You are here

Functional Imaging to Develop Outstanding Service-Dogs (FIDOS)


OBJECTIVE: This effort will capitalize on first-of-its-kind neural imaging feasibility work; demonstrating functional brain activation in unrestrained dogs in response to handler cues. The objective of this effort is two-fold; first, to optimize the selection of ideal service dogs, both in operational military and therapy environments, and second, to use real-time neural feedback to optimize canine training, shortening training duration, reducing costs, and increasing learned responses. DESCRIPTION: Military working dogs are used in a variety of operations, including bomb detection, search and rescue, and drug interdiction. Service dogs are also utilized in clinical settings, as therapy dogs to mitigate symptoms of Post-Traumatic Stress Disorder (PTSD) and Traumatic Brain Injury (TBI) in returning service members. Certain dog breeds, ideal for this work, are rare (a limiting resource) and variability in dog training and performance limits their effectiveness in operational environments. Better selection and screening of service dogs would ensure that the best dogs, with optimal motivation and trainability are selected for service. Current training regimens are driven by classic behaviorism, involving simple reward/punishment conditioning. They are time-intensive and costly, costing more than $20,000 per dog-human pair (in clinical animal-human work pairs). Therefore, this program will address these two problems by 1) providing quantitative means for selecting service dogs for training and 2) providing quantifiable evidence-based methods for optimal canine training techniques. Canine training paradigms require a cognitive revolution to take advantage of recent advances in brain imaging. A first-ever functional Magnetic Resonance Imaging (fMRI) study in awake, unrestrained dogs (Berns et al., 2012) confirmed that the dog brain reward system in the caudate nucleus reacts to a primed reward hand signal. These results provide a first-ever window into the brain of man"s best friend, providing a glimpse of how dogs functionally process human trainer signals and to what extent different brain networks are activated by these signals. In both the operational and clinical settings, better canine selection and screening methods would reduce training time and costs and result in more effective service dogs. Now, with this state-of-the-art canine neuroimaging tool in hand, potential high-value service dogs could be screened based on their neural activation to specific handler training cues. The hypothesis is that dogs with greater activation in the caudate nucleus in response to handler cues will be faster and easier to train. These experiments would result in better use of scarce canine resources (e.g., the rare Belgian Malinois breed) and better service dogs. The mechanisms of canine TBI and PTSD therapy effectiveness are not well understood, and training and implementation methods are ripe for improvement. One hypothesis is that effective therapy dogs are better able to sense their owner"s mental state and emotions. Some research suggests that dogs may, in fact, possess robust"theory of mind"(the ability to attribute mental states such as beliefs, desires and attitudes, to another person). For example, dogs are known to follow human"s gaze and pointing (Kirchhofer et al., 2012) and show contagious yawning, more so with their owners than with strangers (Silva et al., 2012). This method of canine neuroimaging could be used to identify"brain hyper-social"dogs ideally suited to TBI or PTSD therapy work. Correlating the dog's brain activity within the caudate nucleus with neurophysiological markers of handler stress and anxiety could provide a screening tool for dogs ideally suited to therapy work. In addition to optimizing the selection of ideal service dogs, a canine neuroimaging tool could pave the way for revolutionized service dog training paradigms. The identification of caudate activation by Berns et al. (2012) demonstrates that canine-imaging studies could quantify the relationship between handler signal and intrinsic reward without reliance on a behavioral proxy. Currently, handlers must reward approximations of a desired behavior to teach a dog a desired task. However, associations can be made prior to behavioral manifestation. Monitoring brain activation in real-time could allow trainers to reward proper brain activation patterns indicative of associative learning. These methods could be used to quickly measure how effective a given training technique is and provide quantitative, evidence-based rationale for selecting superior training methods. They would also increase the speed and efficacy of canine training, useful in both operational and therapy domains. Advances in understanding the influence of dog training techniques on the canine brain 1) will enhance selection of highly-trainable work dogs and 2) will enable faster, cheaper, and more effective training of military work and therapy dogs. Overall, this project represents a radical new method of quantitatively measuring cross-species communication, coordination, and therapeutics. PHASE I: Develop a reproducible training method for imaging canines while awake and unrestrained. Previous work involved training only two dogs and was trial-and-error based (Berns et al., 2012). A more stream-lined method will be needed for quickly and reliably training service dogs. The small business"s expertise in dog training will be particularly crucial for this objective. The large size of some training dogs also presents a technical challenge for positioning in the scanner. Equipment modifications will be completed in Phase I to ensure proper access to the scanner and canine comfort in the experiments. Phase I deliverables will include 1) a technical report and brief describing the training method and 2) a set of experiments demonstrating proof of concept for scanning reward-related brain activity in the service dogs. PHASE II: Finalize and validate a training method for training large dogs to undergo fMRI scanning while awake and unrestrained. Establish performance parameters through experiments that produce reliable brain images of responses to trainers"hand cues, which can be used to test ideal military dog training methods. Develop, demonstrate, and validate protocols that combine canine imaging and simultaneous human neurophysiological measures in order to examine dog-human interactions. Apply these protocols to study what dogs are best suited for use in therapeutic situations and quantify the relationship between this suitability and clinical outcomes. Required Phase II deliverables will include a technical report and brief describing the training methods and findings from canine brain scans, as well as feasibility of use in future commercial and/or military applications. PHASE III: Law enforcement agencies train and use work dogs for many of the same operations as the military; therefore, transition to this customer would be seamless. Improved canine training techniques could also be used by commercial dog training organizations, for behavior improvement or therapy applications. Training techniques developed in Phase I and II will substantially reduce canine training time and costs by selecting ideal dogs and optimizing training techniques. These techniques will be transitioned to the US Air Force, which runs the DoD Military Working Dog Program from Lackland AF Base, San Antonio, TX (341st Training Squadron). Advances from this program could also be transitioned to the Veterans Administration, which is running a clinical trial on the impact of therapy dogs on the lives of veterans diagnosed with PTSD. Understanding the mechanism and communication between canine and human will facilitate therapy dog support for veterans with PTSD. REFERENCES: 1) Berns GS, Brooks AM, Spivak M. 2012. Functional MRI in awake unrestrained dogs. PLoS One. 2012;7(5):e38027. Epub 2012 May 11. 2) Kirchhofer KC, Zimmermann F, Kaminski J, Tomasello M. 2012. Dogs (Canis familiaris), but not chimpanzees (Pan troglodytes), understand imperative pointing. PLoS One. 7(2):e30913. 3) Silva K, Bessa J, de Sousa L. 2012. Auditory contagious yawning in domestic dogs (Canis familiaris): first evidence for social modulation. Anim Cogn. Jul;15(4):721-4.

US Flag An Official Website of the United States Government