MRI Texture Analysis for Differentiation Between Healthy and Golden Retriever Muscular Dystrophy Dogs at Different Phases of Disease Evolution
Abstract
In this study, a texture analysis is applied to T2-weighted Magnetic Resonance Images (MRI) of canine pelvic limbs in order to differentiate between Golden Retriever Muscular Dystrophy (GRMD) dogs and healthy ones. The differentiation is performed at three phases of canine growth and/or disease development: 2-4 months (the first phase), 5-6 months (the second phase), and 7 months and more (the third phase). Eight feature extraction methods (statistical, model-based, and filter-based) and five classifiers are tested. Four types of muscles are analyzed: the Extensor Digitorum Longus (EDL), the Gastrocnemius Lateralis (GasLat), the Gastrocnemius Medialis (GasMed) and the Tibial Cranialis (TC). The experiments were performed on five healthy and five GRMDdogs. Each of themuscles was considered separately. The best classification results were 95.81% (the EDL muscle), 97.19% (GasLat), and 91.37% (EDL) correctly recognized cases, for the first, second and third phase, respectively. These results were obtained with an SVM classifier.
Origin | Files produced by the author(s) |
---|