Radiology AI Podcasts | RSNA Radiology AI Papers in a Capsule Series-Episode 2
9 snips
May 30, 2025 Dive into groundbreaking research focusing on AI's role in brain MRI for Alzheimer's assessment. Discover how the NeuroHarmony model tackles scanner variability for more reliable imaging. Learn the significance of harmonization in neuroimaging and its impact on early disease detection. Explore the benefits of using multi-class models to enhance Alzheimer’s classification. Understand the advancements in integrating volumetric MRI data across different scanners, paving the way for improved evaluation tools in neurodegenerative disease assessment.
AI Snips
Chapters
Transcript
Episode notes
NeuroHarmony Tackles Scanner Variability
- NeuroHarmony extension addresses scanner variability in brain MRI for Alzheimer's assessment. - This improves reliability of quantitative imaging in diverse clinical settings.
Quantitative Imaging Overcomes Subjectivity
- Quantitative brain volumetric measurements provide objective assessments over subjective visual atrophy assessment. - Scanner differences can cause volumetric variability up to 15%, obscuring disease-related changes.
Need to Account for Disease Pathology
- Original NeuroHarmony used image quality metrics trained on healthy controls only. - Disease pathology affects metrics, so new harmonization must consider cognitive status.