Spiral trajectory design: A flexible numerical algorithm and base analytical equations

James G. Pipe, Nicholas R. Zwart

Research output: Contribution to journalArticlepeer-review

27 Scopus citations


Purpose Spiral-based trajectories for magnetic resonance imaging can be advantageous, but are often cumbersome to design or create. This work presents a flexible numerical algorithm for designing trajectories based on explicit definition of radial undersampling, and also gives several analytical expressions for charactering the base (critically sampled) class of these trajectories. Theory and Methods Expressions for the gradient waveform, based on slew and amplitude limits, are developed such that a desired pitch in the spiral k-space trajectory is followed. The source code for this algorithm, written in C, is publicly available. Analytical expressions approximating the spiral trajectory (ignoring the radial component) are given to characterize measurement time, gradient heating, maximum gradient amplitude, and off-resonance phase for slew-limited and gradient amplitude-limited cases. Several numerically calculated trajectories are illustrated, and base Archimedean spirals are compared with analytically obtained results. Results Several different waveforms illustrate that the desired slew and amplitude limits are reached, as are the desired undersampling patterns, using the numerical method. For base Archimedean spirals, the results of the numerical and analytical approaches are in good agreement. Conclusion A versatile numerical algorithm was developed, and was written in publicly available code. Approximate analytical formulas are given that help characterize spiral trajectories.

Original languageEnglish (US)
Pages (from-to)278-285
Number of pages8
JournalMagnetic Resonance in Medicine
Issue number1
StatePublished - Jan 2014


  • MRI
  • spiral
  • trajectory

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Spiral trajectory design: A flexible numerical algorithm and base analytical equations'. Together they form a unique fingerprint.

Cite this