In the April 1999 issue of Phytopathology News, Larry Stowell argues that the use of photographs and digital images as the sole source of information for a plant health diagnosis is poor practice (view article). He then lists many negative consequences of incorrect diagnoses and cites shrinking budgets as the reason that university extension programs have adopted this technology.
Somehow Stowell has missed the message. Most of his arguments would be valid if not for a single erroneous assumption: that images are used as the sole source of information in disease diagnosis. Nothing to this effect was stated or implied in our published abstract (Phytopathology 88:S39) or in the oral presentation delivered at the 1998 APS Annual Meeting. Similarly, no evidence of such a statement is found in Briere's abstract (Phytopathology 88:S11) or oral presentation nor was evidence detected in the symposium by the University of Georgia.
Digital images are new and powerful tools in disease diagnosis. Their use is not the result of shrinking budgets but rather of increased funding to explore methods that improve the speed and accuracy of current methods.
At North Carolina State University, digital images are typically accompanied by live samples and appropriate supporting information. The advantages of digital images are tremendous. Images allow the diagnostician to preview the sample before it arrives at the clinic (often in a deteriorated state). Based on this information the diagnostician can provide tailored information on which plant part to sample, how much sample to provide, appropriate packaging, etc. Images can be of whole fields, plants too large to submit as samples (e.g., trees and shrubs), or microscopic structures. Often, a telephone conversation occurs between the diagnostician and sample collector while images are viewed simultaneously by both parties as e-mail attachments or on the Internet. Not only does this result in improved diagnoses, but the sample collector receives training in plant pathology above and beyond what was possible without this technology.
In cases where a digital image is not accompanied by additional information, the diagnostician would be unable to go beyond the level of certainty that the image provides, which is no different than what happens with poor specimens. Thus, in terms of the ethics and professionalism involved, nothing has changed. A digital image does not give plant pathologists permission to be unethical or unprofessional, and it never will.
Stowell also argues that he "cannot imagine a situation where an image alone would be suitable for diagnostic purposes…" There are many diseases that possess characteristic symptoms that are quite diagnostic. Diseases such as powdery mildew, root knot, mosaic, etc., are readily recognized by skilled diagnosticians familiar with the host and its common diseases, especially for a given area and season.
Analogously, these same arguments must have been raised when APS began publishing the Compendia series full of color photographs. Some plant pathologists must have shuddered at the thought of untrained people armed with a compendium, making all kinds of blunders by relying on small photos to diagnose diseases. And yet, despite this possibility (and it has surely occurred), the compendia are tremendous tools in the hands of trained professionals.
One drawback to the use of digital images for plant disease diagnosis is poor-quality images, but they are the result of poor photography - something that will always be with us. Another problem may result from a dramatic rise in number of samples. But these problems in no way "ignore biological systems" or "ensure inferior information" as Stowell suggests.
The art and science of disease diagnosis will always be an art and a science. The technology that allows us to exchange digital images has increased our ability to diagnose diseases rapidly and accurately. It does not replace existing methods but rather adds a powerful new component to them. It is a tool that is here to stay.