PYCAD BLOG

Insights on Medical Imaging, AI & DICOM

Guides, project breakdowns, and lessons learned from building medical imaging platforms — written by the PYCAD team.

Comparison built from Fig. 6a of arXiv:2611.06471 by Yulu Xie, Zilong Li, Hao Luo, Fan Wang and co-authors: clinical CT slices of laryngeal, lung and thyroid lesions from three hospitals, with reference outlines in red and zero-shot masks in yellow from VoxTell (top row) and HoloCT-8B (bottom row).

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.

November 5, 2026

Comparison built from Extended Data Fig. 6a,b of arXiv:2611.05011 by Aashish Chapagai, Angad Singh Gurtatta and Erik Cambria: two NIH ChestX-ray14 test radiographs where DenseNet-121 flagged infiltration, one correct and one wrong, each shown with Grad-CAM, LIME and SHAP heatmaps of nearly equal concentration on the chest.

Saliency Maps Do Not Flag Chest X-Ray AI Errors at New Hospitals

A three-site study finds Grad-CAM, LIME and SHAP look the same on right and wrong calls. Model confidence is the better alarm.

November 4, 2026

Before and after frames from the project-page video of arXiv:2611.03996 by Necati Sefercioglu, Mehmet Ozan Unal, Metin Ertas, Altan Cakir and Isa Yildirim: sagittal, coronal and axial slices of IXI test volume t0145 with simulated bias field, spike and noise, and the volume returned by the Dx-Rx agent.

Dx-Rx: An MRI Agent That Measures Artifacts Before Fixing Them

Dx-Rx gives a language model five MRI artifact detectors and four classical repairs, lets it undo its own steps, and checks its report against the log.

November 4, 2026

Before and after figure built from Figure 2 of arXiv:2611.01902 by Wei Jiang, Hongli Chen, Tianyi Ding, Feng Liu, Nan Ye and Hongfu Sun: central sagittal brain MRI slices from a real motion-corrupted scan and a simulated case, shown uncorrected, after the MotionTTT baseline, after EventMC, and the motion-free reference.

EventMC: Training-Free Motion Correction for 3D Brain MRI

EventMC finds when the head moved, fits each k-space segment to a clean reference and rebuilds one sharp brain volume.

November 3, 2026

Before and after figure built from Figure 3 of arXiv:2610.40722 by Nishi Kanta Paul, Md Shihabul Islam Shovo and Désiré Sidibé: two frontal chest radiographs (effusion and atelectasis cases) shown as original, autoencoder reconstruction, LAPIS anonymized release and a change map that is smooth and mostly peripheral.

LAPIS: Chest X-Ray Anonymization Beyond the DICOM Header

LAPIS swaps a small learned identity code inside each chest X-ray so released images are harder to link back to the patient.

November 3, 2026

Before and after figure built from Figure 5 of arXiv:2610.37740 by Jiening Zhu, Yabo Fu and colleagues (Memorial Sloan Kettering Cancer Center): axial, coronal and sagittal chest slices reconstructed from 10 projections with standard FDK (streaks), NAF per-scan fitting (smooth, soft tissue lost), EUReCA (vessels and body wall recovered) and the reference CT.

EUReCA: One Network for Sparse-View Linac CBCT

Any 8 to 100 projections on a Halcyon arc, one set of weights, and a CBCT volume in about a second instead of minutes.

November 2, 2026

We build custom medical imaging platforms — advanced DICOM viewers, AI segmentation, and the clinical systems around them.