John Snow Labs is the De-facto Industry Leader for Medical Large Language Models.
State of the Art Medical Large Language Models
Clinical Note Summarization
is 30% more accurate than BART, Flan-T5 and Pegasus.
Clinical Entity Recognition
John Snow Labs’ models make half the errors that ChatGPT does.
Extracting ICD-10-CM Codes
is done with a 76% success rate versus 26% for GPT-3.5 and 36% for GPT-4.
Putting Healthcare LLMs to Production Use
Using Healthcare-Specific LLM’s for Data Discovery from Patient Notes & Stories
The US Department of Veterans Affairs, a health system which serves over 9 million veterans and their families. This collaboration with VA National Artificial Intelligence Institute (NAII), VA Innovations Unit (VAIU) and Office of Information Technology (OI&T) show that while out-of-the-box accuracy of current LLM’s on clinical notes is unacceptable, it can be significantly improved with pre-processing, for example by using John Snow Labs’ clinical text summarization models prior to feeding that as content to the LLM generative AI output.
Text-Prompted Patient Cohort Retrieval: Leveraging Healthcare LLM Models for Precision Population Health Management
Using John Snow Lab’s Healthcare LLM models, the ClosedLoop platform enables users to retrieve cohorts using free-text prompts. Examples include: “Which patients are in the top 5% of risk for an unplanned admission and have chronic kidney disease of stage 3 or higher?” or “Which patients are in the top 5% risk for an admission, older than 72, and have not undergone an annual wellness checkup?”
Lessons Learned Applying Large Language Models
This session shares currently deployed software, lessons learned, and best practices that John Snow Labs has learned while enabling academic medical centers, pharmaceuticals, and health IT companies to build LLM-based solutions. It covers what generative AI use cases are healthcare organizations deploying today, which solution architectures are used to deliver these use cases, and general-purpose LLMs perform versus healthcare-specific LLMs in these use cases.
Lots of companies make claims about healthcare-specific LLM’s. John Snow Labs are the only ones who publish reproducible accuracy benchmarks and have Medical LLM systems in production.
What’s in the Box
Entity Recognition
40 units
DOSAGE
of
insulin glargine
drug
at night
FREQUENCY
De-Identification
Algorithms
Information Extraction
- Document Classification
- Entity Disambiguation
- Contextual Parsing
- Patient Risk Scoring
Clinical Grammar
- Deep Sentence Detector
- Medical Spell Checking
- Medical Part of Speech
- Terminology Mapping
Entity Linking
Suspect diabetes
SNOMED-CT:
473127005
Lisinopril 10 MG
RxNorm:
316151
Hyponatremia
ICD-10:
E87.1
Question Answering
Algorithms
Data Obfuscation
- Name Consistency
- Gender Consistency
- Age Group Consistency
- Format Consistency
Zero-Shot Learning
- Entities by Prompt
- Relations by Prompt
- Classification by Prompt
- Relative Data Extraction
Assertion Status
Fever and sore throat
PRESENT
No stomach pain
ABSENT
Father with Alzheimer
FAMILY
Summarization
Content
Medical
Language Models
Language Models
BioGPTBioBERTJSL-BERTJSL-sBERTClinicalBERTGloVe-MedT5Flan-T5
Relation Extraction
Ora
NAME
a
25
AGE
yo
cashier
PROFESSION
from
Morocco
LOCATION
Data Enrichment
Content
Medical
Terminologies
Terminologies
SNOMED-CTCPTUMLSICD-10-CMRxNormHPOICD-10-PCSICD-OLOINC
2,000+ Pretrained Models
Clinical Text
Signs, Symptoms, Treatments, Findings, Procedures, Drugs, Tests, Labs, Vitals, Sections, Adverse Effects, Risk Factors, Anatomy, Social Determinants, Vaccines, Demographics, Sensitive Data
Biomedical Text
Clinical Trial Design, Protocols, Objectives, Results; Research Summary & Outcomes; Organs, Cell Lines, Organisms, Tissues, Genes, Variants, Expressions, Chemicals, Phenotypes, Proteins, Pathogens
Trainable & Tunable
Scalable to a Cluster
Fast Inference
Hardware Optimized
Community