３８】
このようにして推定された散乱分布を、例えば半値幅２５ｍｍのガウス型フィルタにより充分に平滑化して、３次元投影データから差し引く。この差し引いた結果、散乱補正された３次元投影データが得られる。そして、この３次元投影データについて、吸収補正を行い、適当な３次元再構成アルゴリズムにより画像が再構成される。
【００３９】
以上に説明した「差分法」は、スライスセプタ撤去可能型ＰＥＴ装置の為に提案されたものであり、その妥当性が実際の装置で確認されている。しかし、スライスセプタ撤去可能型ＰＥＴ装置における「差分法」には、以下のような２つの大きな限界がある。第１の限界は、２次元投影データと３次元投影データとは互いに別の測定により蓄積されることから、線源の陽電子放出能の分布が速く変化する場合や動態研究の場合には応用できないことである。第２の限界は、大きい傾斜角の投影の散乱分布の推定精度が悪いことである。しかし、本実施形態に係るＰＥＴ装置では、２次元投影データと３次元投影データとは１回の測定により同時に蓄積されることから、第１の限界は問題にはならない。また、本実施形態に係るＰＥＴ装置は、以下の方法（「加算法」と呼ぶ）により第２の限界を克服する。
【００４０】
図７は、散乱補正および画像再構成の手順を説明するフローチャートである。このフローチャートで説明する「加算法」は、２次元投影データに基づいて低い空間周波数成分の画像を再構成するとともに、３次元投影データに基づいて高い空間周波数成分の画像を再構成し、これら２つの再構成画像を加算して、最終的な再構成画像を得るものである。このようにすることにより、大きい傾斜角の投影データにおける散乱補正が可能となる。また、低い空間周波数成分の画像は、２次元投影データに基づいて得られるので、測定空間１以外の空間からの散乱の寄与が小さい。
【００４１】
３次元投影データについては、初めに、上記の「差分法」またはより簡便な方法（例えば「ガウス関数フィッティング法」など）により、散乱補正が施される。この散乱補正は近似的なもので充分である。「ガウス関数フィッティング法」は、被検体以外の領域（散乱のみが計測される領域）の投影データをガウス関数でフィッティングすることにより、被検体における散乱成分を推定する方法である。なお、この散乱補正は、続く吸収補正を適切に行う為に必要である。続いて、この散乱補正された３次元投影データについて、通常の吸収補正が施され、ハイパスフィルタｈ(ｒ)により低周波数成分が除去され、３次元再構成アルゴリズムにより高周波画像が画像再構成される。ハイパスフィルタは、３次元投影データに含まれる散乱成分の殆どを除去するように設計される。
【００４２】
一方、２次元投影データについては、例えば「２エネルギーウィンドウ法」や「重畳積分差し引き法」等により散乱補正が施され、続いて、吸収補正が施され、ローパスフィルタｇ(ｒ)により高周波数成分が除去され、２次元再構成アルゴリズムにより低周波画像が画像再構成される。このローパスフィルタｇ(ｒ)の周波数レスポンスＦ[ｇ(ｒ)]は、ハイパスフィルタｈ(ｒ)の周波数レスポンスＦ[ｈ(ｒ)]に対し相補的であるように設計される。すなわち、Ｆ[ｇ(ｒ)]＋Ｆ[ｈ(ｒ)]＝１なる関係式が成り立つ。なお、Ｆ[・]はフーリエ変換を表す。
【００４３】
最終的な再構成画像は、３次元同時計数情報および２次元同時計数情報それぞれの検出感度を考慮した上で、３次元投影データに基づいて得られた高周波画像と、２次元投影データに基づいて得られた低周波画像とを、適当な荷重をかけて加算することにより得られる。
【００４４】
図８は、散乱補正および画像再構成の他の手順を説明するフローチャートである。このフローチャートで説明する方法は、上記「加算法」で３次元投影データについて「フーリエリビニング（Fourier rebinning:ＦＲＢ）法」を採用したものであり、計算効率が大幅に改善される。
【００４５】
３次元投影データについては、初めに、上記の「差分法」またはより簡便な方法（例えば「ガウス関数フィッティング法」など）により散乱補正が施される。この散乱補正は近似的なもので充分である。なお、この散乱補正は、続く吸収補正を適切に行う為に必要である。続いて、この散乱補正された３次元投影データについて、通常の吸収補正が施され、「ＦＲＢ法」による処理がなされる。
【００４６】
図９は、ＦＲＢ法を説明する図である。ＦＲＢ法では、ダイレクトプレーンに対して傾斜した投影（図９（ａ））について得られた３次元投影データ（図９（ｂ））は変数ｒおよびθに関して２次元フーリエ変換されて、変数ｎおよびωに関する２次元フーリエ変換マップ（図９（ｃ））が得られる。この２次元フーリエ変換マップは、「周波数の距離の関係（Frequency-distance relation）」すなわち「γ＝−ｎ／ω」を用いて、ダイレクトプレーンの２次元フーリエ変換マップ（図９（ｄ））に変換される。このようにして得られた各ダイレクトプレーンの２次元フーリエ変換マップは２次元逆フーリエ変換されて、ダイレクトプレーンの投影データ（図９（ｅ））が得られる。そして、この各ダイレクトプレーンの投影データは、２次元画像再構成されて、再構成画像（図９（ｆ））が得られる。
【００４７】
図８に示したフローチャートでは、近似的な散乱補正および吸収補正が施された３次元投影データに基づいてＦＲＢ法により作成されたダイレクトプレーンの投影データ（図９（ｅ））は、ハイパスフィルタｈ(ｒ)により低周波数成分が除去される。このハイパスフィルタは、３次元投影データに含まれる散乱成分の殆どを除去するように設計される。
【００４８】
一方、２次元投影データについては、例えば「２エネルギーウィンドウ法」や「重畳積分差し引き法」等により散乱補正が施され、続いて、吸収補正が施され、ローパスフィルタｇ(ｒ)により高周波数成分が除去される。このローパスフィルタｇ(ｒ)の周波数レスポンスＦ[ｇ(ｒ)]は、ハイパスフィルタｈ(ｒ)の周波数レスポンスＦ[ｈ(ｒ)]に対し相補的であるように設計される。すなわち、Ｆ[ｇ(ｒ)]＋Ｆ[ｈ(ｒ)]＝１なる関係式が成り立つ。
【００４９】
そして、３次元投影データの低周波数成分と２次元投影データの高周波数成分とは、各ダイレクトプレーン毎に、３次元同時計数情報および２次元同時計数情報それぞれの検出感度を考慮した上で、適当な加重をかけて加算される。最終的な再構成画像は、この加算された投影データに基づいて、２次元再構成アルゴリズムにより得られる。
【００５０】
以上のように、この方法では３次元画像再構成を行わないので計算時間が短い。また、ＦＲＢ法は、ダイレクトプレーンに対して傾斜した投影について得られた３次元投影データに含まれる非常に低い周波数成分について誤差を生ずることが知られているが、本実施形態では低周波数成分が除去されるので、この欠点は問題とはならない。
【００５１】
次に、同時計数情報の統計的変動に因る低周波画像のｒｍｓ（root mean square）誤差について述べる。低周波画像のｒｍｓ誤差の大きさは、高周波画像のｒｍｓ誤差の大きさと比較して充分に小さいことが望ましい。「フィルタ逆投影（filtered backprojection）法」による通常の２次元画像再構成では、解像力（半値幅）をａとし、全計数値をＴとすると、得られる画像の相対的ｒｍｓ雑音は、（ａ³Ｔ）^-1/2に略比例する。したがって、低周波画像および高周波画像それぞれのｒｍｓ雑音の比ratio_rmsは、
【００５２】
【数５】

なる式で与えられる。ここで、ａ_3Dは、３次元投影データに基づいて得られる画像の解像力であり、ａ_2Dは、２次元投影データに基づいて得られる画像の解像力である。Ｒ'は、３次元同時計数情報および２次元同時計数情報それぞれの全計数値の比である。それぞれの場合の散乱比率が異なることから、Ｒ'は、３次元同時計数情報および２次元同時計数情報それぞれの検出感度の比Ｒ（上記(２)式）より幾らか大きい。
【００５３】
例えば、断面有効視野が直径２５６ｍｍである頭部（脳）用ＰＥＴ装置を考えると、３次元投影データに含まれる散乱成分のガウス関数成分の最も狭い半値幅は約１００ｍｍ以上であると想定される。そこで、もし、ａ_2D＝５０ｍｍとし、ａ_3D＝３ｍｍとし、Ｒ'＝５０とすれば、ratio_rmsは０．１０４となる。この数値例は、充分な計数値が得られる高解像度測定の場合のものである。計数値が少ない場合には、ａ_3Dを大きくする必要があり、その結果、ratio_rmsは増大する。例えば、ａ_3D＝１０ｍｍとすると、ratio_rmsは０．６３２となる。
【００５４】
以上のように、本実施形態に係るＰＥＴ装置は、２次元投影データおよび３次元投影データを１回の測定で同時に得ることができるので、線源の陽電子放出能の分布が速く変化する場合や動態研究を行う場合に応用が可能である。また、本実施形態に係るＰＥＴ装置は、光子対を高感度に同時計数して３次元投影データを得ることができ、且つ、２次元投影データに基づいて散乱補正を有効に行うことができる。すなわち、本実施形態に係るＰＥＴ装置は、低い空間周波数成分の画像を２次元投影データに基づいて再構成し、高い空間周波数成分の画像を３次元投影データに基づいて再構成し、これら２つの再構成画像を加算して、最終的な再構成画像を得ることにより、大きい傾斜角の投影データにおける散乱補正が可能となる。
