The role of artificial intelligence in clinical imaging and workflows.

Evidence-based medicine, outcomes management, and multidisciplinary systems are laying the foundation for radiology on the cusp of a new day. Environmental and operational forces coupled with technological advancements are redefining the veterinary radiologist of tomorrow. In the past several years, veterinary image volumes have exploded, and the scale of hardware and software required to support it seems boundless. The most dynamic trend within veterinary radiology is implementing digital information systems such as PACS, RIS, PIMS, and Voice Recognition systems. While the digitization of radiography imaging has significantly improved the workflow of the veterinary radiology assistant and radiologist, tedious, redundant tasks are abundant and mind-numbing. They can lead to errors with a significant impact on patient care.  Today, these boring and repetitious tasks continue to bog down patient throughput and workflow. Artificial intelligence, particularly machine learning, shows much promise to rocket the workflow and veterinary clinical imaging into a new day where the AI management of mundane tasks allows for efficiency so the radiologist can better concentrate on the quality of patient care. In this article, we briefly discuss the major subsets of artificial intelligence (AI) workflow for the radiologist and veterinary radiology assistant including image acquisition, segmentation and mensuration, rotation and hanging protocol, detection and prioritization, monitoring and registration of lesions, implementation of these subsets, and the ethics of utilizing AI in veterinary medicine.

Mucopolysaccharidosis-like phenotype in feline Sandhoff disease and partial correction after AAV gene therapy.

Sandhoff disease (SD) is a fatal neurodegenerative disease caused by a mutation in the enzyme ß-N-acetylhexosaminidase. Children with infantile onset SD develop seizures, loss of motor tone and swallowing problems, eventually reaching a vegetative state with death typically by 4years of age. Other symptoms include vertebral gibbus and cardiac abnormalities strikingly similar to those of the mucopolysaccharidoses. Isolated fibroblasts from SD patients have impaired catabolism of glycosaminoglycans (GAGs). To evaluate mucopolysaccharidosis-like features of the feline SD model, we utilized radiography, MRI, echocardiography, histopathology and GAG quantification of both central nervous system and peripheral tissues/fluids. The feline SD model exhibits cardiac valvular and structural abnormalities, skeletal changes and spinal cord compression that are consistent with accumulation of GAGs, but are much less prominent than the severe neurologic disease that defines the humane endpoint (4.5±0.5months). Sixteen weeks after intracranial AAV gene therapy, GAG storage was cleared in the SD cat cerebral cortex and liver, but not in the heart, lung, skeletal muscle, kidney, spleen, pancreas, small intestine, skin, or urine. GAG storage worsens with time and therefore may become a significant source of pathology in humans whose lives are substantially lengthened by gene therapy or other novel treatments for the primary, neurologic disease.

AAV-mediated gene delivery in a feline model of Sandhoff disease corrects lysosomal storage in the central nervous system.

Sandhoff disease (SD) is an autosomal recessive neurodegenerative disease caused by a mutation in the gene for the ß-subunit of ß-N-acetylhexosaminidase (Hex), resulting in the inability to catabolize ganglioside GM2 within the lysosomes. SD presents with an accumulation of GM2 and its asialo derivative GA2, primarily in the central nervous system. Myelin-enriched glycolipids, cerebrosides and sulfatides, are also decreased in SD corresponding with dysmyelination. At present, no treatment exists for SD. Previous studies have shown the therapeutic benefit of adeno-associated virus (AAV) vector-mediated gene therapy in the treatment of SD in murine and feline models. In this study, we treated presymptomatic SD cats with AAVrh8 vectors expressing feline Hex in the thalamus combined with intracerebroventricular (Thal/ICV) injections. Treated animals showed clearly improved neurologic function and quality of life, manifested in part by prevention or attenuation of whole-body tremors characteristic of untreated animals. Hex activity was significantly elevated, whereas storage of GM2 and GA2 was significantly decreased in tissue samples taken from the cortex, cerebellum, thalamus, and cervical spinal cord. Treatment also increased levels of myelin-enriched cerebrosides and sulfatides in the cortex and thalamus. This study demonstrates the therapeutic potential of AAV for feline SD and suggests a similar potential for human SD patients.

