Selective breeding of dogs has reached a critical juncture where welfare and quality of life are

compromised due to a limited understanding of associated diseases and disorders. This thesis

demonstrates how artificial intelligence (AI) can be harnessed to investigate the morphological

changes linked to developmental disorders in Cavalier King Charles Spaniels, specifically focus-

ing on chiari-like malformation (CM) and syringomyelia (SM).

The first contribution presents a novel protocol for extracting morphological data from MR

imaging and integrating it into a fully data-driven AI model. The results are mapped back onto

the MR images, enabling visualisation of the specific deformations and affected regions of the

head. The machine drew findings related to CM and SM morphologies with sensitivities of 89%

and 84% respectively with specificities exceeding 76% in both cases.

Building on this, the second study analyses topographical details from 3D surfaces derived from

CT imaging. Advanced techniques from mathematics and physics are employed to transform

these structures into an AI-compatible dataset, revealing morphological changes associated with

CM and SM and attaining accuracies exceeding 75%. This approach also allows researchers to

re-interpret these findings, simplifying the identification of key biomarkers in CM/SM-affected

dogs.

Aside from the CM/SM-related learnings, the key achievements subtending from this thesis

involve developing a successful AI model despite a shortfall in data available for such experi-

ments, as well as providing a robust protocol to enable these methods to be more widely applied.

The strong confidence in each AI model’s results, in both experiments, supports the assertion

that AI has a role to play in aiding the healthcare community in understanding the pathogenesis

and pathophysiology of complex diseases and disorders. Furthermore, the multi-disciplinary

approach encompassing mathematics, engineering, artificial intelligence and veterinary science

proves to be an insightful combination and highlights promising avenues for future research.