Voxel level dense prediction of acute stroke territory in DWI using deep learning segmentation models and image enhancement strategies

Koska I. O., SELVER M. A., Gelal F., Uluc M. E., Çetinoğlu Y. K., Yurttutan N., ...More

Japanese Journal of Radiology, 2024 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Publication Date: 2024
  • Doi Number: 10.1007/s11604-024-01582-8
  • Journal Name: Japanese Journal of Radiology
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Biotechnology Research Abstracts, CINAHL, MEDLINE
  • Keywords: Bias field correction, Deep learning, Diffusion weighted imaging, Stroke territory, Voxel wise prediction
  • Dokuz Eylül University Affiliated: Yes


Purpose: To build a stroke territory classifier model in DWI by designing the problem as a multiclass segmentation task by defining each stroke territory as distinct segmentation targets and leveraging the guidance of voxel wise dense predictions. Materials and Methods: Retrospective analysis of DWI images of 218 consecutive acute anterior or posterior ischemic stroke patients examined between January 2017 to April 2020 in a single center was carried out. Each stroke area was defined as distinct segmentation target with different class labels. U-Net based network was trained followed by majority voting of the voxel wise predictions of the model to transform them into patient level stroke territory classes. Effects of bias field correction and registration to a common space were explored. Results: Of the 218 patients included in this study, 141 (65%) were anterior stroke, and 77 were posterior stroke (35%) whereas 117 (53%) were male and 101 (47%) were female. The model built with original images reached 0.77 accuracy, while the model built with N4 bias corrected images reached 0.80 and the model built with images which were N4 bias corrected and then registered into a common space reached 0.83 accuracy values. Conclusion: Voxel wise dense prediction coupled with bias field correction to eliminate artificial signal increase and registration to a common space help models for better performance than using original images. Knowing the properties of target domain while designing deep learning models is important for the overall success of these models.