A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
A pilot study comparing three-dimensional models of tumor histopathology and magnetic resonance imaging
Tekijät: Koivuholma, Anne; Sistonen, Heli J.; Aro, Katri; Mäkitie, Antti; Hagstrom, Jaana; Atula, Timo
Kustantaja: Springer Nature
Kustannuspaikka: BERLIN
Julkaisuvuosi: 2025
Journal: Scientific Reports
Tietokannassa oleva lehden nimi: SCIENTIFIC REPORTS
Lehden akronyymi: SCI REP-UK
Artikkelin numero: 1888
Vuosikerta: 15
Numero: 1
Sivujen määrä: 8
eISSN: 2045-2322
DOI: https://doi.org/10.1038/s41598-025-85478-1
Verkko-osoite: https://www.nature.com/articles/s41598-025-85478-1
Rinnakkaistallenteen osoite: https://research.utu.fi/converis/portal/detail/Publication/484286550
Three-dimensional (3D) modeling is often used to provide better visual understanding. This has become an everyday tool especially in medical imaging. However, modeling soft tissue histopathology in 3D is in its early stages, thus making 3D comparison between radiology and histopathology difficult. Here, we sought to create a model that combines a model based on histopathological tumor borders and a model reformatted from magnetic resonance image (MRI). The 3D models were constructed from an oral squamous cell carcinoma (OSCC) resection specimen and a preoperative MRI scan. Potential challenges were identified, and respective solutions proposed. The 3D model based on histopathological tumor borders was constructed by a method that uses a 3D table scanner and modeling software. The 3D MRI model was constructed using 3D Slicer software. The two models were fitted together by anatomical landmarks in each model. We compared the volume and dimensions between the two 3D models. As a result, we created an image fusion that presents the soft-tissue resection specimen, its histopathological findings, and the MRI findings combined in a digital 3D form. Presenting these two 3D models within the resection specimen can serve as a tool to improve the multidisciplinary discussion of patient management. To create these models, only a few additional steps to a normal protocol are required.
Ladattava julkaisu This is an electronic reprint of the original article. |  |
