Reference Contrasts In, Target MRI Out When You Only Have a Handful of Raw Scans
Hang fully sampled reference contrasts beside an undersampled target. CoSMo-RecNet freezes a public-DICOM content/style prior, trains a light RecNet on a handful of paired raw scans, and returns the reconstructed target for low-field and ultra-low-field brain MRI. arXiv:2609.01959.
Parasternal Long-Axis Cine In, Ejection Fraction Out When Apical Windows Fail
Hang a PLAX cine when the apical window is missing or foreshortened. Gao, Yurk, and Abu-Mostafa map that view to study-level EF, with optional A4C fusion. arXiv:2609.02969.
Lesion Frame In, Background-Quieted Overlay Out Before Encoding
Hang a lesion frame with cluttered background. FreNet injects a SAM visual prior into the pixels before encoding, reconfigures features in frequency and space during encoding, and returns a cleaner lesion overlay across ultrasound, dermoscopy, and endoscopy. arXiv:2609.03535.
Portable Neonatal Brain MRI In, Seven Artifact Severity Grades Out
Hang a 0.064 T neonatal brain T2. LoFi RADIO distills three foundation teachers into one ViT-S student and returns 0/1/2 severity for seven acquisition artifacts. arXiv:2609.02676.
Low-Dose CT In, Cleaner Stack Out Without Paired Normal-Dose Training
Hang one low-dose CT measurement. PIPG-LDCT builds two physics-driven noisy training branches from that single acquisition, trains a Noise2Noise-style U-Net, and returns a cleaner stack without paired normal-dose labels. arXiv:2609.02654.
Rough Bone Brush In, Refined Ultrasound Overlay Out
Hang a bone US frame, paint a rough brush on the interface. ExiL completes and refines that imperfect mask with a mask-conditioned U-Net and expert-mode updates. arXiv:2609.00473.
Look at the Structure: Gaze Fixations Become Overlays Without Retraining
Hang a colonoscopy frame or prostate MRI slice, look at the structure. GazeRefine turns duration-weighted fixations into a dense overlay on frozen DINOv3. No labels, no fine-tuning. arXiv:2609.01310.
CT or MRI In, Seven Whole-Heart Overlays Without a Modality Retrain
Cardiac CT or MRI in. Seven chamber and vessel overlays out from one 3D nnU-Net, no modality flag. Label-to-image mix (LISynSeg), Vanderbilt/Mayo. arXiv:2608.31073.
Axial CT Slice In, Organ Boxes and an Auditable Left/Right Call Out
One axial CT slice plus a binary spatial question. YOLO boxes, centroid geometry, then a 1/0 with an audit trail. YOLO weights ship; VLM checkpoints do not.
How to Convert a NIfTI File into a DICOM Series Using Python
This blog is about how to use the nifti2dicom functin to convert nifti file or files into dicom series.