We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Hematoma expansion prediction: still navigating the intersection of deep learning and radiomics.
- Authors
Le, Nguyen Quoc Khanh
- Abstract
A recent study published in European Radiology explores the potential of deep learning (DL) algorithms in predicting hematoma expansion (HE) in spontaneous intracerebral hemorrhage (sICH). The study demonstrates the efficiency and accuracy of DL algorithms in automating image segmentation and classification tasks, surpassing traditional radiomics approaches. By incorporating semantic features and optimizing radiomics features, the DL radiomics model outperforms single models in predicting HE. However, the study acknowledges limitations such as potential selection biases, small sample size, and subjectivity in manual segmentation. Future research should focus on prospective study designs, larger sample sizes, and standardization of segmentation protocols to enhance the reliability and generalizability of DL in radiomics for sICH. Overall, this study highlights the transformative potential of DL in improving clinical outcomes and patient care in neuroimaging.
- Subjects
RADIOMICS; DEEP learning; HEMATOMA; MACHINE learning; COMPUTER-assisted image analysis (Medicine); INTRACEREBRAL hematoma
- Publication
European Radiology, 2024, Vol 34, Issue 5, p2905
- ISSN
0938-7994
- Publication type
Article
- DOI
10.1007/s00330-024-10586-x