MediLabel, a healthcare startup, needs to label sensitive X-ray images with a private group of radiologists and ensure label consolidation across multiple annotators (use consensus/aggregation). Which setup and configuration will meet their security and consolidation requirements when creating a SageMaker Ground Truth labeling job?

This option correctly addresses both security and consolidation. A SageMaker Ground Truth private workforce (work team) ensures only authorized radiologists access sensitive X-ray images, meeting the security requirement. Setting numberOfHumanWorkersPerDataObject to 3 provides multiple annotations per image, and enabling automated annotation consolidation (Ground Truth's built-in algorithm) efficiently combines these into a single, reliable label. The incorrect options fail on either security or consolidation: Using the public Mechanical Turk workforce is unsuitable for sensitive data, even with pre-signed URLs, due to the broad access. Relying on an external vendor or offline consolidation is unnecessary, as Ground Truth supports consolidation for private workforces. Using a public workforce with a 'private' flag on S3 objects is a misconfiguration for sensitive data, and a single annotator per object with manual QA is inefficient and doesn't leverage Ground Truth's consolidation features.