John Snow Labs

09/03/2026

John Snow Labs’ AI Governance Framework conforms to 300+ laws, regulations, and industry standards worldwide, covering nine pillars including risk management, safety, privacy, transparency, fairness, and acceptable use.

The framework is built for regulated healthcare environments:
• Privacy: On-premises, private cloud, and air-gapped deployments keep data within your environment.
• Transparency: Published benchmarks and peer-reviewed research support independent evaluation.
• Fairness: LangTest provides 100+ test types for bias, fairness, robustness, toxicity, representation, and accuracy.

The result: regulatory-grade healthcare AI that is measurable, auditable, and designed for real-world deployment.

Learn more: https://hubs.ly/Q04wwmRz0

09/02/2026

EHR migrations break down at one specific point: translating medical codes across vocabularies.

You have ICD-10. The target system needs SNOMED. Or OMOP. Or UMLS. Someone has to map them - manually, at scale - error-prone and historically without a clean automated path.

The Code-to-Code Mapping feature in the Snow Labs Medical Terminology Server handles this automatically. Provide a source code. Select a target vocabulary - or let the system search across all 16+. Receive the mapped codes with their relationships in seconds.

Mappings are rule-based and curated, not inferred. The same code always maps to the same target. No hallucinated equivalences.

3-minute live UI demo in the comments.
https://hubs.li/Q04wd_FR0

Watch a live walkthrough of the Terminology Server's Concept Maps f...

09/01/2026

Before you route medical coding through a large language model, understand what you're actually getting.

An LLM generates responses. The same input can return different codes on different runs. It has no native access to ICD or SNOMED unless you explicitly provide it. Every API call incurs cost. In a regulated healthcare environment, sending clinical data to a cloud service creates compliance barriers that are hard to justify at scale.

A terminology server is a deterministic lookup engine. Enter "type 2 diabetes" today or next year: you get E11, every time - because it queries a validated, actively maintained vocabulary database. It runs on-premise, behind your firewall. One license. Unlimited lookups. No data leaving your environment.

The two tools complement each other. Use a terminology server for precise, auditable term-level mapping. Use an LLM for document-level inference on unstructured clinical notes. The strongest clinical NLP pipelines use both.

Full video breakdown in the comments.
https://hubs.li/Q04w3QWs0

ChatGPT can describe a disease — but it won't return the same ICD-1...

08/31/2026

Ask a general-purpose LLM to write a DICOM de-identification pipeline and you'll likely get code that looks right and isn't. We tested it: one 10-stage pipeline called a deprecated DicomToImage stage and passed unsupported parameters to PositionFinder and DicomDrawRegions — 7 errors in a single output.

DICOM de-identification is workflow-specific: PHI shows up in metadata tags, free-text fields, burned-in pixels, and embedded PDFs, each needing its own handling, grounding no general-purpose LLM has built in.

The Visual NLP DICOM Skill grounds the same models in verified Visual NLP patterns instead: DicomToImageV3 for frame extraction, ImageTextDetector or ImageTextDetectorV2 for region detection, DicomDrawRegions for redaction and reconstruction, matching Visual NLP's current API surface instead of guessing at it. Type /dicom_tasks for the 8 supported workflows: metadata inspection, strategy-file generation, metadata and free-text de-identification, pixel redaction (zero-shot or blanket), and encapsulated PDF handling.

Generated code still needs review before it touches production data. The skill sits between slow manual discovery and unreliable, ungrounded generation. It doesn't replace validating output against your own data and compliance requirements.

Full walkthrough and download: https://hubs.li/Q04vYKzc0

Generate grounded DICOM de-identification pipelines with the Visual NLP DICOM Skill, supporting metadata, pixel, free-text, PDF, and zero-shot PHI de-identification workflows.

08/25/2026

How do you maintain audit trails when annotation workflows use multiple LLM providers?

Every request needs authentication handling, credential management, rate limiting, and usage logging. When teams connect custom models, they typically build separate infrastructure for each. Quality control fragments. Compliance officers cannot trace which prompts went where.

