| dc.description.abstract | Clinical work in modern hospitals involves extensive use of digital medical imaging, with many specialities becoming reliant on real-time access to detailed imaging. The widespread use of technology such as digital imaging in hospitals, and the difficulty of sterilising computer controls and peripherals has increased opportunities for the spread of pathogens. Increased contact with shared surfaces, such as keyboards, can potentially increase bio-contamination between users. In turn, this can result in increased rates of healthcare-acquired infections (HAI), leading to poorer patient outcomes.
Currently, while imaging provides healthcare professionals (HP) with essential insights into a patient's condition, the requirement for sterility in some situations means that users face issues such as being unable to directly manipulate key imaging in aseptic environments. Touchless interaction provides an attractive potential means for providing aseptic access to medical imaging, by reducing contact with shared surfaces without reducing HP ability to interact with imaging systems. As touchless technologies reach maturity, it is an opportune time to investigate their application to this problem, and in this thesis, I investigate the design of touchless systems for accessing medical imaging.
The thesis investigates the needs and requirements for these systems through a systematic search of the literature and through a qualitative interview study with a broad range of clinicians. Starting with the motivations for touchless control, asepsis provides a theme and motivation for the field as a whole, but other advantages, such as solving the hands-busy problem, are also proposed. Existing practices aim to maintain sterility where possible through covering input devices, for example, but often at the cost of functionality. For example, it is common for the surgeon to instruct a clinical assistant to interact with imaging systems on their behalf in the operating room (OR). This interchange can become complex, resulting in the surgeon having to talk the assistant through what is required, increasing cognitive workload for the surgeon, as well as causing delays to procedures. The interview study shed further light on clinician perspectives and attitudes towards touchless control, as well as practices and problems surrounding the existing use of PACS.
In order to better understand the process of designing and developing touchless image viewing systems, two distinct prototype systems were developed. The process of developing an initial touchless prototype (using the Kinect V2) provided insight into some of the challenges and decisions faced when designing a touchless interaction system. The insights from HP interviews were combined with the learnings from the initial prototype to develop a second prototype (using the Azure Kinect DK). HP experiences with this second prototype were explored. As unintended input is a particular challenge for touchless interaction in medical contexts, the performance of different interface latching (clutching) techniques is investigated. Analysis of semi-structured interviews with HPs following the experiment revealed an appetite for reliable touchless interfaces, a strong desire to reduce shared surface contact, and proposed potential improvements such as combined authentication and touchless control.
Given the wide variety of issues arising from the literature review, the interview study, the construction and empirical exploration of the prototypes, I further provide an framework the development process for touchless medical imaging systems, drawing also on existing research into the design of touchless user interfaces. This provides an overview of the design space, and provides a structured means for developing the design rationale for individual touchless interfaces for medical imaging.
Overall, the findings presented in this dissertation can inform the development of novel touchless medical systems and help identify opportunities for future research. | en |
