AI Analysis

ChartInsight automatically generates multiple analysis outputs from your medical records, making it easy to understand and navigate complex medical documentation.

Narrative Summary

Overview

A plain-language prose summary of the entire medical record, synthesizing key findings, diagnoses, treatments, and outcomes.

Features

  • AI-generated prose that reads naturally
  • Markdown formatting for readability (headings, lists, emphasis)
  • Focus-aware – prioritizes content based on your focus instructions
  • Comprehensive coverage of the record's key medical events

Use Cases

  • Quick overview before detailed review
  • Executive summary for legal briefs
  • Patient history overview for consultations

Chronology View

Overview

A time-ordered index of all medical events, appointments, and procedures extracted from the record.

Entry Details

Each chronology entry includes:

  • Date of service
  • Service provider/facility name
  • Primary classification (e.g., "Office Visit", "Lab Results")
  • Secondary classification (if applicable)
  • Extract title – brief summary heading
  • Detailed extract – full description of the event
  • Page references – clickable links to source document pages

Search and Filtering

  • Full-text search across all entries
  • Date range filtering with start and end date pickers
  • "Relevant only" toggle to show/hide AI-determined non-relevant entries
  • Entry count showing filtered vs. total entries

Editing Capabilities

  • Inline editing – correct AI-extracted data directly:
    • Date of service
    • Service provider
    • Extract title
    • Extract text
  • Relevance controls – manually mark entries as relevant or not relevant
  • Revert capability – restore original AI-extracted values after corrections
  • Correction indicator – visual badge showing which entries have been modified

Navigation

  • Click any page reference to view that page in the PDF viewer
  • Page ranges displayed for multi-page events (e.g., "pg. 15-18")

Vitals Timeline

Overview

A tabular display of all vital signs extracted from the medical record, organized by date and provider.

Tracked Vital Types

VitalDescription
Blood PressureSystolic/diastolic measurements
Heart RateBeats per minute
Body TemperatureFahrenheit or Celsius
Respiratory RateBreaths per minute
Oxygen SaturationSpO2 percentage
Pain ScalePatient-reported pain level
HeightPatient height
WeightPatient weight
Body Mass IndexCalculated BMI

Features

  • Column visibility controls – show/hide specific vital types
  • Search across provider names and values
  • Date filtering with start and end date pickers
  • Provider attribution – see which provider recorded each reading
  • Auto-column selection – only shows vitals present in the record

Display

  • Responsive table layout
  • Chronological ordering by date
  • Clear column headers with vital type labels

Medications List

Overview

A consolidated inventory of all medications mentioned throughout the medical record.

Entry Details

Each medication entry includes:

  • Medication name
  • Date range – from/to dates showing prescription period
  • Page references – links to source document locations

Features

  • Search by medication name
  • Date filtering to find medications in a time period
  • Expandable page lists for medications mentioned on many pages
  • "Show all" toggle for long page lists

Use Cases

  • Medication reconciliation
  • Drug interaction review
  • Treatment timeline analysis

Custom Sections

Overview

Team-configurable extraction sections defined in templates, allowing specialized content extraction for your workflow.

Configuration

  • Defined in Templates
  • Customizable icons for visual identification
  • Section-specific extraction prompts

Features

  • Specialized content extraction based on section configuration
  • Displayed in sidebar alongside standard outputs
  • Per-template configuration for different record types

Examples

  • Surgical history
  • Radiology findings
  • Physical therapy notes
  • Expert opinions

Output Quality

AI Accuracy

  • Continuous improvement from user feedback
  • Classification models trained on medical terminology
  • Focus instructions improve relevance detection

User Controls

  • Edit extracted content to correct errors
  • Mark entries as relevant/not relevant
  • Revert changes to restore original extraction
  • Reprocess records after file changes