【００５５】
なお、本実施形態に係るＰＥＴ装置においては、全て光子検出器の個数Ｎに対するスライスセプタ２０の後方に存在する光子検出器の個数ｎの比ｓ（＝ｎ／Ｎ）が重要な設計事項となる。(２)式から判るように、ｓの値を大きくすると、(１−２ｓ)に比例して３次元同時計数情報の検出感度が低下し、２ｓに比例して２次元同時計数情報の検出感度が増大する。したがって、ｓの値は、３次元同時計数情報の検出感度と散乱補正の正確度とのバランスを考慮して決定される必要がある。ｓの好適な値は１／２以下であり、より好適な値は１／１０〜１／６程度である。
【００５６】
本発明は、上記実施形態に限定されるものではなく種々の変形が可能である。特に、検出部およびスライスセプタの構成は、以下のように種々の変形が可能である。
【００５７】
図１０は、検出部およびスライスセプタの構成の第１の変形例を説明する図である。この図は、中心軸に平行な方向に検出部を見たときの図を示している。この変形例では、検出部１０の内側に、スライスセプタ２０Ａ，２０Ｂおよび２０Ｃを備えている。スライスセプタ２０Ａ，２０Ｂおよび２０Ｃそれぞれは、図１におけるスライスセプタ２０と同様の構成のものであり、中心軸を中心とする円周上に略等間隔で配置されている。この変形例では、スライスセプタ２０Ａ，２０Ｂおよび２０Ｃそれぞれの回転バランスが優れているので好適である。
【００５８】
図１１は、検出部およびスライスセプタの構成の第２の変形例を説明する図である。図１１（ａ）は、中心軸に平行な方向に検出部を見たときの図を示しており、図１１（ｂ）は、中心軸を含む面で検出部を切断したときの断面を示している。この変形例では、検出部１０Ａおよび１０Ｂそれぞれは、平面上に光子検出器が２次元配列されたものである。また、スライスセプタ２０は、回転軸に垂直な複数のシールド板からなり、一方の検出部１０Ａの一部の内側に固定して設けられている。そして、検出部１０Ａおよび１０Ｂそれぞれは、相対的位置関係を保持したまま被検体２の周りを回転することで、２次元同時計数情報および３次元同時計数情報を検出する。
【００５９】
図１２は、検出部およびスライスセプタの構成の第３の変形例を説明する図である。この図は、中心軸に平行な方向に検出部を見たときの図を示している。この変形例では、検出部１０Ａ，１０Ｂ，１０Ｃおよび１０Ｄそれぞれは、平面上に光子検出器が２次元配列されたものである。また、スライスセプタ２０Ａおよび２０Ｄそれぞれは、回転軸に垂直な複数のシールド板からなり、検出部１０Ａ，１０Ｄの一部の内側に固定して設けられている。そして、検出部１０Ａ〜１０Ｄそれぞれは、相対的位置関係を保持したまま被検体２の周りを回転することで、２次元同時計数情報および３次元同時計数情報を検出する。
【００６０】
【発明の効果】
以上、詳細に説明したとおり、本発明によれば、複数組の検出器リングそれぞれを構成する複数の光子検出器のうち一部のものの測定空間の側に回転自在のスライスセプタを設けたことにより、２次元投影データおよび３次元投影データが１回の測定で同時に得られる。したがって、このように２次元投影データおよび３次元投影データを同時に獲得して画像再構成を行うことにより、光子対を高感度に同時計数することができ、且つ、散乱補正等を行うことができる。また、線源の陽電子放出能の分布が速く変化する場合や動態研究を行う場合に応用が可能である。
【００６１】
特に、２次元投影データのうち空間周波数が低い成分、および、３次元投影データのうち空間周波数が高い成分に基づいて、画像の再構成を行うことにより、光子対を高感度に同時計数して３次元投影データを得ることが好適にでき、且つ、２次元投影データに基づいて散乱補正を有効に行うことが好適にできる。しかも、大きい傾斜角の投影データにおける散乱補正が可能となる。
【００６２】
本発明に係るＰＥＴ装置は、スライスセプタの測定空間の側に中心軸と平行な棒状の校正線源を設けて２次元投影データを蓄積し、この２次元投影データに基づいて再構成画像を補正することにより、検出器感度補正および吸収補正が好適に行われ、良好な再構成画像が得られる。
【図面の簡単な説明】
【図１】本実施形態に係るＰＥＴ装置の検出部およびスライスセプタの構成を説明する図である。
【図２】本実施形態に係るＰＥＴ装置の全体構成を概念的に説明するブロック図である。
【図３】回転校正線源を用いたブランク測定およびトランスミッション測定の説明図である。
【図４】校正線源の説明図である。
【図５】検出器感度補正の説明図である。
【図６】検出器感度補正の説明図である。
【図７】散乱補正および画像再構成の手順を説明するフローチャートである。
【図８】散乱補正および画像再構成の手順を説明するフローチャートである。
【図９】ＦＲＢ法を説明する図である。
【図１０】検出部およびスライスセプタの構成の第１の変形例を説明する図である。
【図１１】検出部およびスライスセプタの構成の第２の変形例を説明する図である。
【図１２】検出部およびスライスセプタの構成の第３の変形例を説明する図である。
【図１３】２次元ＰＥＴ装置の検出部の構成を説明する図である。
【図１４】３次元ＰＥＴ装置の検出部の構成を説明する図である。
【図１５】スライスセプタ撤去可能型３次元ＰＥＴ装置の検出部の構成を説明する図である。
【符号の説明】
１…測定空間、２…被検体、３…校正線源、１０…検出部、１１，１２…シールド、２０…スライスセプタ、３０…セプタ回転駆動部、４０…回転位置検出センサ、５１…２次元投影データ蓄積部、５２…３次元投影データ蓄積部、６０…データ処理部、７０…画像表示部、Ｄ₁〜Ｄ_N…検出器、Ｒ₁〜Ｒ₁₀…検出器リング、Ｓ₁〜Ｓ₉…シールド板。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a PET apparatus capable of imaging the behavior of a trace amount substance labeled with a positron emitting isotope (RI radiation source).
[0002]
[Prior art]
A PET (positron emission tomography) apparatus detects a photon pair that is generated in association with the annihilation of electrons and positrons in a living body (subject) into which an RI radiation source has been introduced and flies in opposite directions. It is an apparatus that can image the behavior of a trace amount substance in a specimen. The PET apparatus includes a detection unit having a large number of small photon detectors arranged around a measurement space in which a subject is placed, and simultaneously counts photon pairs generated as a result of electron / positron pair annihilation. And reconstructing an image representing the spatial distribution of the occurrence frequency of photon pairs in the measurement space based on the accumulated coincidence information, that is, projection data. This PET apparatus plays an important role in the field of nuclear medicine and the like, and can be used to study, for example, biological functions and higher-order brain functions. Such a PET apparatus is roughly classified into a two-dimensional PET apparatus, a three-dimensional PET apparatus, and a slicer detachable three-dimensional PET apparatus.
[0003]
