Design Thinking Methods

Design thinking methods are a set of structured, creative techniques used to move from understanding user problems to generating and refining potential solutions. In the Define phases of co-design, methods such as How Might We questions, fishbone diagrams, and design sketches help teams break down complex challenges, spark diverse ideas, and translate user insights into concrete design concepts for further prototyping.

Application Example

​

In a study aiming to design “Aliado,” an AI-based decision support system for oncological liver surgery, researchers applied a series of design thinking methods to move from problem identification to concept development. The design process began with background research on machine learning, AI, liver malignancies, and oncological liver surgery, followed by observations along the entire treatment path and interviews with surgeons, radiologists, pathologists, and AI experts to uncover pain points and unmet needs.

​

Based on these insights, the team used several design thinking tools to structure and generate ideas:

• How Might We (HMW) questions - to reframe identified challenges into actionable design opportunities.

• Cause–effect (Ishikawa/Fishbone) diagrams - to break down complex problems into root causes and subproblems.

• Insight cards - to capture each problem or insight (with supporting evidence on the front) and propose solution strategies (on the back).

• 1-Minute Ideas - to rapidly generate creative implementation ideas for each insight card.

• Crazy 8 - to quickly sketch eight variations of ideas in eight minutes, expanding the design space.

• Design sketches - to consolidate promising ideas into visual concepts.

​​

These design sketches were then discussed with surgical residents and senior physicians during co-creation sessions, where feedback was incorporated through several rounds of iteration. The final concept was visualised in Figma using an artificial patient case to illustrate how the system could work in practice. (Schulze et al., 2024)