PROBLEM TO BE SOLVED: To provide a high-speed bed positioning method with the six degrees of freedom concerning a radiotherapy.SOLUTION: X-ray fluoroscopic images are imaged from a plurality of directions, so as to generate DRRs corresponding to the imaging directions of the respective images, and to temporarily generate a patient deviation amount with the six degrees of freedom by an optimization algorithm. The out-of-plane rotation component of the deviation amount is calculated, so as to reproduce the DRRs when a change from the DRR generation is larger than a regulated value, and to approximate the deviation by the planar conversion of the respective DRRs when the change is smaller. The deviation amount is projected to the planar conversion components of the respective DRRs, and then, the DRRs are planarly converted when the change is performed from the DRR generation by a regulated position or more. The total sum (a total similarity degree) of similarity degrees between the converted DRRs and the X-ray fluoroscopic images is calculated, so as to update a search range by the optimization algorithm to increase the total similarity degree. When the search range satisfies a convergence condition, the calculation is terminated. When not satisfying the convergence condition, a temporary deviation amount is generated by the optimization algorithm again, thereby repeating the calculation. Thus, the problem is solved.COPYRIGHT: (C)2011,JPO&INPIT【課題】放射線治療における、高速な6自由度のベッド位置決め法を提供する。【解決手段】本発明では、複数方向からX線透視画像を撮影し、各画像の撮影方向に対応するDRRを生成し、最適化アルゴリズムにて6自由度の患者ずれ量を仮生成する。ずれ量の面外回転成分を計算し、DRR生成時からの変化が既定値より大きい場合にDRRを再生成し、小さい場合にずれを各DRRの平面変換で近似する。ずれ量を各DRRの平面変換成分へ射影し、DRR生成時から既定位置以上変化した場合にDRRを平面変換する。変換されたDRRとX線透視画像の類似度の総和(総類似度)を計算し、総類似度を増大させるよう最適化アルゴリズムで探索範囲を更新する。探索範囲が収束条件を満たす場合に計算を終了し、満たさない場合、再び最適化アルゴリズムで仮のずれ量を生成し、計算を繰り返す。これにより上記課題を解決する。【選択図】 図2
