Shevs Connect | Full-Width Navigation
Mon – Fri: 08:00am – 05:00pm Magen Plaza, Juja – Along Thika Road, next to Juja Flyover. info@shevsconnectinstitute.com +254 780 253160

Medical Devices (SaMD — Software as a Medical Device)

SCILabel | Medical Imaging & Radiology AI

Medical Devices (SaMD — Software as a Medical Device)

Meeting the data quality bar that regulators require for AI-powered medical devices

Industry Challenge | SCILabel

Industry Challenge

AI-embedded medical devices — from ECG analysis software to surgical planning tools and wearable monitoring algorithms — are regulated as Software as a Medical Device (SaMD) by the FDA, CE Mark authorities, and other national bodies. These regulations impose strict requirements on training data quality, annotation provenance, and model validation. SaMD developers need a trusted data annotation partner that understands both clinical requirements and regulatory obligations.

Radiology AI Challenge
How SCILabel Serves This Industry | Radiology AI

How SCILabel Serves This Industry

Data Collection

We source physiological signal datasets (ECG, EEG, PPG), wearable sensor streams, surgical video recordings, and medical device output logs from clinical research partners and device manufacturers. All data is handled under Business Associate Agreements where required and with regulatory-aligned chain-of-custody documentation.

Data Annotation & Labeling

Our specialist workforce annotates ECG arrhythmia events (ANSI/AAMI EC57-aligned), epileptiform discharge detection on EEG, surgical instrument tracking on surgical video, fall detection events on wearable sensor data, and glucose trend classification on CGM outputs. All annotation is performed to client-specified annotation guidelines and documented for regulatory submission.

Data & Model Evaluation

Evaluators benchmark SaMD AI performance using FDA-aligned validation frameworks — sensitivity, specificity, PPV/NPV on pre-defined test sets. We provide model card templates, bias analyses across demographic subgroups, and documentation packages suitable for regulatory submissions.

Annotation Types & Formats

  • ECG arrhythmia event annotation (ANSI/AAMI EC57 types)
  • EEG epileptiform discharge detection and classification
  • Surgical video instrument detection and tracking
  • Wearable sensor event annotation (fall, activity, sleep stage)
  • CGM glucose trend classification and event labeling
  • Regulatory-documentation-ready annotation provenance records

Specialist Workforce Tracks

Track 2 (Medical Imaging) and Track 5 (AI Safety & Regulatory Compliance): Biomedical Engineers, Clinical Officers, Biomedical Scientists, Regulatory Affairs Professionals.

Example Deliverable | SCILabel

Example Deliverable

Client Output Example
A 10,000-recording annotated ECG dataset with arrhythmia event labels to ANSI/AAMI EC57 taxonomy, annotator qualification records, inter‑annotator agreement report, and a regulatory submission package template.