RadStream: An interactive visual display of radiology workflow for delay detection in the clinical imaging process

Abstract

Given the high caseload most radiology departments face on a daily basis, workflow optimization becomes a necessity to avoid delays and poor health outcomes. This requires detailed analysis of workflow data to identify problem areas in the process. Analysis of the clinical imaging process demands an understanding of temporal intervals and temporal event sequences and relationships. Working with radiology staff, we seek to provide a tool to help monitor and improve radiology department workflow in order to increase efficiency and productivity and ensure the delivery of timely clinical imaging reports. In this thesis, I present RadStream: a web-based retrospective, exploratory, interactive data visualization tool that provides a comprehensive overview of the radiology department's daily activities. I worked closely with radiology staff to analyze the department workflow and classify the analytical tasks required by domain experts in order to inform the design of the tool. Together, we abstracted the steps involved in the clinical imaging process. We also identified factors affecting the workflow (such as personnel, machine availability, and resources) and noted the relationship between the different factors as it plays an important role in increasing productivity. RadStream depicts the steps involved in the process of clinical imaging and shows the flow of processes from one step to the next. The visual representation emphasizes the time intervals between the different steps and uses colour coding to denote the status of a process (on time, acceptably late, late) in compliance with standard radiology turnaround times (TATs). The main focus of RadStream is on monitoring performance with special attention to duration, delays, and compliance with standard TATs. RadStream was evaluated by radiology staff using hospital data and real scenarios to evaluate its effectiveness, efficiency, and usability. The initial feedback received was very promising. And based on results collected from the evaluation studies, I sensed a general acceptance and excitement about the system as a quality assurance tool. I have also collected some constructive feedback to build upon for future releases. Finally, I reflect on lessons l learned from iteratively designing RadStream, and present design guidelines for the design of visual analytics tools for health care.

EHR STAR: The State‐Of‐the‐Art in Interactive EHR Visualization

“…The size of an electronic healthcare dataset is often huge. The rate of data growth exceeds the capacity of algorithms and software developed to visualize it [AKDA15].…”

“…High-dimensionality: Closely related to heterogeneity, healthcare datasets are high-dimensional and complex [GS14,TRL*17]. The ability to visualize large datasets with many attributes effectively remains a challenging problem [AKDA15].…”