How It Works
A classic scan-to-confirm workflow for faster head lice decisions
Head Lice Checker is built as a decision engine, not a black box. You capture better evidence, review AI signals with confidence context, then move to clinic follow-up only when the confirmation gate indicates escalation.
At-a-glance outcomes
- Evidence-first AI analysis in under a minute
- Confidence tiers with visible detection overlays
- Clinic finder appears after confirmation gating
The same non-diagnostic safety boundaries apply across scan results, guidance content, and clinic escalation pathways.
The 6-step AI analysis sequence
This page follows a standard SaaS process model: input quality, processing transparency, confidence framing, and controlled escalation.
Step 1
Capture a quality scalp photo
Part hair near the scalp, use strong direct light, and take close shots from multiple angles. This reduces false uncertainty before analysis begins.
Output: an image suitable for high-confidence screening.
Step 2
Run AI analysis
The scan evaluates visual signals tied to lice-related indicators and returns candidate detections with class labels, confidence, and box coordinates.
Output: structured detection payload for interpretation.
Step 3
Normalize and validate detections
Detection aliases are mapped to stable labels and malformed entries are filtered. This protects consistency when provider payload formats vary.
Output: clean, render-safe detection set.
Step 4
Generate confidence-led summary
Results are converted into practical tiers so families can understand likely signal strength without overconfident medical language.
Output: headline label, confidence tier, and indicator count.
Step 5
Apply confirmation gate
Users review evidence overlays and symptom context before escalation. Low-confidence outcomes prompt recapture guidance instead of immediate clinic action.
Output: clear decision path to monitor, rescan, or escalate.
Step 6
Open clinic follow-up when needed
When indicators repeat or confidence is elevated, users move to the clinic finder and submit structured enquiry details for professional confirmation.
Output: faster, better-context clinic handoff.
Inside the AI process
The workflow is intentionally auditable in plain language so families and partners understand what the system is doing before any escalation decision is made.
Image intake and preprocessing
Images are validated for format and minimum usable dimensions before inference. We avoid aggressive enhancement transforms and instead emphasize better source capture for reliable interpretation.
Parsing and normalization
Provider outputs can include nested predictions and alias labels. We normalize these payloads into a stable internal structure so overlays, summaries, and logic remain consistent across releases.
Confidence and fallback policy
Confidence tiers communicate signal strength in plain language. Where evidence is weak, guidance shifts to recapture quality controls and symptom-aware follow-up rather than binary certainty.
Safety boundaries by design
The product is triage support, not diagnosis. Messaging, result framing, and CTA sequencing explicitly direct users to clinical confirmation when repeat indicators or persistent symptoms remain.
Confirmation gate before clinic follow-up
Clinic finder appears after confirmation logic, not as step one. This protects users from over-escalation and keeps professional follow-up focused on stronger or persistent risk cases.
Expanded FAQ
Run scan first, then escalate with confidence
Start with an evidence-quality scan. If the confirmation gate points to escalation, move directly to clinic follow-up in a structured flow.
This tool provides an indicative AI screening result only and is not a medical diagnosis.