FIG. 13 is a diagram illustrating the configuration of the detection unit of the two-dimensional PET apparatus. This figure shows a cross section when the detection unit is cut along a plane including the central axis. The detection unit 10 of the two-dimensional PET apparatus includes a detector ring R stacked between the shield 11 and the shield 12. ₁ ~ R ₇ have. Detector ring R ₁ ~ R ₇ Each has a plurality of photon detectors arranged in a ring shape on a plane perpendicular to the central axis. Each photon detector is, for example, BGO (Bi _Four Ge _Three O ₁₂ The scintillation detector is a combination of a scintillator and a photomultiplier tube, and detects photons that have arrived from the measurement space including the central axis. A slice scepter 20 is provided inside the detection unit 10. The slice ceptor 20 is composed of six ring-shaped shield plates S arranged at positions between adjacent detector rings. ₁ ~ S ₆ Consists of. The detection unit 10 of the two-dimensional PET apparatus configured in this way can simultaneously count only photon pairs flying from a direction whose angle with the central axis is approximately 90 degrees by the collimating action of the slice ceptor 20. That is, the coincidence counting information obtained by the detection unit 10 of the two-dimensional PET apparatus, that is, the two-dimensional projection data, is a pair of photon detections included in the same detector ring or adjacent (or very close) detector rings. Limited to equipment. Therefore, the two-dimensional PET apparatus can efficiently exclude scattered rays in which photon pairs generated at positions outside the measurement space are scattered, and absorption correction and sensitivity correction can be easily performed on two-dimensional projection data. It can be carried out.
[0004]
FIG. 14 is a diagram illustrating the configuration of the detection unit of the three-dimensional PET apparatus. This figure also shows a cross section when the detection unit is cut along a plane including the central axis. The configuration of the detection unit 10 of the three-dimensional PET apparatus is the same as that of the two-dimensional PET apparatus. However, the three-dimensional PET apparatus is not equipped with a slice ceptor. The detection unit 10 of the three-dimensional PET apparatus configured as described above can simultaneously count photon pairs flying from all directions. That is, the coincidence counting information obtained by the detection unit 10 of the three-dimensional PET apparatus, that is, the three-dimensional projection data, can be obtained by a pair of photon detectors included in an arbitrary detector ring. Therefore, the three-dimensional PET apparatus can simultaneously count photon pairs with a sensitivity about five to ten times higher than that of the two-dimensional PET apparatus.
[0005]
FIG. 15 is a diagram illustrating the configuration of the detection unit of the slice scepter removable type three-dimensional PET apparatus. This figure also shows a cross section when the detection unit is cut along a plane including the central axis. The configuration of the detection unit 10 of the slice ceptor removable type three-dimensional PET apparatus is the same as that of the two-dimensional PET apparatus. However, in the slice scepter removable type three-dimensional PET apparatus, the slice scepter 20 can be retracted in a retracting space provided on the shield 12 side. In other words, the slice scepter removable type 3D PET apparatus has a slice ceptor 20 with a detector ring R. ₁ ~ R _Five Is equivalent to the two-dimensional PET apparatus (FIG. 15A), and is equivalent to the three-dimensional PET apparatus when the slice ceptor 20 is retracted in the retreat space (FIG. 15 ( b)). Therefore, the slice scepter removable type three-dimensional PET apparatus is used as either one of the two-dimensional PET apparatus or the three-dimensional PET apparatus depending on the purpose.
