Nitya Talreja #3: Cancer Trials Part 1

Three dogs at the Penn Working Dog Center are currently being trained to detect ovarian cancer at various stages. Those three dogs are McBaine (spaniel mix), Ffoster (yellow lab), and Tsunami (german shepherd). As you can see, the list above is not specific to any breed of dog, which concludes that the ability to detect cancer (and other diseases) is not based on the dog's senses, but on it's ability to be trained.


The cancer training trials were conducted promptly at 9:30 am every morning and ended around 11:30am. All three dogs were trained on a scent detection wheel such as this one here:




There were 12 ports on the wheel and the cancer team would watch all 10 trials from a dark room where the dogs could not see us (to prevent any distractions). Three of the ports were filled with blood plasma samples in which there were malignant, benign, and normal samples stored in glass containers from three different people. The other 9 ports served as controls and were filled with either blank glass containers or miscellaneous pieces of plastic or rubber bands.  All the ports were covered with a metal case so that exterior variables (such as human saliva or dog drool) do not interfere with the scent of the sample. I was specifically told to treat the 3 experimental samples as powered sugar and to be extremely careful when handling them to prevent its scent from spreading. Data was collected on the cancer trials through a flase/true negative/positive basis. For instance, if a dog signaled at the malignant cancer port, it would be labeled as a true positive. If a dog signaled at a benign, normal, or control sample, it would be called a false positive. If a dog passed a benign or control sample, it would be a true negative. And, so on. It was a bit tricky for me to grasp at first, but it did not take long for me to get accustomed to.

Which ever dog began first, the dog had to be walked before and after each cancer session to prevent any accidents and limit and variables of the data. Also, the cancer team had to clean each of the ports between every dog with rubbing alcohol (70% concentration) and after cancer trials were completed each day.

As far as the actual people on the cancer team concerns, there had to be one person collecting data, one person spinning the wheel (with a stick) in-between each of the 10 sessions, and one person clicking when the dog sat at the correct cancer port. I was lucky enough to have tried all three tasks, and from experience, I can say that collecting data is by far the most difficult. You have to be alert at all times and observing the dogs and all of their reactions. But just in case, all cancer sessions are video recorded to refer back to in times of doubt.

I am genuinely enjoying my time at the Penn Working Dog Center and cannot believe my time here is almost over! I have had a great time with Dr. Lorenzo and am going to miss the puppies so much! Shortly after I had been accustomed to the cancer trials, Lorenzo and the rest of his team had abruptly changesdit to improve the study. Stay tuned for the interesting change in task :).