PYCAD builds custom web DICOM viewers and medical-imaging AI — annotation through model deployment — for teams that already have a scanner, a PACS, or a product and need that piece built. This page is what the company actually does. It is not an all-in-one imaging platform. It is not a SKU list. It is not a course catalog.
If you meant founders, podcasts, press → about us. If you meant named client work → case studies. If you meant the hospital software stack (viewer / PACS / VNA) → medical imaging software. If you meant how to start learning the field → medical imaging starter.
The 2022 version of this URL was a first-person index of conversion notebooks and a MONAI waitlist. That index is not the company. The public tutorials that still live are linked at the bottom. The waitlist landing is gone.
What the work is
| Job | What you get | What you do not get |
|---|---|---|
| Custom web DICOM viewer | A browser viewer for a real series: hang, window, MPR, sometimes a store. Built to the product, not a white-label of a public demo | A boxed PACS. A cleared workstation unless the contract says so. The public OHIF tour is a different URL |
| Imaging AI, end to end | The unglamorous chain: labels → train → package → sit next to the viewer. One organ, one question | A foundation model for every scan. A “diagnosis API.” MONAI as a PYCAD product — the toolkit is this page |
| The glue | DICOM in, a result the clinician can see. Conversion, de-id, a hanging protocol — when the viewer or the model needs them | A conversion SaaS. /pycad-convert/ 404s. The conversion how-tos stay on their own slugs |
A typical brief is “we have a dental or surgical planning app and the viewer is the hole,” or “we have a model and nowhere honest to show it.” The case studies page is the named list. This page is the job description.
What PYCAD is not
- Not an all-in-one imaging platform you subscribe to. There is no PYCAD Cloud, no PYCAD PACS, no PYCAD Viewer SKU on a price card.
- Not a radiologist group, a teleradiology service, or a staffing desk. Workforce is shortage of radiologists.
- Not a photo-x-ray app, a cine workstation, or a Gaussian-filter product. Those words are other URLs.
- Not the 2022 “premium MONAI course with lifetime support.” That landing 301s into machine learning for medical imaging.
When a sentence on this site says “we integrate into unified medical imaging platforms,” read it as: the viewer or the model is built to sit in your product. It is not a claim that PYCAD ships the platform.
Where the rest of the site lives
Founders and talks: about us. Named work: case studies — not /portfolio. How a web DICOM viewer is assembled in public: OHIF and the viewer software list. How a model is trained in public: the MONAI preprocess and aug notebooks, the mandible write-up, the liver-segmentation series.
Public conversion how-tos that still open (the 2022 index, cleaned):
- DICOM → JPG/PNG
- DICOM series → NIfTI
- NIfTI → DICOM series
- NumPy array → NIfTI
- MONAI 3D preprocess and 3D aug
GitHub that matches those posts lives under amine0110. YouTube is youtube.com/c/pycad. Neither is a product.
If the missing piece is a viewer or a model in a real product, that is the work. Case studies.