September 2004


Canine olfactory detection of cancer

Interesting “proof of principle” study from a group in the UK. Here is the article:
Canine olfactory detection of cancer

In this study, six dogs were trained to discriminate between urine from patients with bladder cancer and urine from diseased and healthy controls. The authors claim that the dogs successfully discriminated urine samples from patients with bladder cancer 41% of the time. They believe error was introduced due to an inability to identify the chemical composition of a particular cancer odour signature. Since there is no known “pure source” to train on, they expected their success rate would be less than that of dogs trained for explosives or drugs. They tested one cohort of dogs on dry samples and one cohort on wet samples. The dogs trained with wet samples performed significantly better suggesting that volatile chemicals are part of the signature odor of bladder cancer.

Sep 25 2004 05:25 pm | K-9 SAR and Scent and Search Theory | No Comments »

aged track in mount airy

We went back to mount airy last night to run some tracks we lay two nights before. BB is still not where I want him with urban trailing, so I chose the track in the park. It’s a busy park with tons of people and dogs, but about 3/4 of the track was on mowed grass. The temperature was around 75% with a purported 90% humidity. We had a very slight breeze of 1-2 mph. I was very surprised to find that BB ran the trail with no sense of confusion about where the trail was. He was on it very well, though his nose does not stay glued to the ground the way Titan’s does. He runs head low, but off the ground.

The original track is in yellow and BB’s trail in red. What I really need are some icon conventions and standard graphics for annotating these trails. For example, I wouldn’t have to label everything if I had icons for things like gazebos and portajohns. Also important track features such as original track vs. trail, wind direction, distratactors, etc. could be standardized such that a tool palette would associate meta-data with vector graphics. And wouldn’t it be way cool if distances could be calculated directly from a GPS track overlay on my image? I was too lazy to carry a GPS last night. It takes significant effort to take that track data and overlay it on an image.

Sep 25 2004 04:07 pm | K-9 SAR and Training log | No Comments »

Next Page »