[0006]
[Problems to be solved by the invention]
However, the conventional PET apparatus has the following problems. That is, since the two-dimensional PET apparatus only counts photon pairs flying from a direction whose angle with the central axis is approximately 90 degrees, it cannot simultaneously count photon pairs with high sensitivity. On the other hand, the three-dimensional PET apparatus cannot efficiently exclude scattered rays in which photon pairs generated in a space other than the measurement space are scattered, and scattering correction, absorption correction, and sensitivity correction are difficult or complicated. Therefore, a good image cannot be reconstructed.
[0007]
Further, since the slice scepter removable type 3D PET apparatus acquires 2D projection data and 3D projection data by different measurements, each of the 2D PET apparatus and 3D PET apparatus described above has each. The problem cannot be solved at the same time. In addition, the apparatus configuration is complicated and expensive.
[0008]
The present invention has been made to solve the above-described problems, and enables two-dimensional projection data and three-dimensional projection data to be acquired simultaneously to enable highly sensitive photon pair coincidence and effective scattering correction. It is an object of the present invention to provide a PET apparatus capable of performing the above.
[0009]
[Means for Solving the Problems]
The PET apparatus according to the present invention includes (1) a plurality of sets of detector rings in which a plurality of photon detectors that respectively detect photons flying from a measurement space including the central axis are arranged on a plane perpendicular to the central axis. A detection unit in which a plurality of sets of detector rings are stacked in a direction parallel to the central axis, and (2) a measurement space of a part of a plurality of photon detectors constituting each of the plurality of sets of detector rings. And a slice scepter that is arranged so as to be rotatable about the central axis, and allows only the photons that have come in the direction perpendicular to the central axis to pass, and (3) a pair of photon detectors included in the detector A slice sceptor position determination means for determining whether or not a slice ceptor is present on the side of at least one of the pair of photon detectors when the photon detector simultaneously counts photon pairs; (4) At least one of the pair of photon detectors Two-dimensional projection data accumulating means for accumulating coincidence information of photon pairs by a pair of photon detectors when it is determined by the slice sceptor position determining means that a slice ceptor is present on one measurement space side; (5) When it is determined by the slice sceptor position determination means that no slice scepter is present on either measurement space side of the pair of photon detectors, the photon pair of the pair of photon detectors is determined. 3D projection data storage means for storing coincidence count information and (6) 2D projection data storage means for generating and generating coincidence count information 2D projection data And image reconstruction means for reconstructing an image representing the spatial distribution of the occurrence frequency of photon pairs in the measurement space based on the 3D projection data generated by accumulating the coincidence count information by the 3D projection data storage means And.
[0010]
According to this PET apparatus, when photon pairs flying from the measurement space are simultaneously counted by the pair of photon detectors of the detection unit, the slice ceptor is placed on the side of at least one measurement space of the pair of photon detectors. Is determined by the slice scepter position determination means. This determination is performed based on, for example, the rotational position of the slice scepter detected by the rotational position detection sensor. If the slice ceptor position determination means determines that a slice ceptor exists on the side of at least one measurement space, the photon pair coincidence information by the pair of photon detectors is stored in the two-dimensional projection data. Accumulated by means. On the contrary, when it is determined by the slice sceptor position determination means that no slice ceptor exists on the side of the measurement space, the coincidence information of the photon pairs by the pair of photon detectors is the three-dimensional projection data storage means. Is accumulated. The coincidence information is accumulated and generated by the image reconstruction means by the two-dimensional projection data accumulation means. 2D projection data Based on the three-dimensional projection data generated by accumulating the coincidence counting information by the three-dimensional projection data storage means, an image representing the spatial distribution of the occurrence frequency of photon pairs in the measurement space is reconstructed. As described above, the two-dimensional projection data and the three-dimensional projection data are obtained simultaneously by one measurement. Therefore, by simultaneously acquiring 2D projection data and 3D projection data and performing image reconstruction in this way, photon pairs can be simultaneously counted with high sensitivity, and scattering correction and the like can be performed. .
[0011]
The image reconstruction unit of the PET apparatus according to the present invention performs image reconstruction based on a component having a low spatial frequency in the two-dimensional projection data and a component having a high spatial frequency in the three-dimensional projection data. Features. In this case, it is preferable to simultaneously count photon pairs with high sensitivity to obtain three-dimensional projection data, and it is preferable to effectively perform scattering correction based on the two-dimensional projection data. In addition, it is possible to correct scattering in projection data with a large tilt angle.
[0012]
The PET apparatus according to the present invention is provided with a rod-shaped calibration radiation source parallel to the central axis on the measurement space side of the slice ceptor, and based on the two-dimensional projection data stored in the two-dimensional projection data storage means, The image processing apparatus further includes a correction unit that corrects the image reconstructed by the configuration unit. In this case, detector sensitivity correction and absorption correction are preferably performed, and a good reconstructed image is obtained.
[0013]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.
[0014]
FIG. 1 is a diagram for explaining a configuration of a detection unit and a slice scepter of a PET apparatus according to the present embodiment. 1A shows a view when the detection unit 10 is viewed in a direction parallel to the central axis, and FIG. 1B shows a cross section when the detection unit 10 is cut along a plane including the central axis. Is shown.
[0015]
The detection unit 10 of the PET apparatus according to this embodiment includes a detector ring R stacked between a shield 11 and a shield 12. ₁ ~ R _Ten have. Detector ring R ₁ ~ R _Ten Each of the N photon detectors D arranged in a ring shape on a plane perpendicular to the central axis ₁ ~ D _N have. Photon detector D ₁ ~ D _N For example, BGO (Bi _Four Ge _Three O ₁₂ The scintillation detector is a combination of a scintillator and a photomultiplier tube, and detects photons that have arrived from the measurement space 1 including the central axis.
[0016]
A slice ceptor 20 is provided inside the detection unit 10, that is, on the measurement space 1 side. The slice ceptor 20 includes nine shield plates S arranged at positions between adjacent detector rings. ₁ ~ S ₉ including. Shield plate S ₁ ~ S ₉ Each is made of a photon pair that is generated when the pair of electrons and positrons annihilate and flies in opposite directions, that is, a material that absorbs γ-rays having energy of 511 keV (for example, tungsten, lead). The slice ceptor 20 has a collimating effect, and causes only a photon pair flying from a direction whose angle with the central axis is approximately 90 degrees to enter the photon detector behind the photon detector.
[0017]
Shield plate S ₁ ~ S ₉ Each is not ring-shaped, but N photon detectors D constituting each detector ring. ₁ ~ D _N Of them (in FIG. 1, photon detector D _Three ~ D _Ten ) On the measurement space 1 side. If the number of photon detectors behind the slice ceptor 20 is n (eight in FIG. 1), the value of n / N is preferably 1/2 or less, and about 1/10 to 1/6. Is particularly preferred. The slice ceptor 20 is rotatable about a central axis and performs continuous rotation, step rotation, or reciprocal rotation. The rotational position of the slice ceptor 20 is detected by a rotational position detection sensor, or is grasped by a scepter rotation drive unit that controls the rotation.
