Learn
Developer-focused guides to building against DICOM: protocols, hospital networking, mTLS, STOW-RS webhook receivers, and the tooling to test it all.
Start here
Integration path
Building something that receives DICOM from hospitals? Read these six in order: protocols, then network constraints, then security and the receiver itself. Each one also lives in its topic section below.
Protocols
What goes over the wire, and who initiates it.
- 1DICOM C-STORE vs DICOMweb STOW-RSC-STORE and STOW-RS both deliver DICOM instances — here's how the classic and web-based transfer protocols differ, and when to use each.
- 2Push vs pull in medical imaging workflowsWhy imaging modalities and PACS push data upon acquisition instead of exposing pull APIs, and how this shapes cloud ingestion architecture.
Network
Why the hospital side constrains your design.
- 3Why hospitals cannot open inbound ports for cloud servicesWhy hospital network security policies block inbound firewall ports, how NAT tracking works, and how to connect cloud services with outbound-only traffic.
- 4Receiving DICOM from a hospital over the internetHow to accept live DICOM studies from a hospital's scanners over the public internet, and what changes when the sender isn't on your LAN.
Secure & build
Authenticate the sender, then build the receiver.
- 5Mutual TLS for DICOM connectionsWhy mutual TLS (mTLS) is the right way to authenticate and encrypt DICOM connections over the internet, and how it differs from a regular HTTPS client.
- 6Designing a resilient webhook receiver for DICOMweb STOW-RSHow to build a STOW-RS webhook receiver that streams large binary multipart payloads directly to cloud storage without memory buffering or timeouts.
DICOM fundamentals
The data model and wire concepts you need before writing any DICOM code: identifiers, hierarchy, encoding.
What is a DICOM Application Entity (AE) title?
An AE title names a DICOM service in up to 16 characters. Learn how AE titles work, why they fail, and why they provide identity without security.
DICOM Instances, Series, and Studies explained
Understand the four-level Patient-Study-Series-Instance hierarchy in DICOM, how unique identifiers work, and why imaging arrives as many separate files.
Understanding DICOM UIDs
Understand how DICOM Unique Identifiers (UIDs) guarantee global uniqueness across studies, series, and instances without a central coordinator.
Start here · step 1DICOM C-STORE vs DICOMweb STOW-RS
C-STORE and STOW-RS both deliver DICOM instances — here's how the classic and web-based transfer protocols differ, and when to use each.
What is DICOM Modality Worklist (MWL)?
DICOM Modality Worklist connects hospital scheduling with imaging scanners to eliminate manual data entry, prevent patient mismatch, and secure image intake.
The DICOM file format: preamble, DICM, and File Meta
Understand the DICOM Part 10 file format structure, including the 128-byte preamble, the DICM prefix, File Meta Information, and raw dataset differences.
Implicit vs Explicit VR in DICOM
Compare Implicit and Explicit Value Representation in DICOM. Learn why modern systems mandate Explicit VR and how Implicit VR creates parsing bugs.
DICOM Value Representations (VR) explained
Understand DICOM Value Representations, data element length limits, text padding rules, and why picking the wrong VR corrupts attributes silently.
DICOM SOP Classes and Transfer Syntaxes explained
Understand how SOP classes define clinical operations and transfer syntaxes control byte encoding, and why association negotiation depends on both.
DICOM C-ECHO: Verification and connectivity testing
DICOM C-ECHO tests application readiness, AE titles, and TLS handshakes without sending patient data. Learn how it works and how to troubleshoot it.
DICOM Character Sets and Patient Names Explained
Understand Specific Character Set (0008,0005) in DICOM, ISO 2022 escape sequences vs UTF-8, Person Name component groups, and how to avoid garbled text.
DICOM Compression: JPEG 2000 vs JPEG-LS vs HTJ2K
Compare DICOM compression transfer syntaxes including JPEG 2000, JPEG-LS, and HTJ2K by decoding speed, library support, and ingestion architecture.
DICOM Structured Reports (SR) explained
Learn how DICOM Structured Reports organize clinical measurements and AI findings into hierarchical content trees using templates like TID 1500.
DICOM private tags: what they are and how to handle them
Learn how DICOM private tags use odd groups and creator reservations, how to extract proprietary vendor metadata safely, and how to avoid PHI leakage.
DICOM Pixel Data for AI: Rescale, Windowing, and Spacing
Learn how to transform raw DICOM pixel data for machine learning models using Rescale Slope, Photometric Interpretation, window leveling, and pixel spacing.
Enhanced Multi-Frame DICOM: Functional Groups Explained
How enhanced multi-frame DICOM stores a whole scan in one file, how functional groups describe each frame, and what changes for viewers and ML loaders.
Developer tooling & test labs
DCMTK commands, Docker test labs, CI conformance tests, and libraries for building and debugging against DICOM.
Build a local DICOM test lab with Docker
Set up a local DICOM test lab with Docker Compose, Orthanc PACS, and DCMTK to send test studies and verify DIMSE associations without hospital hardware.
pydicom vs dcm4che vs fo-dicom vs dcmtk: which to choose
Compare pydicom, dcm4che, fo-dicom, and DCMTK by actual engineering use case, from Python machine learning pipelines to enterprise Java PACS archives.
DCMTK commands for developers: a working reference
A verified, practical reference for echoscu, storescu, dcmsend, storescp, dcmrecv, dcmdump, dcmodify, and dcmconv — the DCMTK commands you'll actually run.
Testing DIMSE Over Mutual TLS Locally
Generate a throwaway test CA with OpenSSL and test DICOM DIMSE over mutual TLS with DCMTK, decoding certificate, cipher, and AE title errors.
Synthetic DICOM test data and PS3.15 de-identification
Create safe synthetic DICOM test data, implement PS3.15 Annex E de-identification profiles, and remove real patient fixtures from your repositories.
Debugging a failed DICOM association: A-ASSOCIATE-RJ
Learn how to diagnose rejected DICOM associations, decode A-ASSOCIATE-RJ and A-ABORT error bytes, and inspect wire captures using tcpdump and Wireshark.
DICOM Conformance Testing in CI
Learn what DICOM conformance checks to automate in CI pipelines—from association negotiation to transfer syntaxes—and what still needs manual testing.
Load-testing a DICOM receiver: association throughput
Push your DICOM receiver past its comfort zone with reproducible benchmarks, comparing association churn against large-study payloads with DCMTK.
Hospital connectivity & network security
Firewalls, NAT, VPNs, and mTLS: how DICOM actually gets out of a hospital network and reaches your service.
Why hospitals cannot open inbound ports for cloud services
Why hospital network security policies block inbound firewall ports, how NAT tracking works, and how to connect cloud services with outbound-only traffic.
Start here · step 4Receiving DICOM from a hospital over the internet
How to accept live DICOM studies from a hospital's scanners over the public internet, and what changes when the sender isn't on your LAN.
Site-to-site VPN vs internet-facing endpoints for DICOM
Compare site-to-site IPsec VPNs and internet-facing mTLS endpoints for hospital DICOM ingestion, including operational complexity, latency, and costs.
Start here · step 5Mutual TLS for DICOM connections
Why mutual TLS (mTLS) is the right way to authenticate and encrypt DICOM connections over the internet, and how it differs from a regular HTTPS client.
Client certificates vs API keys for healthcare integrations
Why medical devices and imaging gateways use X.509 client certificates instead of bearer tokens or API keys, and what it means for your cloud ingest.
Zero-retention DICOM pipelines for HIPAA and GDPR
Compare stateless forwarding with transient disk caching in DICOM pipelines, and learn how zero retention simplifies HIPAA BAAs and GDPR audits.
PACS, VNA, and DICOM routers: where your product fits
Learn how hospital PACS, VNA, and DICOM routers interact, which system sends studies to your product, and how to identify the team managing the connection.
Hospital security questionnaires for imaging vendors
Learn what hospital security questionnaires ask imaging software vendors, how MDS2 fits, and how to keep answers accurate across each release and provider.
Integration patterns & webhooks
Designing the receiving side: STOW-RS webhooks, streaming, idempotency, and routing once DICOM reaches you.
Push vs pull in medical imaging workflows
Why imaging modalities and PACS push data upon acquisition instead of exposing pull APIs, and how this shapes cloud ingestion architecture.
DICOM C-FIND vs DICOMweb QIDO-RS: Querying for Studies
Compare DICOM C-FIND with DICOMweb QIDO-RS. Learn wire protocols, query levels, return keys, and why cloud ingest pipelines rarely query hospital PACS archives.
Start here · step 6Designing a resilient webhook receiver for DICOMweb STOW-RS
How to build a STOW-RS webhook receiver that streams large binary multipart payloads directly to cloud storage without memory buffering or timeouts.
Idempotency and duplicate detection for DICOM ingest
Learn how to handle duplicate DICOM studies and retries idempotently using SOP Instance UIDs without corrupting multi-slice medical imaging pipelines.
Why DICOM routing does not support multi-destination fan-out
Learn why DICOM C-STORE requires synchronous point-to-point delivery, how multi-destination fan-out causes failure ambiguity, and how to decouple ingest.
Handling large DICOM studies: CT and PET transfers
Learn how to ingest multi-gigabyte CT and PET DICOM studies without timeouts, buffer overruns, or dropped slices in your cloud webhook pipelines.
Sorting DICOM Slices: Image Position vs Instance Number
Learn why sorting DICOM slices by Instance Number fails, and how to use Image Position and Orientation to build anatomically correct 3D imaging volumes.
Regulatory & quality
MDR, FDA, and quality-system basics for teams shipping cloud medical software.
Is your cloud software a medical device?
How the EU MDR and FDA SaMD frameworks decide whether cloud software touching imaging data is a medical device, and how MDR classification works.
MDR Class I vs IIa/IIb: Software Classification Rules
Learn how EU MDR Rule 11 classifies medical software into Class I, IIa, or IIb, and where data pipelines sit when they do not drive clinical decisions.
FDA Software Routes: 510(k), De Novo, and Exemptions
Learn how FDA pathways regulate medical software, when to use 510(k) or De Novo, what 510(k) exempt means, and why imaging AI cannot claim CDS exclusion.
MDCG 2019-11: Software Qualification & Boundaries
Walk through the MDCG 2019-11 decision tree to qualify medical device software, separate communication pipelines, and document valid module boundaries.
IEC 62304 Software Safety Classes: Class A, B, and C
Understand IEC 62304 software safety classification for Classes A, B, and C, and learn what documentation small teams must produce for each tier.
ISO 14971 Risk Management for Imaging Data Pipelines
Learn how to apply ISO 14971 risk management to medical imaging pipelines with worked hazard examples for wrong recipients, lost studies, and silent truncation.
IEC 81001-5-1 & MDCG 2019-16: Medical Device Cybersecurity
Understand what IEC 81001-5-1 and MDCG 2019-16 require for medical device cybersecurity, including SBOM generation and vulnerability disclosure.
Medical Device Software Change Control: Significant Changes
Learn when a software update requires a new FDA 510(k) or notified body review, how to evaluate significant change, and how to maintain lean change control.
Architecture & operations
Build-vs-buy tradeoffs, multi-tenancy, and operational patterns for running a DICOM data pipeline.
Understanding DICOM Conformance Statements
Learn how to read a vendor DICOM Conformance Statement to verify supported SOP classes, transfer syntaxes, and AE roles before starting an integration.
Onboarding a hospital modality: The pre-flight checklist
Exchange AE titles, ports, hostnames, and TLS certificates with a hospital PACS administrator to complete your first DICOM transfer without silent errors.
Building a custom DICOM listener vs a managed gateway
The real engineering cost of a DICOM listener that serves many hospitals at once — associations, transcoding, cert renewal — versus a managed gateway.
Multi-Tenant DICOM Architecture in the Cloud
How to isolate hospital senders in a multi-tenant cloud DICOM architecture, prevent Patient ID collisions, and authenticate connections before parsing bytes.
Cloud DICOM Ingestion: Per-Study vs Per-Gigabyte Pricing
Compare per-study and per-gigabyte cloud DICOM ingestion pricing models. Learn how imaging study size variance impacts network costs and billing.