You wouldn’t dream of downhill skiing without a rating system. The beginners can stay off the tough slopes and the pros get to debate whether the black run of Colorado’s Arapahoe Basin Pallavicini is tougher than any double diamond at Breckenridge. We need the same rating system in image analysis. Many of the double-diamond examples in image analysis are not only difficult to do once on a tissue, but present extreme problems in scaling up to the kind of volumes common in the pharma and medical device industries.
Below is an example of some applications in our experience and how they rate. Even with these there is lots of debate within Flagship. While oncology image analysis generally gets harder the closer you get to real patient samples, Frank Voelker strongly protests (and rightly so) in his talk on Xenograft Image Analysis that even xenografts with their artificial cancer environment are not greens and maybe not blues. But we learn more through disagreement, so we are starting to rank all our applications in this way.
I think that Coralie Apfeldorfer’s dendritic spine density in Alzheimer’s work that she presented at Path Visions 2010 is the toughest double-diamond I have seen. Hats off to her!
Another helpful aspect of difficult rankings is you know what equipment to use. Flagship uses a number of in-house developed software and commercial solutions.
So let the debates begin — I will still prefer skiing Pallavincini late in the season when the cornice is huge than finding glomeruli 🙂
