moco-dl: Fast Slice-Wise Motion Correction for Spinal Cord fMRI

A tiny CNN aligns every axial slice of a spinal cord fMRI run on a CPU in seconds and beats SCT’s current tool at 3T and 7T.
LoopMR: Accelerated MRI Reconstruction With One Looped Block

One Restormer block, run 12 times, rebuilds undersampled brain, knee, heart and carotid MRI with a ninth of PromptMR+’s parameters.
CADS: One Open Model Suite for 167 Whole-Body CT Structures

Trained on 22,022 curated CT scans from 100+ centres, CADS segments head to knees, runs in 3D Slicer and outputs SNOMED CT labels.
MicroBT: Finding and Counting Sub-3 mm Brain Tumors on MRI

A centroid-guided nnU-Net extension segments brain tumors on MRI and counts each lesion, aimed at the tiny metastases that models miss.
P2Echo: One Text-Prompted Model Segments Four Echo Views

UBC’s P2Echo takes a view-and-structure prompt and draws LV, MYO, LA, RV or RA masks in 2CH, 3CH, 4CH and PSAX with one network.
UltraVoxelGS: 3D Ultrasound From Tracked Sweeps in Seconds

Hong Kong Baptist University turns tracked 2D ultrasound frames into a 3D Gaussian scene in about 10 seconds, no per-scan fitting.
BrainGFM: Synthetic Brain Scans From Few Labels

One annotated case adapts a LoRA. BrainGFM then synthesizes image-mask pairs, class images, and sequence transforms for MRI and CT.
LungMend: Fix Incomplete Lung Masks on Severe TB CT

R231 drops cavities and consolidations on severe TB CT. LungMend fills them from a healthy shape prior and bone cues.
Virtual Contrast Breast MRI That Spares Gadolinium for Some Patients

A 3D U-Net predicts virtual CE from non-contrast breast MRI. NC+VCE reaches AUROC 0.901; about 40% of patients could skip acquired CE.
HoloCT: A CT Model That Backs Every Finding With a 3D Mask

Alibaba DAMO’s open CT model screens, answers and drafts reports, and draws a 3D mask for each finding so a reader can check it.