Technology Clinical Data Investors Team Get Access
The Physics

Certainty of Sight.

Polarization-Sensitive Optical Coherence Tomography (PS-OCT) measures the birefringence of cardiac tissue. It sees what others can't.

How It Works

Light vs. Electricity.

Standard mapping systems rely on electrical impedance, which is an indirect surrogate. AblaView® uses a 1310nm swept-source laser to directly image tissue structure. By analyzing the polarization state of backscattered light, we determine structural integrity (birefringence) in real-time.

Resolution

10µm

Axial resolution. 100x higher than ultrasound.

The Only Complete Solution.

Real-time durability assessment across all energy modalities.

Product Features AblaView® J&J MedTech Medtronic Boston Sci Abbott
Energy Delivery RF + PFA RF / PFA PFA PFA PFA
Real-time Visualization Yes — — — —
Lesion Depth Prediction Yes — — — —
Lesion Durability (PFA) Yes — — — —
Gap Detection Yes — — — —
Contact Stability

50µm Resolution.

The only technology able to quantify contact with 50µm resolution and evaluate lateral/vertical displacement. Superior to force sensors and impedance.

Pop Risk

Steam Pop Indicator.

AblaView® is the only company to have developed and validated a model for Pop risk, preventing dangerous complications.

PFA Assessment

Immune to Stunning.

Our model identifies ablation efficacy independent of electrical signals, which are biased by cellular stunning below the electroporation threshold.

Gap Detection

Optical Precision.

Detects preserved tissue between lesions (precursor to reconnections) independent of electrical signals.

The Three Impossibility Walls.

Why competitors are a decade behind.

Wall 1: Hardware

12 Years. €68M+.

To develop the optical multiplexer and catheter integration. We own the only working system.

Wall 2: Data

25TB Clinical Dataset.

15 years of PS-OCR signals paired with histological ground truth. No competing system holds comparable data.

Wall 3: IP

70+ Patents.

9 jurisdictions. Owned outright. No licensing obligations to any third party.

Technical Specifications

Imaging Modality PS-OCR (Polarization-Sensitive Optical Coherence Reflectometry)
Laser Source 1310nm Swept-Source
Scan Rate 50 kHz A-Scan
Parallel Channels 7 PS-OCR Measurements
Update Rate 20 Frames/Second Real-time
Catheter Diameter 7 Fr
RF Generator Up to 50W, 140V max, 500kHz
PFA Generator Up to 2000V @ 22 Amps
AI Processing NVIDIA IGX Orin (275 TOPS)
Latency < 50ms (Photon to Prediction)
Clinical AI Engine, SaMD Datasheet

Transparent by design.

The Clinical AI Engine is Software as a Medical Device under IMDRF and FDA digital-health frameworks. Its intended use, module architecture, performance, and change-control plan are documented, testable, and auditable.

Intended Use and Classification

SaMD Risk Category (IMDRF) Category III. Informs treatment decisions in a critical-care context.
FDA Classification and Pathway Class II medical device software. 510(k) pathway with PCCP. Predicate analysis complete.
EU MDR Classification Class IIb per Rule 11. Software providing information for therapeutic decisions.
Intended User Board-certified cardiac electrophysiologist in an EP-lab environment.
Intended Use Real-time tissue characterization, contact verification, and lesion-quality indicators during catheter ablation for atrial fibrillation.
Clinical Role Advisory. Decisions remain with the treating physician. Outputs are not autonomous actions.
Software Lifecycle IEC 62304 Software Safety Class C. 21 CFR Part 11 compliant.

Module Architecture

M1. Tissue Classification and Gap Detection Identifies tissue type at the catheter tip from the optical return signal and detects preserved tissue between lesions. Trained on the 25TB proprietary dataset.
M2. Contact Assessment Quantifies catheter-to-tissue contact quality in real time from the optical signal, replacing electrophysiological estimation.
M3. Lesion Prediction Detects birefringence loss as the direct optical marker of irreversible tissue change. Predicts durable lesion formation live during PFA and lesion depth during RFA.
M4. Safety Detection Independent CPU-based safety boundary layer. Runs in parallel with the AI stack and is not dependent on it.
Runtime Stack TensorRT FP16 on NVIDIA IGX Orin. Deterministic inference. Sub-50ms end-to-end.

Training Data and Validation

Proprietary Dataset 25TB+ accumulated across 15 years of R&D and clinical work. Bench, preclinical, and first-in-human data combined.
Optical Signal Library Used to train all four module layers. Paired with birefringence calibration across tissue types and energy levels.
Ground Truth Whole-mount histology, TTC staining, blinded pathologist adjudication for preclinical work. Chronic remap confirmation for durability endpoints.
FIH Validation Layer 1 (optical hardware) validated in first-in-human study. Published Europace, February 2025. 100% specificity for durable PFA lesion prediction at three-month follow-up.
Replication Cost Estimate A competitor starting from zero would need 10+ years and €60M+ minimum to build a comparable dataset. None is known to exist today.

Governance and Change Control

GMLP Alignment FDA Good Machine Learning Practice (2021). Ten guiding principles, full mapping.
Predetermined Change Control Plan PCCP drafted per FDA 2024 guidance. Retraining scope, performance bounds, and protocol pre-specified.
Continuous Monitoring Deployed-model drift detection. Anonymized performance telemetry, opt-in, GDPR and HIPAA compliant.
Human-Factors File IEC 62366-1. Formative complete. Summative planned Q3 2026.
Cybersecurity Posture FDA 2023 premarket guidance and EU MDR Article 15(5). SBOM, threat model, signed firmware, vulnerability disclosure policy.
Explainability Saliency overlays and birefringence heatmaps exposed in the clinician UI for every inference.
Request the AI/ML Regulatory Dossier (NDA)
Portfolio Fit & Integration

Built to ship inside your stack.

AblaView® is architected to co-exist with the dominant mapping and ablation platforms. Catheter-only, co-deployed, or SDK-integrated. The hardware, optical pipeline, and Clinical AI Engine expose defined interfaces at each level.

Partner System Interface Integration Mode Status
J&J CARTO 3 (Biosense Webster) Published SDK • signal-level co-registration Catheter-only or co-deployed console Prototype
Abbott EnSite X EnSite Connect API • tracked-tool profile Catheter + optical overlay on EnSite display Roadmap,H2 2026
Medtronic Affera Third-party mapping bridge Co-registration of AI Engine outputs onto Affera map Roadmap,2027
Boston Sci OPAL HDx / Rhythmia IntelliMap bridge • FDM signal export Catheter + overlay; Farapulse-compatible console mode Roadmap,2027
Generic EP recording systems HL7 + DICOM-encapsulated PDF/SR Procedure report export to EHR Available
Deployment Mode A

Catheter-Only

AblaView® catheter used with the partner's existing mapping/generator stack. AI Engine outputs displayed on AblaView® console; procedural workflow unchanged.

Deployment Mode B

Co-Deployed Console

AblaView® console sits alongside the partner's mapping system, time-synchronized and spatially co-registered via published SDK. Side-by-side clinical view.

Deployment Mode C

SDK / OEM Integration

AI Engine outputs streamed directly into the partner's UI as a licensed overlay. Deepest integration; requires bilateral engineering and regulatory alignment.

For Strategic Partners

Medical-device partnerships are long-cycle. Partnership conversations benefit from early architectural alignment. We publish interface contracts under mutual NDA so engineering teams can validate fit before terms.

Request Partnership Briefing (NDA)