Designed around ECG, respiration, and PPG signals arriving from consumer or research wearables.
Concept Prototype
SignalScope
A remote wearable review console for teams monitoring ECG, respiration, and signal quality at scale. SignalScope is framed as a product concept for turning messy biosignal streams into a readable triage workspace with annotations, confidence cues, and fast escalation paths.
Concept note: this is a portfolio exploration designed to show product thinking and front-end direction, not a deployed clinical system.
Highlights degraded signal quality, repeated artifact patterns, and likely review priorities before a clinician opens the full record.
Pairs waveform review with quality markers and short-form notes so teams can move faster without losing context.
Review workflow
From noisy wearable streams to a clean, accountable review surface.
The core idea is a console that helps users decide whether a stream is clinically useful before they spend time interpreting it. Instead of throwing raw charts on a screen, SignalScope groups quality, severity, and context into one place.
- Patient rail prioritizes unresolved events and poor-signal segments first
- Waveform stack makes it easy to compare channels without jumping across views
- Inline quality confidence gives quick feedback on whether the segment is interpretable
- Annotation-first flow keeps the review loop compact for high-volume monitoring teams
SignalScope is presented as a product-direction concept: a realistic interface study for remote monitoring teams, rather than a shipped healthcare deployment.
Interaction Model
The interface is intentionally biased toward scan-then-act behavior: a reviewer should be able to identify signal quality, compare channels, and record a concise note in seconds. The layout therefore keeps priority states and waveform relationships visible at the same time.
Architecture Rationale
Streams arrive as packetized WebRTC data channels, decoded client-side into typed arrays for each signal modality. D3 handles axis scaling and annotation overlays while the raw waveform draws through a direct canvas path for consistent frame pacing. The queue state is event-sourced so multiple reviewers could share triage position without conflicts.
Clinical Signal Pipeline
Each incoming segment passes through a quality gate that scores lead contact, baseline drift, and motion artifact density. Segments below threshold are flagged before a reviewer even opens the record, so the queue naturally front-loads the cases that need human judgment rather than routine confirmation.
Next Step
A believable next iteration would add clip-level annotations, collaborative review comments, and timeline playback controls so the concept can evolve from static case-study material into a richer interactive prototype.