Page History

Papers you should read before you start:

• Fiala JC (2005) Reconstruct: A free editor for serial section microscopy. J Microscopy 218:52-61. (PDF)
• Fiala JC, Harris KM (2001) Cylindrical diameters method for calibrating section thickness in serial electron microscopy. J Microscopy 202(3):468-472. (PDF)
• Harris KM, Spacek J, Bell ME, Parker PH, Lindsey LF, Baden AD, Vogelstein JT, Burns R (2015) A resource from 3D electron microscopy of hippocampal neuropil for user training and tool development. Sci Data Sep 1;2:150046. PMCID: PMC4555877. (PDF; Tables)
• Harris KM (1994) Serial electron microscopy as an alternative or complement to confocal microscopy for the study of synapses and dendritic spines in the central nervous system. In: Three-dimensional confocal microscopy: volume investigation of biological specimens (Stevens JK, Mills LR, Trogadis JE eds), pp 421-445. New York: Academic Press, Inc. (PDF)

Steps for starting a new series from scratch:

1. Have a Dendrite Analysis Spreadsheet ready and record all pertinent information as you work.
2. Make a new series & import images
3. Calibrate pixel size
4. Align EM images & propagate
5. Calibrate section thickness

Step 1: Make a new series & import images

Please see the Starting a New Series PDF for complete details.

Step 2: Calibrate pixel size

Please see the Calibration Protocol PDF for complete details. 

WARNING: CALIBRATION MUST BE DONE BEFORE YOU START TRACING.

Step 3: Alignment of serial EM

• Ch 1 & Ch 10 of the Reconstruct Manual discuss alignment.
• Alignment works by aligning identically named stamps on adjacent sections when they are selected. Stamps must be selected in order to be aligned.
• Use a different name/color for the adjacent sections to avoid confusion. You can use a minimum 3 stamps, recycling through the same name/color, without any overlap.
• If you are working with large-field tSEM images, you might want to use FIJI/TrakEM2 to align the images before having them imported to RECONSTRUCT.

1. In order to be aligned, the sections must be unlocked. Section > Modify > Unlock. To lock a section after aligning, Ctrl-L.
2. When choosing a “golden section” to base all the alignments on, choose a central section with a zone (of several sections) that is free of flaws.
3. Name the stamps Align_$+. Cross-sectioned mitochondria work best. Choose different spots to place stamps each time. Don’t pick the same mitochondria on each section—don’t overlap the stamps. Use a square or cross stamp—more correspondence points.
4. After making the alignment stamps, select the stamps: Trace > Align Section > Linear.

Step 4: Calibrate section thickness

images 

Please see the Reconstruct Alignment Protocol. (PDF, Word)

Protocol for TrakEM2 alignment is in progress... coming soon!

Step 4: Calibrate section thickness

Various Alignment Notes That Need to be Organized...

(1) KH ca. 2010:

• Zoom in close enough that the stamp fills the mitochondrion.
• Don’t pick the far edges—they’re not going to align anyways. Choose center and corners (but not far corners; don’t go too close to the edge).
• Linear.
• If you move away from it at all, you can’t un-do it. You can only do space bar to blend. If you scroll any more, you can’t un-do.
• Traces must be selected in order to align. If they’re all de-selected, you can’t choose Align (it will be grey and you can’t click on it).
• Start in the middle.
• Ctrl-L to lock the section after it’s aligned.
• Propagating—wait till after done aligning. Section > Movement > Propagate

(2) JBB 2012:

• For large fields, work with alignment clusters – subregions.
• Don’t align to a distorted section – go back to a good section.
• Lock your first section – Ctrl-L.
• Trace > Align Section > Linear.
• For sections way out of alignment, do an initial general alignment.

Alignment title: “align_##_$+“ (where the first number is the cluster, and the second number is the mito)
Folder title: “WHKYG” / → “WHKYG D01 D02 D03”
              → “WHKYG D04 D05 D06 D07”    
              …
Method

1. Lock central section (ctrl + L, or “section > list sections > modify > lock”)
2. Place alignment traces on >4 cross sectioned mitochondria. Preferably mito within-or-close to your dendrites of interest.
3. Place alignment traces on identical mitochondria on the adjacent section. It’s helpful to use the “/” key to switch between images when placing stamps.
4. Align with “Trace > Align section > Linear”
5. Check alignment quality by blending with the “space bar”
6. Lock aligned section.
7. Repeat

Tips

• Always endeavour to align sections that do not have major flaws. Example: if section 2 is bisected by a critical flaw,  do your best to align section 2, then use the section list to align section 3 to section 1.
• “Section > Movement > Propagate” can be used to move multiple sections simultaneously. Example: if sections 3-200 are uniformly misaligned, perform conventional alignment on sections 1-2, and propagate movement to sections 3-200

(3) DJW:

1. Start at the central section.
2. Stick to big cross-sectioned structures (i.e. mitochondria are good to use), and try to put the marker on the same place on both sections.  Never use the same structure in pairs, or you could falsely straighten out the field. You will use the 'stamp' tool.
3. Use the Square and + sign stamps because they have more points to align, more useful than say a circle shape.
4. Example, starting on section 103, stamp on 103 then up to 104 (can only lay stamps in 1 direction) for 5-15 points (depends on size of EM)
5. Click Align, Linear to pull 103 into alignment with 104.
6. Check how good the alignment looks each pair at a time by blending with the space bar, the edges will typically not align well.
7. Best to do 'Linear' alignment because the others will distort the image.  Deformal - is useful if something is tilted, Quadratic - only if a section has a lot of distortion, this will force your alignment markers to align perfectly.
8. Then move to align 104 to 105, and so on until you reach the end of the series.
9. At the very end of the alignment, page through the entire series on 'blend' to see if anything seems out of alignment.  Spot repair the misalignment and propagate thru the rest of the sections as needed.  Example, sections 1-22 aligned beautifully and 23 - 61 were also good, but something weird happened between 22 and 23.  Fix the misalignment and propagate back from 22 to 1.
10. Try to align around folds, tears, and stretches.  Use Quadratic as need to force the mid section into alignment.  As in, align section 103 to 105 because 104 has the tear.  Then use quadratic to force 104 to align to 105.
11. Don't begin to trace dendrites, synapses, axons, etc until you are happy with the alignment.