ECEESPE2025 Poster Presentations Bone and Mineral Metabolism (112 abstracts)
Deeplasia: a novel tool for the assessment of bone age in skeletal dysplasia
Alexandra Keller 1 , Behnam Javanmardi 2 , Maesa Al-Hallak 3 , Daniela Avdjieva 4 , Susanne Bechtold-Dalla Pozza 5 , Christoph Beger 6 , Desiree Dunstheimer 7 , Corinna Grasemann 8 , Julie Harvengt 9 , Wolfgang Högler 10 , Angela Hübner 11 , Nicole Muschol 12 , Sean Nader 13 , Katja Palm 14 , Roland Pfäffle 6 , Beata Pyrzak 15 , Tilman Rohrer 16 , Heike Hoyer-Kuhn 17 , Ondřej Souček 18 , Ioana Streata 19 , Thomas Völkl 20 , Katja Wechsung 21 , Joachim Woelfle 22 , Jiří Zeman 23 & Klaus Mohnike 14
1Kinderzentrum Am Johannisplatz, Leipzig, Leipzig, Germany; 2Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Bonn, Germany; 3kbo-Kinderzentrum, Munich, München, Germany; 4SBAL Children’s Hospital Sofia, Sofia, Bulgaria; 5University Hospital Munich, Department. Pediatrics, München, Germany; 6University Hospital Leipzig, Department Pediatrics, Leipzig, Germany; 7Pediatrics, University Hospital Augsburg, Department Pediatrics, Augsburg, Germany; 8Department of Pediatrics, Katholisches Klinikum Bochum, Ruhr University-Bochum, Bochum, Bochum, Germany; 9CHU Génétique Liege, Liege, Belgium; 10Department of Paediatrics and Adolescent Medicine, Johannes Kepler University Linz, Linz, Austria; 11Department of Paediatrics, Faculty of Medicine and University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany; 12Pediatrics, University Hospital Hamburg-Eppendorf, Hamburg, Germany; 13Schön-Clinic Vogtareuth, Pediatric Orthopedics, Vogtareuth, Germany; 14Otto-von-Guericke University, Med. Faculty, Department Pediatrics, Magdeburg, Germany; 15Medical University of Warsaw, Department Pediatrics, Warsaw, Poland; 16University Hospital Saarland, Department Pediatrics, Homburg, Germany; 17University Hospital Cologne, Department Pediatrics, Cologne, Germany; 18Motol Prague, Prague, Czech Republic; 19University of Medicine and Pharmacy of Craiova, Regional Centre of Medical Genetics Dolj, Emergency Clinical County Hospital Craiova, Laboratory of Human Genomics, Dolj, Romania; 20Josefinum Augsburg, Department of Pediatrics and Adolescent Medicine, Augsburg, Germany; 21Charité Universitätsmedizin Berlin, Pediatric Endocrinology, Berlin, Germany; 22University Hospital Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Department of Pediatrics and Adolescent Medicine, Erlangen, Germany; 23VFN Uni Prague, Department Pediatrics, Prague, Czech Republic
JOINT2900
Bone age assessment as a marker of skeletal maturation is a relevant biomarker in treatment decisions. Visual evaluation by the atlas methods according to Greulich-Pyle (GP) or Tanner-Whitehouse are standard procedures. However, these comparators are based on images from healthy children. Another disadvantage is the subjective nature of the procedure and in the evaluation of abnormal bone structure, e. g. skeletal dysplasia. BoneXpert was the first artificial intelligence (AI) radiology system to be marketed, but as outlined in the manual this tool excludes data analysis from subjects with bone dysplasia. To address this gap, we developed Deeplasia; an open-source prior-free deep-learning approach designed for bone age assessment, specifically validated on skeletal dysplasias incl. achondroplasia (Rassmann et al., 2023). To expand its validation scope, Deeplasia was integrated into the CrescNet auxological database (ClinicalTrials. gov ID: NCT03072537). All data were documented directly during the visit at participating endocrine centers. 953 images from different diagnoses were examined by both Deeplasia and by paediatric endocrinologists (Table) and analyzed using Bland-Altman plot. In particular, we included images from subjects with achondroplasia (305 images of 126 patients), hypochondroplasia (62/21), Noonan syndrome (83/16), SHOX deficiency (226/42), Silver-Russell syndrome (SRS, 24/5), Ullrich-Turner syndrome (UTS, 223/57), and others (38/23). Mean visual bone age was 8. 3 years (SD 4. 1) according to Greulich-Pyle and 8. 4 years (4. 2) according to Deeplasia. Analysis of the Bland-Altman plot between the two methods shows a high level of agreement across all ages, with few outliers. In conclusion, our data support that Deeplasia might significantly aid in assessing BA in skeletal dysplasia.
| Achondroplasia | Hypochondroplasia | Noonan | SHOX | SRS | UTS | |
| N | 305 | 62 | 83 | 226 | 24 | 223 |
| Mean difference Deeplasia vs. GP (years) | -0. 25 | 0. 26 | 0. 26 | 0. 47 | 0. 72 | 0. 03 |
| SD (years) | 1. 03 | 1. 66 | 0. 72 | 0. 88 | 0. 96 | 0. 85 |
Volume 110
Article tools
My recent searches
My recently viewed abstracts
Authors
Endocrine Abstracts
ISSN 1470-3947 (print) | ISSN 1479-6848 (online)
© Bioscientifica 2025 |
Privacy policy |
Cookie settings
BiosciAbstracts
Bioscientifica Abstracts is the gateway to a series of products that provide a permanent, citable record of abstracts for biomedical and life science conferences.
Find out more