Impact of Scanner Manufacturer, Endorectal Coil Use, and Clinical Variables on Deep Learning-assisted Prostate Cancer Classification Using Multiparametric MRI.

Journal: Radiology. Artificial intelligence
PMID:

Abstract

Purpose To assess the effect of scanner manufacturer and scanning protocol on the performance of deep learning models to classify aggressiveness of prostate cancer (PCa) at biparametric MRI (bpMRI). Materials and Methods In this retrospective study, 5478 cases from ProstateNet, a PCa bpMRI dataset with examinations from 13 centers, were used to develop five deep learning (DL) models to predict PCa aggressiveness with minimal lesion information and test how using data from different subgroups-scanner manufacturers and endorectal coil (ERC) use (Siemens, Philips, GE with and without ERC, and the full dataset)-affects model performance. Performance was assessed using the area under the receiver operating characteristic curve (AUC). The effect of clinical features (age, prostate-specific antigen level, Prostate Imaging Reporting and Data System score) on model performance was also evaluated. Results DL models were trained on 4328 bpMRI cases, and the best model achieved an AUC of 0.73 when trained and tested using data from all manufacturers. Held-out test set performance was higher when models trained with data from a manufacturer were tested on the same manufacturer (within- and between-manufacturer AUC differences of 0.05 on average, < .001). The addition of clinical features did not improve performance ( = .24). Learning curve analyses showed that performance remained stable as training data increased. Analysis of DL features showed that scanner manufacturer and scanning protocol heavily influenced feature distributions. Conclusion In automated classification of PCa aggressiveness using bpMRI data, scanner manufacturer and ERC use had a major effect on DL model performance and features. Convolutional Neural Network (CNN), Computer-aided Diagnosis (CAD), Computer Applications-General (Informatics), Oncology Published under a CC BY 4.0 license. See also commentary by Suri and Hsu in this issue.

Authors

  • José Guilherme de Almeida
    Champalimaud Foundation, Lisbon, Portugal. jose.almeida@research.fchampalimaud.org.
  • Nuno M Rodrigues
    LASIGE, Department of Informatics, Faculty of Sciences, University of Lisbon, Lisbon, Portugal.
  • Ana Sofia Castro Verde
    Champalimaud Research, Champalimaud Foundation, Avenida Brasilia, Lisboa, Lisboa 1400-038 Portugal.
  • Ana Mascarenhas Gaivão
    Department of Radiology, Champalimaud Clinical Center, Champalimaud Foundation, Lisbon, Portugal.
  • Carlos Bilreiro
    Department of Radiology, Champalimaud Clinical Center, Champalimaud Foundation, Lisbon, Portugal.
  • Inês Santiago
    Radiology Department, Champalimaud Clinical Center, Champalimaud Foundation, Lisbon, Portugal.
  • Joana Ip
    Radiology Department, Champalimaud Clinical Center, Champalimaud Foundation, Lisbon, Portugal.
  • Sara Belião
    Radiology Department, Champalimaud Clinical Center, Champalimaud Foundation, Lisbon, Portugal.
  • Celso Matos
    Champalimaud Foundation, Lisbon, Portugal.
  • Sara Silva
    LASIGE, Faculdade de Ciências, Universidade de Lisboa, Lisboa, Portugal.
  • Manolis Tsiknakis
    Computational BioMedicine Laboratory, FORTH-ICS, Heraklion, Crete, Greece.
  • Kostantinos Marias
    Department of Electrical and Computer Engineering, Hellenic Mediterranean University, Heraklion, Greece.
  • Daniele Regge
    From the 3D Imaging Research Lab, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, 25 New Chardon St, Suite 400C, Boston, MA 02114 (R.T., J.J.N., N.K., T.H., H.Y.); Department of Information Science and Technology, National Institute of Technology, Oshima College, Yamaguchi, Japan (R.T.); Department of Medical Physics and Engineering, Graduate School of Medicine, Osaka University, Suita, Osaka, Japan (J.O.); Department of Medical Physics, University of Applied Sciences Giessen, Giessen, Germany (N.K.); Department of Radiology, Seoul National University Hospital, Seoul, Republic of Korea (S.H.K.); Department of Surgical Sciences, University of Torino, Turin, Italy (D.R.); and Candiolo Cancer Institute, Fondazione del Piemonte per l'Oncologia-Istituto di Ricovero e Cura a Carattere Scientifico (FPO-IRCCS), Candiolo, Turin, Italy (D.R.).
  • Nikolaos Papanikolaou
    Computational Biomedicine Laboratory (CBML), Foundation for Research and Technology Hellas (FORTH), 70013 Heraklion, Greece.