[0018]
In the detection unit 10, when at least one of the pair of photon detectors is behind the slice scepter 20, the pair of photon detectors flies from a direction whose angle with the central axis is approximately 90 degrees. Only photon pairs are counted simultaneously. In addition, these one pair of photon detectors can efficiently exclude scattered rays in which photon pairs generated at positions outside the measurement space are scattered. That is, the coincidence information obtained by the pair of photon detectors is equivalent to that obtained in the two-dimensional PET apparatus. Hereinafter, this coincidence information is referred to as two-dimensional coincidence information.
[0019]
On the other hand, when none of the pair of photon detectors is behind the slice scepter 20, the pair of photon detectors can simultaneously count the photon pairs flying from all directions. That is, the coincidence information obtained by the pair of photon detectors is equivalent to that obtained in the three-dimensional PET apparatus. Hereinafter, this coincidence information is referred to as three-dimensional coincidence information.
[0020]
The geometric detection efficiency of the two-dimensional coincidence information is highest in the same detector ring plane (direct plane), and the difference between the detector ring numbers (value n in the detector ring code Rn), that is, the ring difference δ is large. It decreases as a result of the shielding effect of the slice ceptor 20. Its effective axial viewing angle φ _2D Is
[0021]
[Expression 1]

It is expressed by the following formula. Where D is the inner diameter of the detector ring, d is the axial width of the photon detector, ε (δ) is the relative detection sensitivity of the projection tilted with respect to the direct plane, δ _max Is the maximum ring difference in the accumulation of two-dimensional coincidence information.
[0022]
On the other hand, the viewing angle in the axial direction for accumulating three-dimensional coincidence information is φ _3D Let s (= n / N) be the ratio of the number n of photon detectors existing behind the slice ceptor 20 to the number N of all photon detectors. At this time, the ratio R of the detection sensitivity of each of the three-dimensional coincidence information and the two-dimensional coincidence information is
[0023]
[Expression 2]

It is expressed by the following approximate expression. Here, w is the radial width of the slice ceptor 20. In deriving this equation, the effects of photon absorption and scattering were ignored. For example, φ _2D = 1 °, φ _3D = 10 °, s = 1/8, d = 6 mm, w = 1 mm, R = 36.
[0024]
FIG. 2 is a block diagram conceptually illustrating the overall configuration of the PET apparatus according to the present embodiment. The septa rotation driving unit 30 rotates the slice ceptor 30 around the central axis, and the rotational position detection sensor 40 detects the rotational position of the slice ceptor 20. During a single measurement period in which the subject 2 is placed in the measurement space 1, the slice ceptor 20 is driven and rotated by the scepter rotation drive unit 30, and the rotational position of the slice ceptor 20 is a rotational position detection sensor. 40 is always grasped. When the pair of photon detectors simultaneously count the photon pairs, it is determined whether or not at least one of the pair of photon detectors is located behind the slice scepter 20. This determination is made based on the rotational position of the slice ceptor 20 detected by the rotational position detection sensor 40.
[0025]
If it is determined that one photon detector is located behind the slice scepter 20, the coincidence information detected by the pair of photon detectors is determined to be two-dimensional coincidence information, and the two-dimensional coincidence information is detected. The count information is stored in the two-dimensional projection data storage unit 51. Otherwise, it is determined that the coincidence information detected by the pair of photon detectors is three-dimensional coincidence information, and the three-dimensional coincidence information is accumulated in the three-dimensional projection data accumulating unit 52. . In this way, each of the two-dimensional coincidence information and the three-dimensional coincidence information is accumulated separately from each other to create a histogram. Hereinafter, the histogram of the two-dimensional coincidence information is referred to as two-dimensional projection data, and the histogram of the three-dimensional coincidence information is referred to as three-dimensional projection data.
[0026]
The data processing unit 60 reconstructs an image indicating the spatial distribution of the occurrence frequency of photon pairs in the subject 2 based on the two-dimensional projection data and the three-dimensional projection data. The data processing unit 60 also performs detector sensitivity correction, absorption correction, and scattering correction. The image display unit 70 displays the image reconstructed by the data processing unit 60.
[0027]
Next, blank measurement and transmission measurement will be described. FIG. 3 is an explanatory diagram of blank measurement and transmission measurement using a calibration radiation source. FIG. 4 is an explanatory diagram of the calibration radiation source. These figures show the figure when viewed in the direction parallel to the central axis, as in FIG.
[0028]
The calibration radiation source 3 is, for example, ⁶⁸ It is a Ge-shaped rod and is arranged in contact with the vicinity of the center of the slice ceptor 20 on the measurement space 1 side and in parallel with the central axis. Further, two shields 3A and 3B are provided with the calibration source 3 interposed therebetween. Only the two-dimensional coincidence information is detected by the slice ceptor 20 and the shields 3A and 3B without detecting the three-dimensional coincidence information by the detection unit 10. By doing so, the contribution of scattered radiation is reduced, and an abnormal increase in the counting rate of the photon detector near the calibration source 3 is prevented. Further, the absorption correction method developed for the slice scepter removable type three-dimensional PET apparatus can be applied.
[0029]
Blank measurement is performed by rotating the slice ceptor 20 together with the calibration source 3 without placing the subject 2 in the measurement space 1. The two-dimensional projection data stored in the two-dimensional projection data storage unit 20 in this way is blank data, and detector sensitivity correction is performed based on this blank data. Further, the subject 2 is placed in the measurement space 1 and the slice ceptor 20 is rotated together with the calibration source 3 to perform transmission measurement. The two-dimensional projection data stored in the two-dimensional projection data storage unit 20 in this way is transmission data, and absorption correction is performed based on the transmission data. Note that the subject 2 to which the RI radiation source has been placed may be placed in the measurement space 1, and the slice ceptor 20 may be rotated together with the calibration radiation source 3 to perform the emission measurement and the transmission measurement simultaneously.
[0030]
In addition, it is preferable to use the “indirect sensitivity calibration method” when correcting the detector sensitivity. 5 and 6 are explanatory diagrams of detector sensitivity correction. The detection sensitivity for the straight line that coincides with the photon pair generated from the calibration source 3, that is, the coincidence line L, is a pair of photon detectors D that simultaneously count the photon pair. _i And D _j Each detection efficiency and various geometric factors ε _ij (See FIG. 5). Where the geometric factor ε _ij Is a factor considering the detector ring difference δ, the distance from the center point of the measurement space to the coincidence line L, and the like. Among these factors, the detection efficiency of each photon detector is not stable in time and fluctuates, so it is necessary to perform calibration periodically.
