I still haven't gotten used to the idea that I have a camera in my pocket at all time -- my smart phone. If and when I get my head around this idea, I will be confronted with a new capability. I can use this camera to better understand the world around me. This concept was explained in a recent article from which I quote below (see: Seeing the World Around You Through Your Phone):
The best new thing to hit smartphones is augmented reality. These apps, like Goggles (free, for Android phones) and Layar (free, for Android and Apple devices), are like space-age glasses. Point your smartphone in any direction and look through the camera viewer, and it will reveal information about what it sees. Want to know the artist responsible for the print hanging in that restaurant? ....The last selling price of that house up ahead?....I opened Google’s Goggles on a MyTouch 3G Slide, from T-Mobile, and pointed it at a reproduction of a vintage poster I’ve had hanging in my living room for years. The poster is in French. The app snapped the poster’s photo and I watched it quickly scan the text and image. A moment later the app offered a translation of the text and links where I could find the poster online....Goggles tends to perform better when pointed at text, bar codes, labels and other inanimate objects that get a fair amount of attention on the Web....[A] Google spokesman, said Goggles was especially useful when traveling, and seeking information on a local landmark...without having to pore through a guidebook....Layar, in the meantime, is equally slick, and in many ways even more useful than Goggles....Here’s how it works: Opening the Layar app triggers your device’s camera. From there, you choose which “layer” to view. If you’re curious about real estate values, for instance, you can choose Trulia’s layer. The number of layers is growing fast, as companies discover that it’s a great marketing vehicle....If you’re hungry, there are numerous restaurant-finding layers. From the icons on your screen, you can see which restaurant is closest, and you can hit a button and get more information and directions to the place if it’s not in sight....Layar will also function in the dark, since it works as much from your GPS position as from the image in front of you. Layar is free, but not all layers are free.
This discussion of Goggles and Layar provides a suitable segue to the idea of image search and interpretation in digital pathology (see: Technologies and Tools to Search Images with Images; "Eminence-Based" Surgical Pathology and the Digital Pathology Department). A pathologist selects a field of interest from a current case and asks the computer to search an image database and find identical, diagnosed images. The computer then provides diagnostic and therapeutic suggestions for the case in question. Similar image analysis technology has already gained a foothold with the use of algorithms to score immunohistochemistry (IHC) stained tissue (see: Putting Some Numbers to Digital Pathology Adoption Trends by Pathologists). Regarding the use of such computer algorithms, there is common agreement that this is only a tool to assist a pathologists in interpreting a stained slide and that he or she is the final arbiter of the results.
Regarding image search in surgical pathology, I think that such an approach to diagnostic interpretation will soon become a widely accepted practice, again as a tool to assist the pathologist in the interpretation of a case. What we now lack, however, is a curated histopathology image database. By this I mean an database of pathologic images that have been reviewed and vetted by experts as to the diagnoses used to tag them. If you perform a Google search for the terms pathology, image, and archive today, you get 278,000 hits. Who's to know which of these archives provide universally reliable diagnoses linked to the images. Once such a curated database is created, the next step would be to link it to reliable clinical databases, allowing queries about optimum treatment for patients with the lesion under study across time.
Comments