Automated Analysis of Alignment in Long-Leg Radiographs by Using a Fully Automated Support System Based on Artificial Intelligence
Justus Schock, Daniel Truhn, Daniel B. Abrar, Dorit Merhof, Stefan Conrad, Manuel Post, Felix Mittelstrass, Christiane Kühl, Sven Nebelung
Abstract
PURPOSE: To develop and validate a deep learning-based method for automatic quantitative analysis of lower-extremity alignment. MATERIALS AND METHODS: and intraclass correlation coefficients. RESULTS: < .001). CONCLUSION: © RSNA, 2020See also commentary by Andreisek in this issue.
Topics & Concepts
Intraclass correlationStandard deviationRadiographyInterclass correlationSegmentationTibiaFemurPearson product-moment correlation coefficientArtificial intelligenceCorrelation coefficientNuclear medicineMathematicsMedicineOrthodonticsComputer scienceAnatomyReproducibilityStatisticsRadiologySurgeryTotal Knee Arthroplasty OutcomesHip disorders and treatmentsOrthopedic Infections and Treatments