A medical imaging team has many unlabeled CT scans in S3 and must build a labeling pipeline that only authorized users can access. Which set of steps will build the labeling pipeline with the least development effort?

The correct answer leverages Amazon SageMaker Ground Truth, which is designed for building labeling pipelines with minimal development effort. Creating a private workforce ensures only authorized users can access the labeling tasks. A manifest file points Ground Truth to the S3 objects (CT scans) to be labeled. The built-in bounding-box task type is appropriate for medical imaging where specific regions of interest need to be identified. Writing clear labeling instructions is crucial for consistent and accurate annotations. Incorrect options: Building a custom labeling tool on EC2 and using SQS requires significant development effort, which the question aims to minimize. Mechanical Turk is a public workforce and would not restrict access to authorized users, violating the requirement. Using Amazon Cognito and building a custom web app with Amplify and Lambda also involves substantial development effort, contrary to the goal of least effort.