Biomarkers for disease progression and AAV therapeutic efficacy in feline Sandhoff disease.

The GM2 gangliosidoses, Tay-Sachs disease (TSD) and Sandhoff disease (SD), are progressive neurodegenerative disorders that are caused by a mutation in the enzyme ß-N-acetylhexosaminidase (Hex). Due to the recent emergence of novel experimental treatments, biomarker development has become particularly relevant in GM2 gangliosidosis as an objective means to measure therapeutic efficacy. Here we describe blood, cerebrospinal fluid (CSF), magnetic resonance imaging (MRI), and electrodiagnostic methods for evaluating disease progression in the feline SD model and application of these approaches to assess AAV-mediated gene therapy. SD cats were treated by intracranial injections of the thalami combined with either the deep cerebellar nuclei or a single lateral ventricle using AAVrh8 vectors encoding feline Hex. Significantly altered in untreated SD cats, blood and CSF based biomarkers were largely normalized after AAV gene therapy. Also reduced after treatment were expansion of the lysosomal compartment in peripheral blood mononuclear cells and elevated activity of secondary lysosomal enzymes. MRI changes characteristic of the gangliosidoses were documented in SD cats and normalized after AAV gene therapy. The minimally invasive biomarkers reported herein should be useful to assess disease progression of untreated SD patients and those in future clinical trials.

High resolution MRI anatomy of the cat brain at 3 Tesla.

BACKGROUND: Feline models of neurologic diseases, such as lysosomal storage diseases, leukodystrophies, Parkinson's disease, stroke and NeuroAIDS, accurately recreate many aspects of human disease allowing for comparative study of neuropathology and the testing of novel therapeutics. Here we describe in vivo visualization of fine structures within the feline brain that were previously only visible post mortem. NEW METHOD: 3Tesla MR images were acquired using T1-weighted (T1w) 3D magnetization-prepared rapid gradient echo (MPRAGE) sequence (0.4mm isotropic resolution) and T2-weighted (T2w) turbo spin echo (TSE) images (0.3mm×0.3mm×1mm resolution). Anatomic structures were identified based on feline and canine histology. RESULTS: T2w high resolution MR images with detailed structural identification are provided in transverse, sagittal and dorsal planes. T1w MR images are provided electronically in three dimensions for unrestricted spatial evaluation. COMPARISON WITH EXISTING METHODS: Many areas of the feline brain previously unresolvable on MRI are clearly visible in three orientations, including the dentate, interpositus and fastigial cerebellar nuclei, cranial nerves, lateral geniculate nucleus, optic radiation, cochlea, caudal colliculus, temporal lobe, precuneus, spinocerebellar tract, vestibular nuclei, reticular formation, pyramids and rostral and middle cerebral arteries. Additionally, the feline brain is represented in three dimensions for the first time. CONCLUSIONS: These data establish normal appearance of detailed anatomical structures of the feline brain, which provide reference when evaluating neurologic disease or testing efficacy of novel therapeutics in animal models.

Primary ovarian teratoma and GCT with intra-abdominal metastasis in a dog.

This report describes the simultaneous occurrence of an ovarian teratoma and a granulosa cell tumor (GCT) with intra-abdominal metastasis in a 1.5 yr old female Doberman pinscher. At surgery, a 20 cm, smooth, intact mass associated with the left ovary and multiple 1-2 cm irregular masses in the broad ligament were found. The masses were surgically removed and submitted for histopathology. A histologic diagnosis of a teratoma and a GCT with broad ligament metastasis was made. Further treatment was elected by the owner and included two cycles of carboplatin therapy. The dog was euthanized 6 wk postoperatively for signs related to metastasis and dyspnea. Teratoma of the ovary, although it contains derivatives of all three embryonic germ cell layers, rarely presents together with either ovarian epithelial or sex cord-stromal tumors. To the authors' knowledge, this is the first reported case of an ovarian teratoma coexisting with a primary GCT with intra-abdominal metastasis in the same ovary in a dog.