EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Recommender-based bone tumour classification with radiographs—a link to the past

Florian Hinterwimmer, Ricardo Smits Serena, Nikolas Wilhelm, Sebastian Breden, Sarah Consalvo, Fritz Seidl, Dominik Juestel, Rainer H H Burgkart, Klaus Woertler, Ruediger von Eisenhart-Rothe, Jan Neumann, Daniel Rueckert

European Radiology · 2024

DOI: 10.1007/s00330-024-10672-0 ↗

Kurz gefasst

Fachartikel in der Zeitschrift European Radiology, 2024. Verfasst von Florian Hinterwimmer und 11 weiteren Autor:innen. Beteiligte Einrichtung: Department of Orthopaedics and Sports Orthopaedics, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany. florian.hinterwimmer@t…. Hauptthemen laut MEDLINE: Bone Neoplasms, Algorithms. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Technische Universität München.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objectives

To develop an algorithm to link undiagnosed patients to previous patient histories based on radiographs, and simultaneous classification of multiple bone tumours to enable early and specific diagnosis. Materials and methods For this retrospective study, data from 2000 to 2021 were curated from our database by two orthopaedic surgeons, a radiologist and a data scientist. Patients with complete clinical and pre-therapy radiographic data were eligible. To ensure feasibility, the ten most frequent primary tumour entities, confirmed histologically or by tumour board decision, were included. We implemented a ResNet and transformer model to establish baseline results. Our method extracts image features using deep learning and then clusters the k most similar images to the target image using a hash-based nearest-neighbour recommender approach that performs simultaneous classification by majority voting. The results were evaluated with precision-at- k , accuracy, precision and recall. Discrete parameters were described by incidence and percentage ratios. For continuous parameters, based on a normality test, respective statistical measures were calculated.

Results

Included were data from 809 patients (1792 radiographs; mean age 33.73 ± 18.65, range 3–89 years; 443 men), with Osteochondroma (28.31%) and Ewing sarcoma (1.11%) as the most and least common entities, respectively. The dataset was split into training (80%) and test subsets (20%). For k = 3, our model achieved the highest mean accuracy, precision and recall (92.86%, 92.86% and 34.08%), significantly outperforming state-of-the-art models (54.10%, 55.57%, 19.85% and 62.80%, 61.33%, 23.05%).

Conclusion

Our novel approach surpasses current models in tumour classification and links to past patient data, leveraging expert insights. Clinical relevance statement The proposed algorithm could serve as a vital support tool for clinicians and general practitioners with limited experience in bone tumour classification by identifying similar cases and classifying bone tumour entities. Key Points • Addressed accurate bone tumour classification using radiographic features. • Model achieved 92.86%, 92.86% and 34.08% mean accuracy, precision and recall, respectively, significantly surpassing state-of-the-art models. • Enhanced diagnosis by integrating prior expert patient assessments. Graphical abstract

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Bone NeoplasmsAlgorithmsHumansMaleFemaleAdultRetrospective StudiesAgedMiddle AgedAdolescentAged, 80 and overChildChild, PreschoolYoung AdultRadiographyDeep Learning

Stichwörter der Autor:innen

Bone neoplasmsClassificationDeep learningMachine learningRadiography

Autor:innen und Institutionen

  • Department of Orthopaedics and Sports Orthopaedics, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany. florian.hinterwimmer@tum.de.
  • Institute for AI and Informatics in Medicine, Technical University of Munich, Munich, Germany. florian.hinterwimmer@tum.de.
  • Department of Orthopaedics and Sports Orthopaedics, Klinikum rechts der Isar, Technical University of Munich, Munich, Germany.
  • Institute for AI and Informatics in Medicine, Technical University of Munich, Munich, Germany.
  • Institute of Biological and Medical Imaging, Helmholtz Zentrum München, Neuherberg, Germany.
  • Institute at Helmholtz: Institute of Computational Biology, Oberschleißheim, Germany.

Förderung

  • Technische Universität München

Zitieren

APA

Hinterwimmer, F., Serena, R. S., Wilhelm, N., Breden, S., Consalvo, S., Seidl, F., Juestel, D., Burgkart, R. H. H., Woertler, K., von Eisenhart-Rothe, R., Neumann, J., & Rueckert, D. (2024). Recommender-based bone tumour classification with radiographs—a link to the past. European Radiology. https://doi.org/10.1007/s00330-024-10672-0

Vancouver

Hinterwimmer F, Serena RS, Wilhelm N, Breden S, Consalvo S, Seidl F, et al. Recommender-based bone tumour classification with radiographs—a link to the past. European Radiology. 2024. doi:10.1007/s00330-024-10672-0

BibTeX
@article{Hinterwimmer2024,
  author = {Hinterwimmer, Florian and Serena, Ricardo Smits and Wilhelm, Nikolas and Breden, Sebastian and Consalvo, Sarah and Seidl, Fritz and Juestel, Dominik and Burgkart, Rainer H H and Woertler, Klaus and von Eisenhart-Rothe, Ruediger and Neumann, Jan and Rueckert, Daniel},
  title = {Recommender-based bone tumour classification with radiographs—a link to the past},
  journal = {European Radiology},
  year = {2024},
  doi = {10.1007/s00330-024-10672-0}
}
Kostenlos starten →