Juan S Calle-Toro1, Summer L Kaplan2,3, Savvas Andronikou2,3. 1. Department of Radiology, Children's Hospital of Philadelphia, 3401 Civic Center Blvd., Philadelphia, USA. juan53670@gmail.com. 2. Department of Radiology, Children's Hospital of Philadelphia, 3401 Civic Center Blvd., Philadelphia, USA. 3. Perelman School of Medicine, University of Pennsylvania, 3400 Civic Center Blvd., Philadelphia, PA, 19104, USA.
Abstract
PURPOSE: To retrospectively review imaging planes, number of visible pyloric layers and location of measurements, in infants with suspected (HPS). METHODS: 103 pyloric ultrasound studies for suspected HPS were included. For each study, we recorded whether longitudinal or transverse views were performed, the layers visualized (a schematic was developed for two pediatric radiologists to categorize the interfaces of the anatomic layers a-e) and position of the internal measurement cursor. Categories for the anterior (superficial wall) layers were from external to internal: (a) internal aspect of the muscularis propria; (b) external aspect of the muscularis mucosa; (c) internal aspect of the muscularis mucosa; (d) internal aspect of the mucosa interfacing with a mucosal fold (e) deep aspect of the mucosal fold. Median differences between HPS groups were calculated and inter-reader agreement (kappa score) was performed between both readers. RESULTS: In 100 studies (97 patients), longitudinal (99%) and transverse (69%) views of the pylorus were recorded. For longitudinal views, measurements included muscle thickness (95%), length (97%) and no pyloric diameter. For the transverse view, measurements included muscle thickness (16%) and the diameter (3%). Pyloric layer interfaces were visible: (a) in 64% (b) in 64% (c) in 66% (d) in 30% and (e) in 26%. The internal reference point of cursor placement for measuring the muscle wall thickness in the longitudinal view for one reader was as follows: (a) 46% (b) 27% (c) 30% (d) 1% and (e) 2% of studies. Surgically proven HPS group had a median thickness measurement 0.17 mm greater than the non-HPS studies (CI 95% 0.12-0.21, p < 0.05), and inter-reader agreement was considered as moderate (Kappa 0.5). CONCLUSIONS: We found a variety of thickness measurements performed predominantly in the longitudinal view and a largely abandoned diameter measurement. The latter might offer a solution as it is not defined by any internal interfaces.
PURPOSE: To retrospectively review imaging planes, number of visible pyloric layers and location of measurements, in infants with suspected (HPS). METHODS: 103 pyloric ultrasound studies for suspected HPS were included. For each study, we recorded whether longitudinal or transverse views were performed, the layers visualized (a schematic was developed for two pediatric radiologists to categorize the interfaces of the anatomic layers a-e) and position of the internal measurement cursor. Categories for the anterior (superficial wall) layers were from external to internal: (a) internal aspect of the muscularis propria; (b) external aspect of the muscularis mucosa; (c) internal aspect of the muscularis mucosa; (d) internal aspect of the mucosa interfacing with a mucosal fold (e) deep aspect of the mucosal fold. Median differences between HPS groups were calculated and inter-reader agreement (kappa score) was performed between both readers. RESULTS: In 100 studies (97 patients), longitudinal (99%) and transverse (69%) views of the pylorus were recorded. For longitudinal views, measurements included muscle thickness (95%), length (97%) and no pyloric diameter. For the transverse view, measurements included muscle thickness (16%) and the diameter (3%). Pyloric layer interfaces were visible: (a) in 64% (b) in 64% (c) in 66% (d) in 30% and (e) in 26%. The internal reference point of cursor placement for measuring the muscle wall thickness in the longitudinal view for one reader was as follows: (a) 46% (b) 27% (c) 30% (d) 1% and (e) 2% of studies. Surgically proven HPS group had a median thickness measurement 0.17 mm greater than the non-HPS studies (CI 95% 0.12-0.21, p < 0.05), and inter-reader agreement was considered as moderate (Kappa 0.5). CONCLUSIONS: We found a variety of thickness measurements performed predominantly in the longitudinal view and a largely abandoned diameter measurement. The latter might offer a solution as it is not defined by any internal interfaces.