[0031]
The PET apparatus according to the present embodiment is a photon detector D collimated by a slice ceptor 20 by measurement using a rod-shaped calibration source 3. _j And uncollimated photon detector D _i Since the photon pairs are counted simultaneously, the detection efficiency when each photon detector is collimated and the detection efficiency when each photon detector is not collimated are calculated using a “fun sum method” from one blank measurement. (See FIG. 6). That is, the collimated photon detector D _i Photon detector D based on the average value of coincidence information for a number of coincidence lines passing through _i Obtain the detection efficiency of. Similarly, collimated photon detector D _j Obtain the detection efficiency of. Calibration of the two-dimensional projection data is performed based on the detection efficiency when the photon detector is collimated, while calibration of the three-dimensional projection data is performed based on the detection efficiency when the photon detector is not collimated. .
[0032]
Next, scattering correction and image reconstruction will be described. In general, the response of scattered radiation (distribution of the scattering coincidence count in the projection data for the point source or the rod source) is greatly different between the two-dimensional projection data and the three-dimensional projection data. That is, the scattered radiation in the two-dimensional projection data is mainly caused by scattering inside the detection unit 10, that is, in the measurement space 1 or in the vicinity of the pole, and the scattering response to the rod-shaped radiation source 3 placed parallel to the central axis is It is well approximated by an exponential function. On the other hand, scattered radiation in the three-dimensional projection data is mainly caused by scattering at a location away from the measurement space 1. The scattered response of scattered radiation in the three-dimensional projection data contains almost no high spatial frequency component, mainly contains an extremely low spatial frequency component, and is well approximated by a Gaussian function or a parabolic function. Therefore, in the PET apparatus according to the present embodiment, as described below, by using the two-dimensional projection data, the contribution of scattered radiation in the three-dimensional projection data is accurately corrected (scatter correction), and image reconstruction is performed. To do.
[0033]
The method described below is called the “difference method”. In this method, for a certain direct plane, the amount of increase in the scattering component when changing from the accumulation of 2D coincidence information to the accumulation of 3D coincidence information is the difference between the 3D projection data and the 2D projection data. Assume that it can be estimated from the difference. That is, the increase S ′ (r, θ) of the scattering component is
[0034]
[Equation 3]

Suppose that Here, r is the position coordinate of the projection, θ is the azimuth angle of the projection, and p _2D (r, θ) is two-dimensional projection data, and p _3D (r, θ) is three-dimensional projection data, and ε (r, θ) is an efficiency correction factor.
[0035]
Further, in order to correct the influence of the scattering component included in the two-dimensional projection data, the total scattering component S (r, θ) of the three-dimensional projection data is changed to the scattering distribution S ′ (r, θ) in the above equation (3). Multiply by the correction factor k (θ)
[0036]
[Expression 4]

It is expressed by the following formula. The correction factor k (θ) is determined by comparing the distributions of the two-dimensional projection data and the three-dimensional projection data in regions other than the region where the radiation source exists.
[0037]
Two detector rings R with each detector ring number n, m _n , R _m Scattering distribution p in the tilted projection obtained between _{n, m} (r, θ) is the scattering distribution of the direct plane whose detector ring number is int ((n + m) / 2) and the detector ring number is int ((n + m) / 2) +1 It is obtained by linear interpolation from the scattering distribution of the direct plane. Here, int is an operation symbol indicating that integerization is performed. Here, it is assumed that the scattering distribution of the projection with a small inclination angle is substantially equal to the scattering distribution of the direct plane near the center position.
[0038]
The scattering distribution estimated in this way is sufficiently smoothed by, for example, a Gaussian filter having a half width of 25 mm and subtracted from the three-dimensional projection data. As a result of the subtraction, three-dimensional projection data corrected for scattering is obtained. Then, absorption correction is performed on the three-dimensional projection data, and an image is reconstructed by an appropriate three-dimensional reconstruction algorithm.
[0039]
The “difference method” described above has been proposed for a PET apparatus capable of removing a slice sceptor, and its validity has been confirmed by an actual apparatus. However, the “difference method” in the PET apparatus capable of removing the slice sceptor has the following two major limitations. The first limitation is that 2D projection data and 3D projection data are accumulated by different measurements, and therefore cannot be applied when the positron emission power distribution of the radiation source changes rapidly or in dynamic studies. That is. The second limitation is that the estimation accuracy of the scattering distribution of the projection with a large tilt angle is poor. However, in the PET apparatus according to the present embodiment, the first limit is not a problem because the two-dimensional projection data and the three-dimensional projection data are accumulated simultaneously by one measurement. Moreover, the PET apparatus according to the present embodiment overcomes the second limitation by the following method (referred to as “addition method”).
[0040]
FIG. 7 is a flowchart for explaining the procedure of scattering correction and image reconstruction. The “addition method” described in this flowchart reconstructs an image having a low spatial frequency component based on two-dimensional projection data, and reconstructs an image having a high spatial frequency component based on three-dimensional projection data. One reconstructed image is added to obtain a final reconstructed image. By doing so, it is possible to perform scattering correction in projection data with a large inclination angle. Further, since an image having a low spatial frequency component is obtained based on the two-dimensional projection data, the contribution of scattering from a space other than the measurement space 1 is small.
[0041]
The three-dimensional projection data is first subjected to scatter correction by the above-mentioned “difference method” or a simpler method (for example, “Gauss function fitting method”). For this scattering correction, an approximate one is sufficient. The “Gaussian function fitting method” is a method for estimating a scattered component in a subject by fitting projection data of a region other than the subject (a region where only scattering is measured) with a Gaussian function. This scattering correction is necessary to appropriately perform subsequent absorption correction. Subsequently, normal scattering correction is performed on the scattering-corrected three-dimensional projection data, low-frequency components are removed by the high-pass filter h (r), and a high-frequency image is reconstructed by a three-dimensional reconstruction algorithm. . The high pass filter is designed to remove most of the scattered components contained in the three-dimensional projection data.
[0042]
On the other hand, the two-dimensional projection data is subjected to scattering correction by, for example, “two-energy window method” or “superimposition integration subtraction method”, followed by absorption correction, and a high-frequency component by low-pass filter g (r). Are removed, and a low-frequency image is reconstructed by a two-dimensional reconstruction algorithm. The frequency response F [g (r)] of the low-pass filter g (r) is designed to be complementary to the frequency response F [h (r)] of the high-pass filter h (r). That is, the relational expression F [g (r)] + F [h (r)] = 1 holds. Note that F [•] represents a Fourier transform.
