Perio.AI — Implementation Playbook (DSO)

Executive Summary

Perio.AI is an AI-powered diagnostic imaging platform that automatically detects periodontal disease markers (bone loss, pocket depth indicators, inflammation signatures) from panoramic and periapical radiographs, delivering clinical-grade analysis in 60 seconds. For DSOs managing 15-200+ locations, this creates standardized diagnostic workflows and reduces radiologist dependency while improving case acceptance through confident clinical documentation.

DSOs benefit specifically because: (1) they operate multiple locations with variable diagnostic quality standards, (2) they employ hygienists who need decision-support tools, and (3) they can amortize platform costs across high-volume imaging. Early adopters report 23-31% improvement in perio case acceptance and 40% reduction in radiologist referral costs within 90 days.

Deployment timeline: 14 weeks from kickoff to full enterprise deployment across a 20-50 location DSO, with clinical-grade results achievable by Week 8.

Pre-Implementation Checklist (Weeks 1-2)

Technical Requirements

Infrastructure audit:

• Verify PACS integration capability (Dexis, Schick, Planmeca, Carestream compatibility required)
• Confirm minimum internet bandwidth: 10 Mbps upstream per location
• Audit workstation specs: GPU acceleration optional but recommended for high-volume practices (>50 scans/day)
• Ensure HIPAA-compliant data pathways and review BAA requirements

System compatibility matrix: Document which imaging systems exist at each location. Perio.AI processes DICOM files; if any location uses proprietary formats, request export configuration 4 weeks pre-launch.

Admin access: Designate one IT contact per location with PACS administrator privileges. Perio.AI requires read-only PACS access and directory credentials.

Stakeholder Alignment

Executive sponsor identification:

• Clinical director: owns diagnostic standards
• DSO operations lead: owns deployment timeline and location sequencing
• Finance stakeholder: owns ROI tracking and budget allocation

Clinician buy-in session (2 hours):

• Present clinical validation data: sensitivity 91-94% for moderate+ bone loss, specificity 87-89%
• Show side-by-side comparison videos (Perio.AI output vs. radiologist reports) from 3-5 case studies
• Address the "replacement anxiety" head-on: position as hygienist-enabler and radiologist-offloader, not replacement
• Establish that final diagnosis remains clinician responsibility; AI provides confidence boost

Hygienist and front-desk training schedule: Lock in 3-hour onboarding time per location before Week 3.

Baseline Metrics Capture (Critical)

Establish enterprise baseline before any deployment:

Capture 4 weeks of baseline across pilot locations before Perio.AI activation. This enables statistically valid before/after comparison.

Pilot Wave (Weeks 3-6)

Location Selection Criteria

Choose 2-4 pilot locations balancing these factors:

1. Clinical readiness: Practices where hygienists actively diagnose perio (not fully delegated to radiologists)
2. Tech maturity: Locations with stable PACS, reliable IT support, and broadband
3. Volume: 150-300 intraoral radiographs/month (enough signal to measure, low enough to absorb change friction)
4. Leadership: Practice manager and clinical director express enthusiasm (not ambivalence)
5. Diversity: Represent mix of practice sizes and geographic regions within DSO footprint

Avoid: Practices in active transitions (new software, recent staff turnover) or locations with known technology adoption resistance.

Configuration and Setup

Week 3 activity:

1. PACS integration & testing (IT lead + Perio.AI support, 4 hours)

   • Configure read-only PACS access
   • Run 10-15 test scans through pipeline
   • Validate output appears in clinician-facing dashboard within 90 seconds
   • Confirm no data loss or formatting errors

2. Workflow embedding (practice clinical lead + Perio.AI specialist, 3 hours)

   • Map current diagnostic workflow: hygienist takes scan → awaits radiologist report (or diagnoses independently)
   • Insert Perio.AI decision point: after scan captured, before clinical note entry
   • Define action triggers: If bone loss >3mm detected, hygienist escalates to dentist for case discussion
   • Document in practice-specific SOP

3. Dashboard customization

   • Set up user accounts for 5-7 primary users per location
   • Configure notification routing (alerts for high-confidence positive findings)
   • Brand logos and practice names in reports

Week 4-5: Staged activation

• Week 4: Soft launch on 20% of incoming scans (diagnostic mode only; don't alter clinical decisions yet)
• Week 5: Increase to 100% of incoming scans; integrate into clinical workflow

Training Approach

Three-tier training structure:

Tier 1 — Hygienists (3-hour session, recorded):

• Module 1 (45 min): Platform UI, accessing results, interpretation guide
• Module 2 (60 min): Clinical module — what bone loss percentages mean, false positive/negative scenarios
• Module 3 (30 min): Case study review — 5 mixed cases with clinician commentary
• Module 4 (45 min): Hands-on: each hygienist processes 3 live scans under supervision

Tier 2 — Dentists/Associates (2-hour session):

• Focus on report interpretation, confidence scoring, exception handling
• Review 10 challenging cases showing where AI confidence varies
• Establish escalation protocol (when to order additional views, when to refer)

Tier 3 — Front desk/scheduling (30 min):

• How to mention "advanced imaging analysis" during consults
• No clinical discussion; messaging only

Certification requirement: Each clinician completes 10 test cases with >85% agreement to clinician-provided ground truth before live deployment.

Scaled Rollout (Weeks 7-16)

Wave Planning

Wave 2 (Weeks 7-9): 4-6 locations

• Select mix of high and moderate performers from baseline
• Deployment parallel, not sequential (minimize total timeline)

Wave 3 (Weeks 10-12): 8-12 locations

• Confidence high; deployment accelerates to 2-week per-location cycles
• Peer-training model activates: pilot location champions train Wave 3

Wave 4 (Weeks 13-16): Remaining locations

• Remote training standardized; IT setup fully templated
• 1-week deployment cycles possible

Change Management

Clinician resistance is the primary risk. Mitigate via:

1. Monthly clinical consensus calls: Present aggregate de-identified cases showing where Perio.AI improves diagnostic consistency across locations
2. Radiologist integration model: For DSOs with in-house radiologists, position Perio.AI as "first-pass QA" — radiologists review high-confidence positives, reducing administrative overhead
3. Peer pressure (positive): Share per-location acceptance rate improvements publicly in DSO town halls
4. Financial transparency: Publish direct ROI impact (radiologist cost savings + incremental perio revenue) by location

Support Infrastructure

Establish tiered support model:

• Tier 1 (Location level): Designated power user (usually hygienist/office manager) owns day-to-day troubleshooting
• Tier 2 (DSO central): Regional IT + one clinical champion handles escalations; 24-hour response SLA
• Tier 3 (Perio.AI): Integration issues, model updates, edge cases; 48-hour SLA