Convolutional neural networks (CNNs) are commonly used as artificial intelligence (AI) tools for evaluating radiographs, but published studies testing their performance in veterinary patients are currently lacking. The purpose of this retrospective, secondary analysis, diagnostic accuracy study was to compare the error rates of four CNNs to the error rates of 13 veterinary radiologists for evaluating canine thoracic radiographs using an independent gold standard. Radiographs acquired at a referral institution were used to evaluate the four CNNs sharing a common architecture. Fifty

Objective To compare pin placement accuracy, intraoperative technique deviations, and duration of pin placement for pins placed by free‐hand probing (FHP) or 3D‐printed drill guide (3DPG) technique. Sample population Four greyhound cadavers. Methods Computed tomography (CT) examinations from T6‐sacrum were obtained for determination of optimal pin placement and 3DPG creation. Two 3.2/2.4‐mm positive profile pins were inserted per vertebra, one left and one right from T7–L7 (FHP [n = 56]; 3DPG [n = 56]) by one surgeon and removed for repeat CT. Duration of pin placement and intraoperative deviations (unanticipated deviations from planned technique) were recorded. Pin tracts were graded by two blinded observers using modified Zdichavsky classification. Descriptive statistics were used. Results A total of 54/56 pins placed with 3DPGs were assigned grade I (optimal placement) compared with 49/56 pins using the FHP technique. A total of 2/56 pins placed with 3DPGs and 3/56 pins using the FHP technique were assigned grade IIa (partial medial violation). A total of 4/56 pins placed using the FHP technique were assigned grade IIIa (partial lateral violation). No pins were assigned grade IIb (full medial violation). Intraoperative technique deviations occurred with 6/56 pins placed using the FHP technique and no pins with 3DPGs. Overall, pins were placed faster (mean ± SD 2.6 [1.3] vs. 4.5 [1.8] min) with 3DPGs. Conclusions Both techniques were accurate for placement of spinal fixation pins. The 3DPG technique may decrease intraoperative deviations and duration of pin placement. Clinical relevance Both techniques allow accurate pin placement in the canine thoracolumbar spine. The FHP technique requires specific training and has learning curve, whereas 3DPG technique requires specific software and 3D printers.

ObjectiveTo describe arthroscopic‐assisted hip toggle stabilization (AA‐HTS) in cats, evaluate its feasibility and associated rate of iatrogenic injury, and assess deviations from planned surgical technique.Study designEx vivo study.AnimalsSkeletally mature cat cadavers (n = 7).MethodsPreoperative pelvic computed tomography (CT) was performed for surgical planning and to identify the ideal femoral bone tunnel projection. Ultrasound‐guided transection of ligament of head of femur was performed. Following exploratory arthroscopy, AA‐HTS was performed using a commercially available aiming device. Surgical time, intraoperative complications, and feasibility of technique were recorded. Iatrogenic injury and technique deviations were assessed by postoperative CT and gross dissection.ResultsDiagnostic arthroscopy and AA‐HTS were successfully performed in all 14 joints. Median (range) surgical time was 46.5 (29‐144) min, including 7 (3‐12) min for diagnostic arthroscopy and 40 (26‐134) min for AA‐HTS. Intraoperative complications occurred in 5 hips, related to bone tunnel creation (4) and toggle dislodgment (1). Toggle passage through the femoral tunnel was the most challenging component of technique, recorded as mildly difficult in 6 joints. No damage to periarticular/intrapelvic structures was identified. Minor articular cartilage damage (<10% total cartilage area) was identified in 10 joints. Thirteen deviations (8 major, 5 minor) in surgical technique from preoperative planning were identified in 7 joints.ConclusionIn feline cadavers AA‐HTS was feasible but was associated with a high rate of minor cartilage injury, intraoperative complications, and technique deviations.Clinical significanceHip toggle stabilization using an arthroscopic‐assisted approach may be an effective technique for management of coxofemoral luxation in cats.