Skip to main content
Sleep Medicine Knowledge Graph | DEEPdormir.ai
Skip to knowledge graph
The Foundation of Our AI

The Specialized Knowledge Graph for Sleep Medicine

Discover the interconnected web of sleep medicine intelligence that powers AVA and our platform. Built upon evidence-based guidelines, peer-reviewed literature, and clinical protocols, our Knowledge Graph transforms static medical texts into an active reasoning engine.

AASM Guidelines
Peer-Reviewed Literature
Clinical Pathways
HIPAA Compliant
Explore the Graph Meet AVA
AVA
Disorders
ApneaInsomniaNarcolepsy
Treatments
CPAPCBT-IOral Appliance
Diagnostics
PSGHSATBiomarkers
Research
LiteratureGuidelinesEvidence

Evidence-Based Intelligence

DEEPdormir.ai is grounded in established medical science, ensuring our AI provides clinically sound support.

Clinical Foundations

Our knowledge base architecture is designed alongside board-certified sleep medicine physicians, ensuring that the AI maps relationships between symptoms, disorders, and treatments the way a trained clinician would.

Research Integration

Built upon an extensive ingestion of peer-reviewed research papers, clinical studies, and established diagnostic guidelines (including AASM criteria).

AASM Criteria Peer-Reviewed Journals Clinical Protocols

Practice-Specific Tuning

While the base graph provides standard medical knowledge, the system allows for the ingestion of your clinic's specific Standard Operating Procedures (SOPs).

AVA: Where Knowledge Becomes Intelligence

Our Clinical AI Assistant, AVA, utilizes this knowledge graph to understand context, synthesize medical information, and provide structured insights.

  • Clinical Comprehension

    Understands the complex relationships between comorbidities and sleep disorders.

  • Contextual Reasoning

    Applies knowledge contextually based on the patient history provided.

  • Transparent Citations

    Links recommendations back to the underlying clinical guidelines.

Discover AVA's Capabilities
Clinical Query AVA Assistant
What are the primary indicators for Hypoglossal Nerve Stimulation per current guidelines?

Based on established guidelines for Hypoglossal Nerve Stimulation (HGNS), primary indications include:

  • Moderate to severe OSA (AHI typically 15-65)
  • Documented failure or intolerance of CPAP therapy
  • Absence of complete concentric collapse at the soft palate

Source: Patient Selection Criteria for HGNS.

Explore the Knowledge Base

A structured mapping of sleep medicine concepts allowing our AI to understand clinical context and draw evidence-based conclusions.

Obstructive Sleep Apnea (OSA)

Comprehensive mapping of pathophysiology, comorbidities (e.g., hypertension), diagnostic criteria, and treatment pathways.

CPAPSurgeryOAT

Chronic Insomnia

Evidence-based approaches mapping cognitive behavioral interventions (CBT-I) and pharmacological guidelines.

CBT-IHygieneMeds

Narcolepsy & Hypersomnia

Understanding central disorders of hypersomnolence, diagnostic MSLT criteria, and wake-promoting therapies.

PAP Therapy Protocols

Detailed logic on initiation, titration guidelines, interface selection, and compliance troubleshooting.

TitrationCompliance

Oral Appliance Therapy (OAT)

Patient selection criteria, bite registration guidelines, and titration protocols for dental sleep medicine.

Diagnostic Testing (PSG/HST)

AASM scoring criteria for respiratory events, sleep staging logic, and appropriate testing selection guidelines.

AASM ScoringHSAT

Knowledge Architecture

Interconnected Data

Concepts are linked (e.g., tying OSA to specific cardiovascular risks) allowing the AI to reason contextually.

Continuously Updated

Our models are retrained as new clinical guidelines and landmark studies are published.

Isolated Processing

Your patient data is evaluated against the graph, but never used to train public models.

Powering the Entire Ecosystem

This central intelligence layer ensures clinical consistency across every tool in our platform.

1

Ingestion

Guidelines and SOPs are securely ingested into the vector database.

2

Structuring

NLP maps relationships between conditions and protocols.

3

Application

Knowledge is actively utilized by agents and workspaces.

AVA Clinical Assistant

Core

Uses the knowledge base for diagnostic support, treatment planning, and generating clinical documentation.

AI Receptionist

Front Desk

Leverages basic clinical knowledge to answer patient FAQs accurately regarding studies and equipment.

Custom Agents

Targeted

Specialized bots built on subsets of the knowledge graph (e.g., an agent trained purely on CPAP compliance rules).

Marketing Suite

Growth

Draws on factual clinical data to generate accurate, compliant educational content for patients.

Frequently Asked Questions

Generic AI models are trained on the open internet, leading to hallucinations and inaccurate medical advice. Our models are grounded in a private, specialized knowledge graph built strictly on established clinical guidelines, ensuring safe and accurate reasoning.

No. We employ strict data isolation. If you upload your practice SOPs to customize your workspace, that data remains in a private tenant. Patient Health Information (PHI) is never used to train base models.

The system is continuously updated as new guidelines (such as AASM updates) and landmark peer-reviewed studies are published and verified by our clinical team.

Deploy Intelligence in Your Practice

See how our clinically-grounded AI tools can streamline your workflows and support better patient outcomes.

Evidence-Based HIPAA Compliant Practice Specific
Request a Platform Demo Or call us at (888) 555-0199