ACMIT is pleased to announce a new publication in the Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine (Springer Nature).
Publication details
- Title: Explainable transfer learning ensemble AI model for lung ultrasound pneumothorax detection with expert benchmark
- Authors: Gábor Orosz, Róbert Zsolt Szabó, Marcell Szabó, Pál Gyombolai, József T. Tóth, Tamás Ruttkay, Tamás Ferenci, Tamás Ungi, Gábor Fichtinger & Tamás Haidegger
- Date of publication: 17 April 2026
The paper presents an explainable artificial intelligence (AI) ensemble model for detecting pneumothorax in lung ultrasound images, which was trained on a large, diverse dataset and achieved expert-level—sometimes superior—diagnostic accuracy, while providing interpretable visual outputs to support clinicians and reduce variability in emergency care decision-making.
We are especially proud that the publication includes contributions from Prof. Dr. Gábor Fichtinger, member of the ACMIT Scientific Board and Key Researcher, and Prof. Dr. Tamás Haidegger, ACMIT Research Area Manager. Both Prof. Haidegger and Prof. Fichtinger are long-term partners and supporters of ACMIT, whose continuous collaboration has significantly contributed to strengthening its international research excellence.
For more information, please visit:
