SBOM for FDA Premarket Submissions for Imaging AI & SaMD
FDA-aligned SBOMs for imaging AI and SaMD - Python ML stacks, container layers, model weights, and DICOM toolkits - with optional VEX statements documenting exploitability.
Aligned to FDA's February 3, 2026 premarket cybersecurity guidance.
SaMD products - especially imaging AI - have the messiest software bills of materials in MedTech: a Python ML stack with hundreds of transitive deps, container base layers, CUDA/cuDNN binaries, DICOM toolkits, model weights treated as code, and frequently a JS/TS frontend on top. A naive pip-freeze SBOM fails reviewer scrutiny because it under-reports OS-level components and doesn't describe the inference runtime or model provenance. Our SBOM service for this segment produces our approach to supporting Section 524B compliance: a complete, layered SBOM that includes the OS, the runtime, the Python deps with resolved versions, the model artifacts, and an optional VEX (Vulnerability Exploitability eXchange) document that explains why high-severity CVEs in your stack are or are not exploitable in your specific deployment.
This matters because the imaging-AI stack is full of high-CVSS Python and CUDA CVEs that are not exploitable in your container - and an optional VEX statement can document the applicability assessment. We generate CycloneDX 1.5 / SPDX 2.3 SBOMs from the build pipeline (not from runtime introspection alone), enrich them with NVD/OSV/GHSA data, triage every CVE above your defined threshold, and produce VEX statements with the four CISA VEX statuses: not_affected, affected, fixed, under_investigation. We also document model provenance: training data lineage as FDA AI guidance recommends, weight file hashes, and the model-card link - because reviewers increasingly treat the model itself as a SBOM component.
Layers we exercise in this engagement
The imaging & ai/samd system, from the outermost cloud and clinician surfaces down to the device itself. Highlighted layers are exercised by this medical device sbom services.
- 01OS base image Tested
- 02CUDA / cuDNN Tested
- 03Python ML stack Tested
- 04DICOM toolkits Tested
- 05Model weights Tested
- 06JS / TS frontend Tested
Layers shown outermost (top) to innermost (bottom). Dashed rows are part of the surrounding system but out of scope for this view.
Medical Device SBOM Services engagement, end to end
Four phases, fixed fee, scoped to imaging & ai/samd architecture from kickoff onward.
- 01
Build-pipeline integration
SPDX 2.3 SBOMs generated from your actual build, not from runtime introspection alone.
- 02
Enrichment + triage
Components enriched from NVD, OSV, and GHSA; every CVE above your threshold triaged for exploitability.
- 03
VEX authoring
VEX statements (not_affected, affected, fixed, under_investigation) with documented justifications, where the manufacturer chooses to use VEX.
- 04
Postmarket handoff
SBOM + VEX delivery hooked into your QMS so postmarket monitoring continues after submission.
What we see in Imaging & AI/SaMD medical device sbom services
The patterns we hit in this segment, this service, again and again.
OS-layer components missing from runtime SBOM
pip-freeze captures Python deps but misses base-image OpenSSL, glibc, ImageMagick - exactly where the published CVEs land for this segment.
Model weights not represented
Reviewers asked 'where is the model in your SBOM?' and the answer was nowhere. We add weight artifacts as components with hash + provenance.
DICOM toolkit version drift across deployments
DCMTK / pydicom pinned in dev, floating in production via base-image rebuilds. SBOM diff between submitted and deployed shipped silently.
VEX statements absent for >100 high-CVSS Python CVEs
VEX is optional and can help document exploitability. We triage every high-CVSS finding to one of four statuses with justification.
"Blue Goat Cyber takes the burden off our engineers and makes FDA cybersecurity requirements easy to understand. Their expertise and smooth process mean we can focus on our product, not the paperwork. The organized documentation, perfectly formatted for eSTAR, saves us countless hours."
Standard Medical Device SBOM Services deliverables
The same deliverables the parent Medical Device SBOM Services service ships with - tuned to your imaging & ai/samd architecture.
- SPDX and CycloneDX generation
- Component vulnerability mapping (CVE / KEV)
- End-of-life and replacement planning
- Build-system and binary SCA validation
What lands in your eSTAR submission
Documents structured for the cybersecurity attachments of your submission - no reformatting on your side.
- SPDX and CycloneDX generation
- Component vulnerability mapping (CVE / KEV)
- End-of-life and replacement planning
- Build-system and binary SCA validation
Standards that apply
The Imaging & AI/SaMD baseline, plus the call-outs that matter for medical device sbom services in this segment.
Segment-specific call-outs
FDA Section 524B + 2026 final premarket guidance
SBOM is statutory; a machine-readable format (CycloneDX or SPDX) is recommended. VEX is optional but useful for SaMD given the noise level of the Python ecosystem.
AI/ML PCCP guidance
Model artifacts must be tracked as SBOM components if your PCCP allows model updates - otherwise reviewers can't tell what's actually deployed at any given time.
What's not in scope
We scope tightly on purpose. These items are either out-of-scope by design or belong in a separate engagement - we'll tell you up front, not after kickoff.
- Penetration testing of components in the SBOM
- Code refactoring to remove vulnerable dependencies
- License-compliance legal review (we surface, your counsel rules)
Medical Device SBOM Services for Imaging & AI/SaMD - FAQs
The questions buyers in this segment actually ask before scoping a medical device sbom services engagement.
Go deeper on Imaging & AI/SaMD and premarket
A practical, ungated buyer's guide for medical device manufacturers evaluating cybersecurity partners, what goes wrong, why it costs you, and what to demand from your next engagement. Aligned to the FDA February 2026 premarket guidance.
The most common high- and critical-severity findings we surface in medical device penetration tests, what each one looks like in the field, and how to fix it before your FDA submission.
A practical, ungated guide to the threat modeling gaps that trigger FDA cybersecurity questions in 510(k), De Novo, and PMA submissions - and exactly how to close them before reviewers find them.
Medical device cybersecurity should scale with device risk. Christian Espinosa argues why meeting only the Section 524B minimum leaves patients and submissions exposed.
VEX resources for medical device teams: a document guide, SBOM comparison, and common assessment mistakes, with evidence-based status and regulatory context.
Medical device cybersecurity for nanotech neurostimulation: a fictional STRIDE threat model linking controller risks, patient harms, controls, and testing.
Other engagements for Imaging & AI/SaMD
Teams in this segment commonly bundle these alongside medical device sbom services.
Keep going
Scope a Medical Device SBOM Services engagement for your imaging & ai/samd program.
A 30-minute discovery call with Christian Espinosa, Founder and CEO, and Kristy Kennedy, VP of Sales.