Trial for Power Spectrum Analysis in Various Injection Timing for Contrast Medium

  • Takatsu Yasuo
    Department of Radiology, Osaka Red Cross Hospital Division of Health Sciences, Graduate School of Medical Sciences, Kanazawa University
  • Inoue Yasuyoshi
    Philips Electronics Japan
  • Miyati Tosiaki
    Division of Health Sciences, Graduate School of Medical Sciences, Kanazawa University
  • Nakatsuka Takayuki
    Department of Radiology, Kinki University Hospital

Bibliographic Information

Other Title
  • 造影タイミングの変化におけるパワースペクトル解析の試み
  • ゾウエイ タイミング ノ ヘンカ ニ オケル パワースペクトル カイセキ ノ ココロミ

Search this article

Abstract

Frequency analysis is a method of measuring the spatial frequency component in images. We attempted to analyze and assess frequency analyses of images of various contrast medium injection timings using a normalized integral power spectrum. In magnetic resonance imaging, the k-space order was linear and sequential. During the twenty-second scanning time, we injected contrast medium at 3 ml/s and varied the duration of the injection from zero to four seconds, four to eight seconds, eight to twelve seconds, twelve to sixteen seconds, and sixteen to twenty seconds. We then analyzed the images. The rates of change and area values were calculated, and a dendrogram of area value by cluster analysis was made. The element of high frequency area’s power values rose when the contrast medium was in the high-frequency area, and the element of low frequency area’s power values rose when the contrast medium was in the low-frequency area. These data lines crossed at 0.092–0.115 cycles/mm. Five time durations were identified: these were a function of rate of change and area value. Frequency analysis of injection timing for contrast medium using the normalized integral power spectrum can thus be applied; this method can thus be employed as a method of examining injection timing for contrast medium and choosing a k-space order of optimization.

Journal

References(21)*help

See more

Details 詳細情報について

Report a problem

Back to top