Acute stroke demands prompt diagnosis and treatment to achieve optimal patient outcomes. However, the intricate and irregular nature of clinical data associated with acute stroke, particularly blood pressure (BP) measurements, presents substantial obstacles to effective visual analytics and decision-making. Through a year-long collaboration with experienced neurologists, we developed PhenoFlow, a visual analytics system that leverages the collaboration between human and Large Language Models (LLMs) to analyze the extensive and complex data of acute ischemic stroke patients. PhenoFlow pioneers an innovative workflow, where the LLM serves as a data wrangler while neurologists explore and supervise the output using visualizations and natural language interactions. This approach enables neurologists to focus more on decision-making with reduced cognitive load. To protect sensitive patient information, PhenoFlow only utilizes metadata to make inferences and synthesize executable codes, without accessing raw patient data. This ensures that the results are both reproducible and interpretable while maintaining patient privacy. The system incorporates a slice-and-wrap design that employs temporal folding to create an overlaid circular visualization. Combined with a linear bar graph, this design aids in exploring meaningful patterns within irregularly measured BP data. Through case studies, PhenoFlow has demonstrated its capability to support iterative analysis of extensive clinical datasets, reducing cognitive load and enabling neurologists to make well-informed decisions. Grounded in long-term collaboration with domain experts, our research demonstrates the potential of utilizing LLMs to tackle current challenges in data-driven clinical decision-making for acute ischemic stroke patients.

For irregularly measured time-series data, the measurement frequency or interval is as crucial information as measurements are. A well-known time-series visualization such as the line graph is good at showing an overall temporal pattern of change; however, it is not so effective in revealing the measurement frequency/interval while likely giving illusory confidence in values between measurements. In contrast, the bar graph is more effective in showing the frequency/interval, but less effective in showing an overall pattern than the line graph. We integrate the line graph and bar graph in a unified visualization model, called a ripple graph, to take the benefits of both of them with enhanced graphical integrity. Based on the ripple graph, we implemented an interactive time-series data visualization tool, called Stroscope, which facilitates multi-scale visualizations by providing users with a graphical widget to interactively control the integrated visualization model. We evaluated the visualization model (i.e., the ripple graph) through a controlled user study and Stroscope through long-term case studies with neurologists exploring large blood pressure measurement data of stroke patients. Results from our evaluations demonstrate that the ripple graph outperforms existing time-series visualizations, and that Stroscope has the efficacy and potential as an effective visual analysis tool for (irregularly) measured time-series data.