feat(core): add ECGViewport for DICOM waveform rendering

[TFRadicalImaging]

Add a new ECG viewport type that renders 12-lead ECG traces from DICOM Waveform data using a canvas-based custom rendering pipeline (same pattern as VideoViewport and WSIViewport).

• ECGViewport class with grid, trace, and label rendering
• Pan (left-drag) and Zoom (right-drag) tool support
• Per-channel trace visibility toggle
• Supports InlineBinary, BulkDataURI, and retrieveBulkData data sources
• Typed WaveformSequenceInput interface for the setEcg() API
• Interactive example with 12-lead ECG demo data

Context

Adds a new ECG viewport type to Cornerstone3D for rendering 12-lead ECG traces from DICOM Waveform data (SOP Class 1.2.840.10008.5.1.4.1.1.9.1.1). This follows the same custom rendering pipeline pattern used by VideoViewport and WSIViewport — canvas 2D rendering, no VTK.js dependency.

The viewport accepts parsed DICOM WaveformSequence metadata via setEcg() and handles data loading from InlineBinary, BulkDataURI, or retrieveBulkData sources.

Changes & Results

New files:

• packages/core/src/RenderingEngine/ECGViewport.ts — Main viewport class (~940 lines) with grid, trace, and label rendering
• packages/core/src/types/ECGViewportTypes.ts — Type definitions (ECGChannel, ECGWaveformData, WaveformSequenceInput, WaveformDataSource)
• packages/core/src/types/ECGViewportProperties.ts — Viewport properties extending ViewportProperties
• packages/core/src/types/IECGViewport.ts — Type alias for the viewport
• packages/core/examples/ecg/index.ts — Interactive example with 12-lead ECG demo data

Modified files:

• packages/core/src/enums/ViewportType.ts — Added ECG = 'ecg'
• packages/core/src/RenderingEngine/helpers/viewportTypeToViewportClass.ts — Registered ECGViewport
• packages/core/src/index.ts — Exported ECGViewport
• packages/core/src/types/index.ts — Exported new ECG types

Features:

• ECG paper grid rendering (major/minor lines, adaptive spacing at different zoom levels)
• White trace polylines on black background with yellow channel labels
• Pan (left-drag) and Zoom (right-drag) via standard Cornerstone tools
• Per-channel trace visibility toggle
• Dynamic channel scaling to fit canvas aspect ratio
• Viewport auto-resizes to match ECG content dimensions
• Channel min/max cached at data load time for efficient rendering

Testing

1. Run yarn example ecg
2. Verify 12-lead ECG traces render with grid background
3. Left-drag to pan, right-drag to zoom
4. Click "Reset View" to reset camera
5. Toggle individual lead checkboxes to show/hide traces
6. Verify traces rescale when toggling channels on/off

Checklist

PR

• My Pull Request title is descriptive, accurate and follows the
  semantic-release format and guidelines.

Code

• My code has been well-documented (function documentation, inline comments,
  etc.)

Tested Environment

• OS: macOS 25.1.0
• Node version: >=20
• Browser: Versão 144.0.7559.134 (Oficial Version) (arm64)