RATIONALE AND OBJECTIVES: To determine whether observer performance in the localization of tubes and catheters on computed radiography (CR) chest images can be improved by using an automatic image-processing algorithm. METHODS: Comparisons were made of CR hard-copy, CR soft-copy, and CR enhanced soft-copy images obtained with an image-processing algorithm. The enhanced images used gray-level optimization and nonlinear unsharp masking to emphasize the edges of the devices. Chest radiologists (n = 4), general radiologists (n = 4), and interns (n = 6) read 45 images that contained endotracheal tubes, pulmonary artery catheters, and central venous catheters. RESULTS: Chest radiologists had the smallest mean interobserver localization variability (4 mm), followed by general radiologists (6 mm) and interns (8 mm). Localization variability was greatest for hard-copy images and least for enhanced soft-copy images. CONCLUSION: Use of an automatic imaging-processing algorithm reduced localization variability and enabled the medical interns to perform at approximately the same level as the chest radiologists.
RATIONALE AND OBJECTIVES: To determine whether observer performance in the localization of tubes and catheters on computed radiography (CR) chest images can be improved by using an automatic image-processing algorithm. METHODS: Comparisons were made of CR hard-copy, CR soft-copy, and CR enhanced soft-copy images obtained with an image-processing algorithm. The enhanced images used gray-level optimization and nonlinear unsharp masking to emphasize the edges of the devices. Chest radiologists (n = 4), general radiologists (n = 4), and interns (n = 6) read 45 images that contained endotracheal tubes, pulmonary artery catheters, and central venous catheters. RESULTS: Chest radiologists had the smallest mean interobserver localization variability (4 mm), followed by general radiologists (6 mm) and interns (8 mm). Localization variability was greatest for hard-copy images and least for enhanced soft-copy images. CONCLUSION: Use of an automatic imaging-processing algorithm reduced localization variability and enabled the medical interns to perform at approximately the same level as the chest radiologists.
Authors: Rebecca A Hammon; Hannes Seuss; Matthias Hammon; Christian Grillhösl; Rafael Heiss; Martin Zeilinger; Nadine Bayerl; Pieter Vuylsteke; Friedrich Wanninger; Michael Schroth; Michael Uder; Oliver Rompel Journal: BMC Med Imaging Date: 2019-01-07 Impact factor: 1.930
Authors: Julia Ley-Zaporozhan; Hassan Shoushtari; Ravi Menezes; Leon Zelovitzky; Devang Odedra; Laura Jimenez-Juan; Karyn Brunet; Yasser Karimzad; Narinder S Paul Journal: PLoS One Date: 2018-12-21 Impact factor: 3.240
Authors: Dilip B Nagarkar; Ezgi Mercan; Donald L Weaver; Tad T Brunyé; Patricia A Carney; Mara H Rendi; Andrew H Beck; Paul D Frederick; Linda G Shapiro; Joann G Elmore Journal: Mod Pathol Date: 2016-05-20 Impact factor: 7.842