SPIE Medical Imaging is the annual symposium run by SPIE (the international society for optics and photonics) where imaging physics, processing, CAD, perception, image-guided procedures, ultrasound, informatics, and digital pathology present papers. It is not “a medical imaging conference” in the generic sense. It is not a PYCAD event. It is not a registration landing for one year.
If you meant the 3D imaging field → 3D medical imaging. If you meant the career on-ramp → medical imaging starter. If you meant CAD / diagnosis models → AI for medical diagnosis. If you meant imaging ML as a job → machine learning for medical imaging.
PYCAD builds custom web DICOM viewers and imaging models. It does not run SPIE, sell a SPIE pass, or ship a “SPIE toolkit.”
What the meeting is
SPIE Medical Imaging is a week of parallel conferences (tracks), not one keynote-plus-expo. People present a paper; the paper is supposed to exist. The audience is physicists, engineers, computer scientists, and the clinicians who will actually use the method. Radiology society meetings (RSNA and siblings) are a different room — more clinic and vendors. This one is the methods meeting.
The city and the dates move. Recent habit is mid-February. 2026 was 15–19 February in Vancouver, BC. Do not treat that row as a live registration CTA; check spie.org/conferences-and-exhibitions/medical-imaging for the year you mean.
| What it is | What it is not | |
|---|---|---|
| The symposium | One week, several named conferences under one badge | A trade-show aisle with a poster afterthought |
| A “conference” | A track with its own chairs, CFP, and proceedings volume | A hallway conversation |
| A paper | What you submitted, presented, and (usually) published in the proceedings | A slide deck that never became a citeable document |
| The Digital Library | Where those volumes live after the week | The Journal of Medical Imaging (that is a separate SPIE journal) |
The conferences (tracks)
SPIE publishes the current list on the program page. Names shift slightly by year; these are the standing ones:
| Conference | The job of that room |
|---|---|
| Physics of Medical Imaging | Detectors, dose, reconstruction from the measurements — the gantry and the inverse problem |
| Image Processing | After the volume exists: denoise, register, reconstruct a display, a 3D from slices |
| Computer-Aided Diagnosis | A mark or a class a reader can accept or reject. Detection and diagnosis as a system, not only a net |
| Image-Guided Procedures, Robotic Interventions, and Modeling | The scan as a map in a room that has tools in it. IGS itself is imaging guided surgery |
| Image Perception, Observer Performance, and Technology Assessment | Whether a human (or a model) actually sees the thing, and how you measure that |
| Clinical and Biomedical Imaging | A method tied to a named clinical or preclinical question, not only a phantom |
| Imaging Informatics | The file, the archive, the worklist, the report — PACS/VNA/DICOM as research |
| Ultrasonic Imaging and Tomography | Ultrasound as a physics and reconstruction problem, not only a bedside loop |
| Digital and Computational Pathology | Whole-slide images. Pathology AI as a method meeting, not a LIS brochure — AI in pathology |
Application tracks (AI/ML, and in some years sustainability or global health) cut across those rooms so you can follow a topic without living in one conference.
How proceedings work
You submit to a conference. Chairs accept a talk, a poster, or both. The written paper is published in a Proceedings of SPIE volume for that conference and year, in the SPIE Digital Library. Two ISSNs you will see: 0277-786X (Proceedings of SPIE) and 1605-7422 (Progress in Biomedical Optics and Imaging — the series many Medical Imaging volumes sit in).
That is a conference paper: citeable, indexed, usually out fast. It is not automatically a journal article. SPIE’s peer-reviewed journal for this community is the Journal of Medical Imaging (JMI). A proceedings paper can later become a JMI paper; they are not the same object. A paid symposium registration has historically included a bundle of Digital Library downloads — the count is on that year’s registration page, not a promise this URL will keep.
If you only need the PDF later, you do not need the badge. If you need the question after your talk, you do.
What this page will not do
- Not a 2026 (or 2027) registration CTA. The meeting is annual; the city is not this URL’s job.
- Not a market size for “medical imaging” or “AI in imaging.” Dropped.
- Not a generic stamp of 515. 515 is how a person starts. This page is what the symposium is.
- Not a PYCAD SPIE product, booth, or “we turn SPIE talks into a platform” pitch. Case studies if the missing piece is a viewer or a model.