[0043]
The final reconstructed image is based on the high-frequency image obtained based on the three-dimensional projection data and the two-dimensional projection data in consideration of the detection sensitivity of each of the three-dimensional coincidence information and the two-dimensional coincidence information. It can be obtained by adding the obtained low-frequency image under an appropriate load.
[0044]
FIG. 8 is a flowchart for explaining another procedure of scattering correction and image reconstruction. The method described in this flowchart employs the “Fourier rebinning (FRB) method” for the three-dimensional projection data in the “addition method”, and the calculation efficiency is greatly improved.
[0045]
The three-dimensional projection data is first subjected to scatter correction by the above-mentioned “difference method” or a simpler method (for example, “Gauss function fitting method”). For this scattering correction, an approximate one is sufficient. This scattering correction is necessary to appropriately perform subsequent absorption correction. Subsequently, the normal absorption correction is performed on the scattering-corrected three-dimensional projection data, and processing by the “FRB method” is performed.
[0046]
FIG. 9 is a diagram for explaining the FRB method. In the FRB method, the three-dimensional projection data (FIG. 9B) obtained for the projection inclined with respect to the direct plane (FIG. 9A) is subjected to a two-dimensional Fourier transform with respect to the variables r and θ, and the variables n and A two-dimensional Fourier transform map relating to ω (FIG. 9C) is obtained. This two-dimensional Fourier transform map is converted into a direct plane two-dimensional Fourier transform map (FIG. 9D) using “Frequency-distance relation”, that is, “γ = −n / ω”. Converted. The two-dimensional Fourier transform map of each direct plane thus obtained is subjected to a two-dimensional inverse Fourier transform to obtain direct plane projection data (FIG. 9 (e)). Then, the projection data of each direct plane is reconstructed into a two-dimensional image to obtain a reconstructed image (FIG. 9 (f)).
[0047]
In the flowchart shown in FIG. 8, the direct plane projection data (FIG. 9 (e)) created by the FRB method based on the three-dimensional projection data subjected to the approximate scattering correction and absorption correction is the high-pass filter h. (r) removes low frequency components. This high-pass filter is designed to remove most of the scattered components contained in the three-dimensional projection data.
[0048]
On the other hand, the two-dimensional projection data is subjected to scattering correction by, for example, “two-energy window method” or “superimposition integration subtraction method”, followed by absorption correction, and a high-frequency component by low-pass filter g (r). Is removed. The frequency response F [g (r)] of the low-pass filter g (r) is designed to be complementary to the frequency response F [h (r)] of the high-pass filter h (r). That is, the relational expression F [g (r)] + F [h (r)] = 1 holds.
[0049]
The low-frequency component of the three-dimensional projection data and the high-frequency component of the two-dimensional projection data are appropriate for each direct plane after considering the detection sensitivity of the three-dimensional coincidence information and the two-dimensional coincidence information. It is added with a certain weight. A final reconstructed image is obtained by a two-dimensional reconstruction algorithm based on the added projection data.
[0050]
As described above, since this method does not perform 3D image reconstruction, the calculation time is short. Further, the FRB method is known to cause an error for a very low frequency component included in the three-dimensional projection data obtained for the projection inclined with respect to the direct plane. This defect is not a problem because it is eliminated.
[0051]
Next, the rms (root mean square) error of the low-frequency image due to the statistical fluctuation of the coincidence information will be described. The magnitude of the rms error of the low-frequency image is desirably sufficiently smaller than the magnitude of the rms error of the high-frequency image. In normal two-dimensional image reconstruction by the “filtered backprojection method”, when the resolution (half-value width) is a and the total count value is T, the relative rms noise of the obtained image is (a ^Three T) ^-1/2 Is approximately proportional to Therefore, the ratio rms noise ratio between the low frequency image and the high frequency image, respectively. _rms Is
[0052]
[Equation 5]

It is given by Where a _3D Is the resolution of the image obtained based on the three-dimensional projection data, and a _2D Is the resolution of the image obtained based on the two-dimensional projection data. R ′ is the ratio of the total count values of the three-dimensional coincidence information and the two-dimensional coincidence information. Since the scattering ratio in each case is different, R ′ is somewhat larger than the detection sensitivity ratio R of the three-dimensional coincidence information and the two-dimensional coincidence information (Equation (2) above).
[0053]
For example, when considering a head (brain) PET apparatus having a cross-sectional effective field of view of 256 mm in diameter, the narrowest half-value width of the Gaussian function component of the scattering component included in the three-dimensional projection data is assumed to be about 100 mm or more. . So, if a _2D = 50 mm, a _3D = 3mm and R '= 50, the ratio _rms Becomes 0.104. This numerical example is for high-resolution measurement that provides a sufficient count value. If the count is small, a _3D As a result, the ratio _rms Will increase. For example, a _3D = 10mm ratio _rms Becomes 0.632.
[0054]
As described above, since the PET apparatus according to the present embodiment can simultaneously obtain two-dimensional projection data and three-dimensional projection data by one measurement, the distribution of the positron emission ability of the radiation source changes rapidly. It can be applied when conducting dynamic research. In addition, the PET apparatus according to the present embodiment can obtain three-dimensional projection data by simultaneously counting photon pairs with high sensitivity, and can effectively perform scattering correction based on the two-dimensional projection data. That is, the PET apparatus according to the present embodiment reconstructs an image having a low spatial frequency component based on two-dimensional projection data, and reconstructs an image having a high spatial frequency component based on three-dimensional projection data. By adding the reconstructed images to obtain a final reconstructed image, it is possible to correct scattering in projection data with a large inclination angle.
[0055]
In the PET apparatus according to this embodiment, the ratio s (= n / N) of the number n of photon detectors existing behind the slice scepter 20 to the number N of photon detectors is an important design item. . As can be seen from equation (2), when the value of s is increased, the detection sensitivity of the three-dimensional coincidence information decreases in proportion to (1-2s), and the detection sensitivity of the two-dimensional coincidence information in proportion to 2s. Will increase. Therefore, the value of s needs to be determined in consideration of the balance between the detection sensitivity of the three-dimensional coincidence information and the accuracy of the scattering correction. A preferable value of s is 1/2 or less, and a more preferable value is about 1/10 to 1/6.
[0056]
The present invention is not limited to the above embodiment, and various modifications can be made. In particular, the configuration of the detection unit and the slice ceptor can be variously modified as follows.
[0057]
FIG. 10 is a diagram illustrating a first modification of the configuration of the detection unit and the slice ceptor. This figure shows a view when the detection unit is viewed in a direction parallel to the central axis. In this modified example, slice septa 20A, 20B, and 20C are provided inside the detection unit 10. Each of the slice scepters 20A, 20B, and 20C has the same configuration as that of the slice scepter 20 in FIG. 1, and is arranged at substantially equal intervals on a circumference centered on the central axis. This modification is preferable because the rotation balance of each of the slice scepters 20A, 20B, and 20C is excellent.