Generative AI Lab 8.1 uses a proxy-based architecture:

- All LLM requests route through a centralized proxy
- Encrypted API keys, never exposed to users
- Request/response normalization across providers
- Complete audit logs: timestamp, user, model, tokens consumed

Deploy the proxy in your infrastructure. PHI-containing prompts never leave your compliance boundary.

Model flexibility. Complete governance.

Learn more: https://hubs.li/Q04v9flM0

08/21/2026

Type "heart attack." Get SNOMED concept 22298006 and ICD-10 code I21. Type "MI." Same result. Type "myocardial infarction." Same result - because they are the same clinical fact.

The John Snow Labs Medical Terminology Server sits between the free-text your clinicians write and the standardized codes your downstream systems require. It understands clinical meaning across abbreviations, shorthand, typos, and patient-friendly language.

16+ medical vocabularies. On-premise. Deterministic. Full video overview in the comments.
https://hubs.li/Q04twkpQ0

Complete guide to using John Snow Labs' Medical Terminology Server ...

08/20/2026

Your NLP model extracts "metformin 500mg" from a discharge summary. The extraction is correct. Without standardization, the entity cannot connect to any downstream system.

- Your medication reconciliation system needs NDC codes
- Insurance claims require HCPCS
- Research databases expect RxNorm
- EHR integration demands standardized codes

Generative AI Lab's Medical Terminologies solve the translation step directly in the annotation workflow:
- ICD-10, LOINC, CPT, SNOMED CT, RxNorm, and MeSH integrated into annotation
- Automated code resolution during pre-annotation
- Manual lookup available in the interface for human review
- Entity, label, and code travel together through the workflow

Medical codes are the shared language that makes clinical NLP outputs connect to the systems that actually use them.

Learn more: https://hubs.ly/Q04twpV-0

08/19/2026

Annotated clinical datasets stay reusable even when your label schema changes. A concept labeled "Diagnosis" in 2019 and "Primary_Diagnosis" in 2023 no longer blocks the import.

Healthcare AI programs run parallel annotation teams, integrate vendor pre-annotations, and bootstrap new models on earlier training data. Annotations move between projects and systems constantly, and schema drift comes with it.

Generative AI Lab reconciles it at import with no-code label mapping:
- Proposes matches, you approve, it applies them as data is imported
- Human-in-the-loop, every mapping auditable
- Works across NER, assertion, classification, and relation labels
- Source data untouched

Carry years of annotation investment forward instead of rebuilding it.

Link: https://hubs.ly/Q04twkYc0

08/15/2026

Reuse 5,000 already-validated clinical notes in a new project without writing a single script to rename labels.

The blocker is usually naming. One project labeled a concept "diagnosis," another "Diagnosis," a third "Primary_Diagnosis." The clinical meaning is identical, the import fails anyway, and teams lose days transforming data that was already annotated and validated.

Generative AI Lab handles naming at import with no-code label mapping: case-insensitive matches like "diagnosis" to "Diagnosis" resolve automatically, while more complex variations like "Medication_Name" to "Medication" surface as suggestions for human review. Administrators approve the final mappings, every mapping is auditable, and the source annotation files are never modified.

The hard part is reusing annotations. Import-time mapping is what makes it practical.

https://hubs.li/Q04sJDq90

08/14/2026

Your EHR contains more data than your analytics can see.

The gap is in how the coding layer reads clinical language. A system built on keyword matching recognizes one spelling, one abbreviation, one phrasing. "HTN" and "hypertension" describe the same condition. "T2DM" and "E11" point to the same ICD-10 code. "MI" and "myocardial infarction" map to the same SNOMED concept.

Keyword-dependent systems fail on all of these - silently. No error message. No flag. No alert. The code does not get assigned. The record disappears from the denominator. The cohort shrinks. Nobody finds out until a downstream audit, if they find out at all.

The Snow Labs Medical Terminology Server was built to close this gap. 16+ medical vocabularies. Semantic search, spelling correction, abbreviation handling. On-premise. Deterministic: the same input always returns the same code.

2-minute explainer on why traditional approaches break down (link in comments).
https://hubs.li/Q04sJGkJ0

When a clinician types "GERD" and the system returns nothing becaus...