K-space explorer
A pulse sequence built from real gradient events, and the path its gradients trace through k-space: k(t) = γ̄ ∫ G(t) dt. Play it and watch each gradient move the k-space position.
K-space: the path so far and the samples taken
Image from the samples taken so far
Anatomy: Colin 27 Average Brain, high-resolution version 2008: the tissue classes of one normal adult's head, covering the head down to about the skull base. Copyright (C) 1993–2009 Louis Collins, McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University. Only tissue membership is stored; the simulator computes the signal.
Cite Colin27
- C.J. Holmes, R. Hoge, D.L. Collins, R. Woods, A.W. Toga, A.C. Evans: Enhancement of MR images using registration for signal averaging. Journal of Computer Assisted Tomography 22(2):324-333, 1998.
- B. Aubert-Broche, A.C. Evans, D.L. Collins: A new improved version of the realistic digital brain phantom. NeuroImage 32(1):138-145, 2006.
Anatomy: BrainWeb (McGill), subject 04. BrainWeb covers the head down to about the skull base.
Cite BrainWeb
- C.A. Cocosco, V. Kollokian, R.K.-S. Kwan, A.C. Evans: BrainWeb: Online Interface to a 3D MRI Simulated Brain Database. NeuroImage 5(4) part 2/4, S425, 1997.
- R.K.-S. Kwan, A.C. Evans, G.B. Pike: MRI simulation-based evaluation of image-processing and classification methods. IEEE Transactions on Medical Imaging 18(11):1085-97, 1999.
- D.L. Collins, A.P. Zijdenbos, V. Kollokian, J.G. Sled, N.J. Kabani, C.J. Holmes, A.C. Evans: Design and Construction of a Realistic Digital Brain Phantom. IEEE Transactions on Medical Imaging 17(3):463-468, 1998.
- B. Aubert-Broche, M. Griffin, G.B. Pike, A.C. Evans, D.L. Collins: Twenty new digital brain phantoms for creation of validation image data bases. IEEE Transactions on Medical Imaging 25(11):1410-1416, 2006.
The blue line is the path k(t) traced during this TR; the dot is where k is now. During the ADC windows the signal is sampled and k-space fills in. Idle time (the shaded breaks in the diagram) is skipped during playback, except in real time, which plays every millisecond of the scan. The phantom image is noise-free here so that the effect of the sampling stands out.
Slices in spatial order (click one to select it)
Each thumbnail is built from the same TRs as the main image, with the signal change its neighbours' pulses cause. All use one window and level, so a dimmer slice has lost signal. Labels give the slice number, its place in the acquisition order and the white-matter signal change. With several concatenations, P gives each slice's pass; the TRs shown play the selected slice's pass, and the slices of other passes are dimmed (an earlier pass is already acquired, a later one has not started).
Each gold curve is one slice's excitation profile (|Mxy| after its pulse), simulated with the Bloch equations; click a curve to select that slice. Slices are excited one after another within a TR, so a pulse for slice j also partly tips the edges of its neighbours. That is cross-talk: the blue curve shows the longitudinal magnetisation (Mz) left in the steady state just before the selected slice's excitation, and the dips at the neighbours' positions are the saturation they caused. Dips shrink as the gap grows. The pulses are Hann-windowed sincs (time-bandwidth 4), as scanners typically play windowed pulses, and with them cross-talk lowers a slice's signal: most for long-T1 tissue such as CSF at short TR and high flip angles (8 slices of 5 mm, gradient echo 90°, TR 300 ms, no gap: CSF −41 %, WM −20 %), and much less for spin echo. With gaps up to 20 % it never raises the signal. Only at wide gaps (about 30–70 %) can a 90° gradient echo or EPI gain a little, at most about +8 % (gradient echo, CSF, short TR) and about +1 % for EPI: just outside each slice the 90° profile keeps a small transverse part out of phase with its centre, and a neighbour's pulse that saturates it cancels less of the signal. The information panel shows the change as a signed percentage. Dashed red curves (spin echo and fast spin echo) are the refocusing pulses' profiles. Not modelled: magnetisation transfer, stimulated echoes (SE and FSE), and relaxation during the pulses. Interleaved order starts with slice 1 (some scanners start with slice 2 when the number of slices is even).