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.

PILOT STUDY TO ASSESS MEAL PROGRESSION THROUGH THE GASTROINTESTINAL TRACT OF HABITUATED DOGS DETERMINED BY FLUOROSCOPIC IMAGING WITHOUT SEDATION OR PHYSICAL RESTRAINT.

A limiting factor of radiographic contrast studies is the requirement for restraint of the animal in order to reduce movement artifacts. To demonstrate that gastrointestinal transit can be analyzed by a barium meal in nonsedated and unrestrained dogs, a pilot study of six adult Labrador retriever dogs was undertaken. Study subjects were selected by convenience sampling from an available population of Labrador dogs and were trained to stand motionless during radiographic fluoroscopy. Following a meal containing 7% w/w powdered barium sulfate, radiographic images were generated using a digital fluoroscope C-arm, at intervals of 5, 15, and 30 min, and at 1, 2, 3, 4, 5, 6, 7, and 8 h. A qualitative assessment of fill density using a 5-point scale was made for the stomach, small intestine, and ascending, transverse, and descending regions of the colon at each timepoint. Gastric emptying half-time occurred between 1 and 2 h postmeal. Mean fill density of the small intestine increased from 15 min postmeal and reached a peak at 3 h postmeal. Mean fill density of the proximal large intestine mirrored that of the small intestine. The distal large intestine remained empty for the first 2 h postmeal, then increased between hours 2 and 5 postmeal, and was subsequently at maximum fill density from hour 6 postmeal onwards. Fluoroscopic observation of a barium contrast meal provided an effective indication of the amount and progression of ingested food through the various regions of the gastrointestinal tract in habituated, fully conscious dogs.

Intra-abdominal fat burden discriminated in vivo using proton magnetic resonance spectroscopy.

OBJECTIVE: To assess proton magnetic resonance spectroscopy (1H-MRS) as a means to distinguish among mice with disparate intra-abdominal body fat compositions, and to measure changes in intra-abdominal fat burden during weight loss and regain. RESEARCH METHODS AND PROCEDURES: Intra-abdominal fat burden was analyzed as a ratio of integrated areas under the curves of fat to water (1)H-MRS signals collected from a region of interest standardized across B6.V-Lep(ob), C57BL/6, and A-ZIP/F mice that exhibited various genotypically related body fat compositions, ranging from obese (B6.V-Lep(ob)) to minimal body fat (A-ZIP/F). 1H-MRS analysis of fat burden was compared with intra-abdominal fat volume and with a single cross-sectional intra-abdominal fat area calculated from segmented magnetic resonance images. Similar measurements were made from obese B6.V-Lep(ob) mice before, during, and after they were induced to lose weight by leptin administration. RESULTS: Relative amounts of intra-abdominal fat analyzed by 1H-MRS differed significantly according to body composition and genotype of the three strains of mice (p < 0.05). Intra-abdominal fat assessed by 1H-MRS correlated with both intra-abdominal fat volume (r = 0.88, p < 0.001) and body weight (r = 0.82, p < 0.001) among, but not within, all three genotypes. During weight loss and regain, there was a significant overall pattern of changes in intra-abdominal fat quantity that occurred, which was reflected by 1H-MRS (p = 0.006). DISCUSSION: Results support the use of localized 1H-MRS for assessing differences in intra-abdominal fat. Refinements in 1H-MRS voxel region of interest size and location as well as instrument precision may result in improved correlations within certain body compositions.