Explainable Precision Medicine in Breast MRI: A Combined Radiomics and Deep Learning Approach for the Classification of Contrast Agent Uptake.

Autor: Nowakowska S; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland., Borkowski K; b-rayZ AG, Wagistrasse 21, 8952 Schlieren, Switzerland., Ruppert C; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland.; b-rayZ AG, Wagistrasse 21, 8952 Schlieren, Switzerland., Hejduk P; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland., Ciritsis A; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland.; b-rayZ AG, Wagistrasse 21, 8952 Schlieren, Switzerland., Landsmann A; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland., Marcon M; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland., Berger N; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland., Boss A; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland., Rossi C; Diagnostic and Interventional Radiology, University Hospital Zürich, University Zürich, Rämistrasse 100, 8091 Zürich, Switzerland.; b-rayZ AG, Wagistrasse 21, 8952 Schlieren, Switzerland.
Jazyk: angličtina
Zdroj: Bioengineering (Basel, Switzerland) [Bioengineering (Basel)] 2024 May 31; Vol. 11 (6). Date of Electronic Publication: 2024 May 31.
DOI: 10.3390/bioengineering11060556
Abstrakt: In DCE-MRI, the degree of contrast uptake in normal fibroglandular tissue, i.e., background parenchymal enhancement (BPE), is a crucial biomarker linked to breast cancer risk and treatment outcome. In accordance with the Breast Imaging Reporting & Data System (BI-RADS), it should be visually classified into four classes. The susceptibility of such an assessment to inter-reader variability highlights the urgent need for a standardized classification algorithm. In this retrospective study, the first post-contrast subtraction images for 27 healthy female subjects were included. The BPE was classified slice-wise by two expert radiologists. The extraction of radiomic features from segmented BPE was followed by dataset splitting and dimensionality reduction. The latent representations were then utilized as inputs to a deep neural network classifying BPE into BI-RADS classes. The network's predictions were elucidated at the radiomic feature level with Shapley values. The deep neural network achieved a BPE classification accuracy of 84 ± 2% ( p -value < 0.00001). Most of the misclassifications involved adjacent classes. Different radiomic features were decisive for the prediction of each BPE class underlying the complexity of the decision boundaries. A highly precise and explainable pipeline for BPE classification was achieved without user- or algorithm-dependent radiomic feature selection.
Databáze: MEDLINE
Nepřihlášeným uživatelům se plný text nezobrazuje