Update collections/_posts/2026-01-17-mad-ad.md

Co-authored-by: Nathan Painchaud <23144457+nathanpainchaud@users.noreply.github.com>

Author: olivier-bernard-creatis

collections/_posts/2026-01-17-mad-ad.md

@@ -31,7 +31,7 @@ The full presentation, including links to the diffusion model (DDPM), is availab
 
 # Highlights
 
-* Extending the idea proposed in the THOR method [1] to the training procedure 
+* Extending the idea proposed in the THOR method [^1] to the training procedure 
 * Learning to remove regions with anomalies of varying sizes using a diffusion process
 * Evaluation from three public datasets: `IXI Dataset` (brain MRI scans from approximately 600 healthy subjects), `ATLAS 2.0` (655 T1-weighted MRI scans accompanied by expert-segmented lesion masks), and `BraTS'21` (1251 brain scans across four modalities: T1-weighted, contrast-enhanced T1-weighted - T1CE, T2-weighted, and T2 Fluid Attenuated Inversion Recovery - FLAIR)