Results for "software as medical device"

Software regulated as a medical device.

US approval process for medical AI devices.

AI applied to X-rays, CT, MRI, ultrasound, pathology slides.

Software pipeline converting raw sensor data into structured representations.

Compromising AI systems via libraries, models, or datasets.

Software simulating physical laws.

AI that ranks patients by urgency.

Training across many devices/silos without centralizing raw data; aggregates updates, not data.

Devices measuring physical quantities (vision, lidar, force, IMU, etc.).

Practices for operationalizing ML: versioning, CI/CD, monitoring, retraining, and reliable production management.

Central system to store model versions, metadata, approvals, and deployment state.

Models whose weights are publicly available.

Mechanism to disable AI system.

Governance of model changes.

Storing results to reduce compute.

Mathematical guarantees of system behavior.

Of predicted positives, the fraction that are truly positive; sensitive to false positives.

Harmonic mean of precision and recall; useful when balancing false positives/negatives matters.

When some classes are rare, requiring reweighting, resampling, or specialized metrics.

Probabilities do not reflect true correctness.

AI systems assisting clinicians with diagnosis or treatment decisions.

Automated assistance identifying disease indicators.

Ability to correctly detect disease.

Grouping patients by predicted outcomes.

Patient agreement to AI-assisted care.