[0058]
FIG. 11 is a diagram illustrating a second modification of the configuration of the detection unit and the slice ceptor. FIG. 11A shows a view when the detection unit is viewed in a direction parallel to the central axis, and FIG. 11B shows a cross section when the detection unit is cut along a plane including the central axis. ing. In this modification, each of the detection units 10A and 10B is a two-dimensional array of photon detectors on a plane. The slice ceptor 20 is composed of a plurality of shield plates perpendicular to the rotation axis, and is fixedly provided inside a part of the one detection unit 10A. Then, each of the detection units 10A and 10B detects the two-dimensional coincidence information and the three-dimensional coincidence information by rotating around the subject 2 while maintaining the relative positional relationship.
[0059]
FIG. 12 is a diagram illustrating a third modification of the configuration of the detection unit and the slice ceptor. This figure shows a view when the detection unit is viewed in a direction parallel to the central axis. In this modification, each of the detection units 10A, 10B, 10C, and 10D has a two-dimensional array of photon detectors on a plane. Each of the slice ceptors 20A and 20D includes a plurality of shield plates perpendicular to the rotation axis, and is fixedly provided inside a part of the detection units 10A and 10D. Then, each of the detection units 10A to 10D detects the two-dimensional coincidence information and the three-dimensional coincidence information by rotating around the subject 2 while maintaining the relative positional relationship.
[0060]
【The invention's effect】
As described above in detail, according to the present invention, a rotatable slice sceptor is provided on the measurement space side of some of the plurality of photon detectors constituting each of the plurality of sets of detector rings. Two-dimensional projection data and three-dimensional projection data are obtained simultaneously in one measurement. Therefore, by simultaneously acquiring 2D projection data and 3D projection data and performing image reconstruction in this way, photon pairs can be simultaneously counted with high sensitivity, and scattering correction and the like can be performed. . It can also be applied to the case where the distribution of positron emission ability of the radiation source changes rapidly or to study dynamics.
[0061]
In particular, by reconstructing an image based on a component having a low spatial frequency in two-dimensional projection data and a component having a high spatial frequency in three-dimensional projection data, photon pairs can be simultaneously counted with high sensitivity. It is preferable to obtain three-dimensional projection data, and it is preferable to effectively perform scattering correction based on the two-dimensional projection data. In addition, it is possible to correct scattering in projection data with a large tilt angle.
[0062]
The PET apparatus according to the present invention provides a rod-shaped calibration radiation source parallel to the central axis on the measurement space side of the slice ceptor, accumulates two-dimensional projection data, and corrects the reconstructed image based on the two-dimensional projection data. By doing so, detector sensitivity correction and absorption correction are suitably performed, and a good reconstructed image is obtained.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a configuration of a detection unit and a slice scepter of a PET apparatus according to the present embodiment.
FIG. 2 is a block diagram conceptually illustrating the overall configuration of the PET apparatus according to the present embodiment.
FIG. 3 is an explanatory diagram of blank measurement and transmission measurement using a rotational calibration source.
FIG. 4 is an explanatory diagram of a calibration radiation source.
FIG. 5 is an explanatory diagram of detector sensitivity correction.
FIG. 6 is an explanatory diagram of detector sensitivity correction.
FIG. 7 is a flowchart illustrating a procedure for scattering correction and image reconstruction.
FIG. 8 is a flowchart illustrating a procedure for scattering correction and image reconstruction.
FIG. 9 is a diagram illustrating the FRB method.
FIG. 10 is a diagram illustrating a first modification of the configuration of the detection unit and the slice ceptor.
FIG. 11 is a diagram illustrating a second modification of the configuration of the detection unit and the slice ceptor.
FIG. 12 is a diagram illustrating a third modification of the configuration of the detection unit and the slice ceptor.
FIG. 13 is a diagram illustrating a configuration of a detection unit of a two-dimensional PET apparatus.
FIG. 14 is a diagram illustrating a configuration of a detection unit of a three-dimensional PET apparatus.
FIG. 15 is a diagram illustrating a configuration of a detection unit of a slice scepter removable type three-dimensional PET apparatus.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Measurement space, 2 ... Subject, 3 ... Calibration source, 10 ... Detection part, 11, 12 ... Shield, 20 ... Slice ceptor, 30 ... Scepter rotation drive part, 40 ... Rotation position detection sensor, 51 ... Two-dimensional Projection data storage unit, 52... Three-dimensional projection data storage unit, 60... Data processing unit, 70. ₁ ~ D _N ... Detector, R ₁ ~ R _Ten ... Detector ring, S ₁ ~ S ₉ ... shield plate.

Claims (3)

A plurality of photon detectors each detecting photons flying from a measurement space including a central axis include a plurality of detector rings arranged on a plane perpendicular to the central axis, and the plurality of sets of detector rings are Detectors stacked in a direction parallel to the central axis;
Among the plurality of photon detectors constituting each of the plurality of sets of detector rings, a part of the plurality of photon detectors is arranged on the measurement space side so as to be rotatable around the central axis, and among the photons that have come in, the center A slice scepter that only allows passage of something substantially perpendicular to the axis;
When a pair of photon detectors among the photon detectors included in the detection unit simultaneously counts photon pairs, the slice ceptor is present on the measurement space side of at least one of the pair of photon detectors. Slice scepter position determination means for determining whether or not
When the slice sceptor position determination means determines that the slice ceptor is present on the measurement space side of at least one of the pair of photon detectors, the photon pair by the pair of photon detectors Two-dimensional projection data storage means for storing the coincidence counting information,
When the slice sceptor position determining means determines that the slice ceptor is not present on any measurement space side of the pair of photon detectors, the photon pair by the pair of photon detectors is determined. Three-dimensional projection data storage means for storing the coincidence counting information,
Two-dimensional projection data coincidence counting information is generated is stored by the two-dimensional projection data storage means, and, based on the three-dimensional projection data coincidence counting information is generated is stored by the three-dimensional projection data accumulating means Image reconstructing means for reconstructing an image representing a spatial distribution of the occurrence frequency of photon pairs in the measurement space;
A PET apparatus comprising: The image reconstruction means reconstructs the image based on a component having a low spatial frequency in the two-dimensional projection data and a component having a high spatial frequency in the three-dimensional projection data. The PET apparatus according to claim 1. A rod-shaped calibration radiation source parallel to the central axis is provided on the measurement space side of the slice ceptor, and the image reconstruction means is based on the two-dimensional projection data stored in the two-dimensional projection data storage means. The PET apparatus according to claim 1, further comprising a correcting unit that corrects the image reconstructed by the processing.

Images