Home About Browse Search

Clern, Rebecka, 2022. Methodologies for training detection dogs and scent sample storing. Second cycle, A2E. Uppsala: SLU, Dept. of Clinical Sciences



Dogs’ spectacular sense of smell has been used as a detection tool by humans in many areas. Their
olfaction was first used as a hunting tool, but dogs’ detection skills have been shown to also be
applicable in, for example, detection of drugs, explosives, and firearms as well as disease detection.
Different types of disease detection are possible due to the body’s ability to produce volatile organic
compounds (VOCs), which reflect the metabolic state of the body. Many illnesses have been shown
to cause the body to produce a distinct pattern of VOCs, and these molecules together make up a
specific smell. Sweat is a major source of VOCs, therefore sweat samples are used for disease
detection by dogs.
The first aim of this study was to examine what track design is most beneficial for detection dogs’
learning. This was done through evaluating the performance of two trained detection dogs in three
different track shapes, when training the ability to detect individuals infected with severe acute
respiratory syndrome coronavirus 2 (SARS-CoV-2) through sweat samples. All tracks consisted of
six consoles. In Track 1 the consoles were positioned in a straight line, Track 2 was a circular track,
and in Track 3 the six consoles were placed in a slightly curved line. The track design was changed
when the trainers discovered risk factors that could potentially lead to false results. The design of
the straight track caused the dogs to have a high level of curiosity which subsequently made them
skip the first couple of consoles in many searches. The design of the circular track made it possible
for the dogs to pick up on visual cues the handlers were unaware of.
Both dogs’ sensitivity was the highest in the first track (96.4% and 97.8% respectively), and the
lowest in the third track (62.9% and 82.6%). The sensitivity for Track 2 was 78.9% and 85.1%.
Similarly, the specificity was also the highest in the first track (99.3% and 99.8%) and the lowest in
the third track (94.3% and 98.5%). The specificity for Track 2 was 96.6% and 99.2%. However, the
results may have been influenced by prerequisites that differed between the tracks, which makes the
results difficult to interpret.
The second aim of the study was to examine if Falcon centrifuge tubes are suitable as storage
containers for scent samples. This was tested through keeping 50 sealed Falcon tubes together in a
freezer for one week, out of which three contained pieces of natural rubber that give off a particular
smell. Three detection dogs trained to detect that smell searched through all sealed tubes. The results
of the study indicate that the scent did not escape the sealed containers as no dog could detect the
scent. Falcon tubes are therefore a promising alternative for storing scent samples since scent
contamination between samples would not be possible if the containers do not leak the scent. Further
studies would be valuable as this could potentially facilitate storing scent samples in a spaceefficient way.
Keywords: detection dogs, scent training, canine scent detection, search dogs, scent samples in a spaceefficient way.

Main title:Methodologies for training detection dogs and scent sample storing
Authors:Clern, Rebecka
Supervisor:Rönnberg, Henrik and Saellström, Sara
Examiner:Hanson, Jeanette
Volume/Sequential designation:UNSPECIFIED
Year of Publication:2022
Level and depth descriptor:Second cycle, A2E
Student's programme affiliation:VY002 Veterinary Medicine Programme 330 HEC
Supervising department:(VH) > Dept. of Clinical Sciences
Keywords:detection dogs, scent training, canine scent detection, search dogs, scent samples
Permanent URL:
Subject. Use of subject categories until 2023-04-30.:Animal husbandry
Animal diseases
Deposited On:24 Aug 2022 09:57
Metadata Last Modified:25 Aug 2022 01:00

Repository Staff Only: item control page


Downloads per year (since September 2012)

View more statistics