Useful Papers - Slice Timing Correction
Henson et. al (1999), "The slice-timing problem in event-related fMRI," NeuroImage 9, S125 PDF
Summary: Describes the slice-timing problem for event-related experiments, including very nice and dramatic pictures of data without the correction and without it. Describes both interpolation and temporal derivative ways of solving problem.
Bottom line: The basic 1-page slice-timing paper: shows correction is needed and that interpolation works well to correct it.
Van de Moortele et. al (1999), "Latencies in fMRI time-series: effect of slice acquisition order and perception," Nuclear Magnetic Resonance in Biomedicine 10, 230-236 PDF
Summary: Demonstrates slice-timing problem in block experiment, and shows very regular effect spread within single clusters. Demonstrates effectiveness of interpolation even on activations within clusters.
Bottom line: Shows that slice-timing can be useful even in block experiments; shows empirically slice-timing problem happens within clusters.
Van de Moortele et. al (1998), "Slice-dependent time shift efficiently corrected by interpolation in multi-slice EPI fMRI series," NeuroImage 7:4 (Supplement), S607 PDF
Summary: Early paper suggesting slice-timing correction algorithm now used by all major programs; uses it in early/late onset contrast.
Bottom line: Yup, interpolation correction works; if you want timecourse information, you should do it.
Calhoun, V. et al (2000), "Improved fMRI slice timing correction: interpolation errors and wrap-around effects," Proceedings, ISMRM, 9th Annual Meeting, Denver, CO, 810 PDF
Summary: Highlights pitfalls of interpolation correction - describes the shape and profile of interpolation errors and talks about improper interpolation at beginning and end of timecourse.
Bottom line: Consider low-pass smoothing your data if interpolation errors are a big deal (but low-pass smoothing probably removes more signal from your data than you'll get back from the error correction), and use an up-to-date program to do slice-time correction.