This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
docs:comparison_with_confocal [2015/06/15 21:17]
Jon Daniels created
docs:comparison_with_confocal [2016/05/09 23:18] (current)
Jon Daniels
Line 1: Line 1:
 ===== Comparison of diSPIM With Confocal ===== ===== Comparison of diSPIM With Confocal =====
-  Confocal has much more photobleaching/phototoxicity than SPIM (i.eSPIM uses the excitation light much more efficiently) +The following comparison was written with diSPIM in mind, but the main points apply to all types of SPIM/LSFM. 
-      * In general for X slices you need X times as much light for confocal as you do for SPIM.  The low light dose is the main reason people use SPIM+ 
-      * [[http://www.nature.com/nbt/journal/v31/n11/fig_tab/nbt.2713_SV2.html|Supplementary Video 2 of Wu et al.]] is a good demonstration; note how quickly the SDCM kills the cells+Spinning disk confocal (SDCM) is just a massively parallel confocal implementation so it's faster but otherwise has the same characteristics as laser scanning confocal. 
-  * Confocal has better Z resolution than single-view SPIM but worse than dual-view SPIM after registration/fusion.  XY resolution is comparable. + 
-  Spinning disk confocal is just massively parallel confocal so it's faster but has the some characteristics otherwise+  **Light dose:** diSPIM has much lower dose and hence less photobleaching/phototoxicity than confocalReducing light dose is the main motivation for most SPIM users. 
-  * diSPIM is comparable speed (per view) to to spinning disk confocal (assuming you increase the laser intensity to make up for the ~3% open area of the spinning disks).+      * A rule of thumb is that you need X times as much light for confocal as you do for SPIM where X is the number of slices in your stack
 +      * [[http://www.nature.com/nbt/journal/v31/n11/fig_tab/nbt.2713_F3.html|Figure 3]] and [[http://www.nature.com/nbt/journal/v31/n11/fig_tab/nbt.2713_SV2.html|Supplementary Video 2]] of Wu et al. offer bleaching comparisons of imaging with diSPIM and SDCM.  Other papers such as [[http://dx.doi.org/10.1002/jbio.201200144|this one]] offer comparisons of light sheet with confocal which should apply to diSPIM
 +  * **Z resolution:** diSPIM after registration/fusion has best Z-resolution (same as XY resolution).  Next best is confocal, which in turn has better Z resolution than single-view SPIM
 +      * Using Bessel beams to create the sheet can improve Z-resolution for single-view SPIM ("optical sectioning"), but won't meaningfully change dual view resolution after registration/fusion. 
 +  * **XY resolution:** confocal and SPIM are comparable
 +  * **Speed:** SPIM (per view) and SDCM have comparable speed if SDCM laser intensity is increased to compensate for the ~3% open area of the disks
 +      * Camera readout speed bounds the maximum achievable frame rate.  For example, 512 pixels high ROI is 2.5 ms readout time for sCMOS.  Allowing 2.5 ms illumination time results in 5 ms total per image or 200 frames per second.  Using a cylindrical lens instead of a scanned light sheet can reduce camera overhead.