George McNamara

Spectra links

mpmicro (zipfile)

Temporal Area Map & Histogram For Analysis of Cell Motility and Chemotaxis (TAM webpage)

Crusade for better micrographs

prism

Boswell-McNamara Fluorescence Spectra Web Site (introduction page)

Tiki God

Tiki Goddess

See also the tiki_goddess website

Geo's favorite places

Geo's EXE's (zip collection)

NIH Biosketch

 

 

 

Crusade for Publishing Better Light Micrographs

Many scientific journals publish light micrographs that are awful. Since micrographs are data, I wonder what other data in those papers (and journals) are also bad. I subscribe to both Science and Nature, and both routinely publish light micrographs that bad. With respect to brightfield histology images, after I received an awful issue on prostate cancer histology (guest edited by someone who I hope never guests edits any journal I -- or anyone else - subscribe to), I contacted the editor. He agreed with me that the published figures were bad. The upshot was I agreed to be "QA editor" of the journal in exchange for an article on how to get the color balance correct in histology images. To make a long story short: set up the microscope correctly each time you use it, white balance the camera, operate the camera in manual color mode, acquire all images with identical (correct!) settings, and acquire all images for an experiment under identical (correct!) conditions. The complete article is here:

http://home.earthlink.net/~geomcnamara/McNamara2005JoH28n2pp81-88.pdf

I thank Dr. James Hendricks, Editor in Chief of Journal of Histotechnology, for permission to post the PDF. 

Light microscopy publication guidelines:

  • Transmitted light brightfield micrographs of histology, immunohistochemistry (IHC), and chromogenic in situ hybridization (CISH) tissue sections, hematological blood smears, Pap smears and other clinical cytology specimen microscope slides specimens, should be acquired with correctly setup microscope and camera. If you are not setting up the microscope in Kohler illumination, get training from someone who does use the microscope correctly. Watson & Crick were once accused of "practicing biochemistry without a license" (ok, they turned out alright, though Rosalind Franklin's crystallography images were the key). If you use a light microscope and don't set up the microscope correctly, please go practice biochemistry instead. 
  • The transmitted light micrographs mentioned above should almost always be acquired using white light. Sometimes you may need to put a color balancing filter ("CB"), blue daylight filter, or otherwise adjust the illumination color balance. If you need to attenuate the intensity, use neutral density filters (available from your microscope representative, Chroma Technology, Melles Griot, Edmund Optical, and photographic supply companies). Do not reduce the numerical aperture control - this by definition means the microscope is no longer in Kohler illumination (if your microscope does not have a numerical aperture control, buy one that does). The numerical aperture control is for optimizing the trade-off between resolution (more is better) and glare (more is worse). 
  • My recommendation for the tungsten-halogen lamp intensity is simple and reproducible: full power. To attenuate the light level, use neutral density filters (see previous paragraph for suppliers). Bulbs last well over a year. Keep a few on hand. You can either buy them for $20 from your microscope dealer, or for $3 from a local hardware store. The only advantage of buying from the dealer is when they visit, you can get refresher training on Kohler illumination. 
  • Learn to use the camera correctly. White balance at the start of the session, then use manual exposure mode for all your images. If you use different magnifications, you will probably need to change exposure time and/or neutral density filters. Most digital cameras have plenty of sensitivity and dynamic range, such that you can adjust the neutral density filter(s) to be comfortable to your eyes (I think of this as "optical ergonomics"). Document the camera and microscope settings you use.
  • Try to do all staining and all image capture in single sessions.
  • When organizing micrographs for publication, keep the controls and experimental images together in sets. You need to publish the contemporary control for whatever micrographs you use. Selecting for publication a control from a different staining and/or image session is fraud. 
  • I suspect that most published micrographs are "exemplary", "best of", or, 'the only one we took", or "the only one that fit our hypothesis" (I call the latter two categories, "N=1 experiments"). If you are putting together figures, and you select for publication a micrograph based on any of these categories, at least be honest to the reviewers and editor and say so (hopefully they'll tell you to go back and collect data correctly ... even better, your coauthors should tell you ... best of all, your inner super-ego should tell you). What you should be publishing are representative micrographs. That means you need to acquire sufficient images to document/quantify the experiment. your specimen and images should be good enough that any of the micrographs can be used. In fact, if you can only publish one micrograph per treatment group, use a random number generator to pick which one (for our purposes, Excel's is fine. If you have 20 micrographs per treatment group, use =RANDBETWEEN(1, 20) for this). 
  • Please take the time to size your micrographs to that the reader does not need a microscope to see what you are claiming. If a "picture is worth a thousand words" it ought to worth close to a thousand words of space!
  • Most scientific journals have Supplementary Online Material (SOM). Instead of publishing one "exemplary", or even "representative" micrograph, take advantage of the SOM to publish many or all of the micrographs.  
  • I would need another website to go into all the sins possible in biomedical fluorescence microscopy images. Rather than do that here, I'll just make two points: 
    • Any results that claim colocalization of two molecules on the basis of a red+green=yellow, is full of itself. At a minimum, the published images should include each source image in monochrome, the "red+green=yellow" color micrograph (I call this the "money shot", since the reader gets what they pay for), a colocalization scatterplot (also called cytofluorogram) and statistical test (correlation coefficient, etc). The authors also need to document that their imaging system was setup such that the detector offset was above zero for each channel and no pixels saturated. If they are using antibodies, they need to specify in the Methods (and ideally in the figure caption) exactly what they included to block Fc receptors and non-specific binding. They should also publish in the paper or SOM the negative controls.
    • Single color fluorescence micrographs should be published in monochrome, not color. I realize color is sexy. However, the goal is supposed to be to transmit scientific information to the reviewer and reader. The human eye is sensitive to about 60 gray levels, somewhat less than this in green and red, and much less in blue. This combined with most printers doing a bad job at printing the "true blue" of a DAPI, Hoechst 33342 or Hoechst 33258 fluorescence, mean that publishing green fluorescein, red Cy3B, and blue DAPI single channel images just results in masking much of the original micrograph's content. 

While not a publication guideline, a crucial optimization for light microscopy is to use the microscope correctly. Take the time to study ergonomics and apply best ergonomic practices to your -- and your colleague's -- use of the microscope. Your Human Resources Department can introduce you to your employers ergonomic expert. Take the time to work with them to use correct posture, get the right kind of seating, arm rests, keyboard 7 mouse supports, and computer monitor stand. Whether you are the Director of a multi-million dollar microscope facility, Chief of Pathology of a famous medical school or hospital, Principal Investigator of a research lab, histology/cytogenetics technologist/technician, staff scientist, postdoc, graduate student, or even a student taking a histology course, there is no excuse for bad seating and bad posture. If your supervisor and employer, or laboratory instructor and University/college/school, are unwilling to provide a proper work environment, I suggest finding one who does before you suffer an workplace/educational place injury.