Comparison of Radio Frequency vs beamformed ultrasound data for infarct classification with CNNs
Ultrasound is the most practical, cost-effective and patient-friendly clinical imaging modality. However, Computed Tomography and Magnetic Resonance Imaging outperform echography when texture or tissue characterization is clinically required. As AI-methodologies could help resolve this, a Deep Learning algorithm was applied to classify localized infarcted tissue from static ultrasound recordings.
Authors
- Jens‐Uwe Voigt (ORCID: https://orcid.org/0000-0002-0575-1888)
- A Balinisteanu (ORCID: https://orcid.org/0009-0003-8502-6903)
- Paulo Tostes
- Jan D’hooge (ORCID: https://orcid.org/0000-0002-2346-142X)
- Sara Ferreira (ORCID: https://orcid.org/0000-0003-2909-9166)
- Helena Williams (ORCID: https://orcid.org/0000-0002-7970-0290)
Institutions
- Hasselt University (BE)
- KU Leuven (BE)
Publication Details
- Published
- 2024-09-22
- DOI
- https://doi.org/10.1109/uffc-js60046.2024.10794162
- Primary Topic
- Brain Tumor Detection and Classification
- Type
- article
- Field-Weighted Citation Impact
- 0.00