Medical systems require highly interactive user interfaces where improving the user experience for optimal usability is crucial. This work presents a design case study on digital histopathology annotation user interface. The case study is based on the identified usability flaws with the widely used tool in the field - QuPath. Based on the findings, a novel prototyping framework to support the design process is proposed. The proposed prototyping framework was used to create a prototype of the histopathology annotation user interface with a focus on mitigating the identified usability flaws with QuPath. The case study uncovers the limitations of rapid high-fidelity prototyping for interactive interface design. To overcome the limitations, an approach supporting realistic interaction with dynamic canvas elements within a high-fidelity user interface prototype is proposed. The main contribution of this study is a hybrid high-fidelity prototyping tool that integrates static user interface components with an interactive JavaScript-based canvas. The prototype was evaluated and refined based on the obtained user feedback. Beyond the specific use case presented in the study, the proposed prototyping framework demonstrates a flexible approach to prototype design ideas of canvas-based applications in an interactive form, including the possibility to prototype AI-assisted workflows.