Healthcare Interaction Design
Role: Team Leader, Design Researcher, and Experience Design
Problem
Dizziness is one of the most common complaints in ER and one of the most difficult to diagnosis. The causes of dizziness can be life threatening (e.g. stroke) or non-life threatening (e.g. vestibular neuritis). Currently, when a patient comes into the ER with dizziness they must go through time-consuming, qualitative exams and expensive testing to determine the cause of dizziness.
Solution
My team and I designed the digital interaction for VertigoMetric's (a startup from Northwestern) novel retinal diagnostic procedure. The interface mirrors existing diagnostic interfaces to fit seamlessly into the existing ER workflow. The final design walks through initial setup of the equipment and how use the interface. The primary features are the image processing screens and the final results page which enable a physician to get objective data quickly so they can make a decision on treatment.
Value Proposition
By providing ER physicians with with accurate, objective information, we are enabling them to more accurately and quickly differentiate between life threatening and non-life threatening conditions (such as stroke and vestibular neuritis). This will save valuable physician time and hospital resources while adding clarity and confidence around an otherwise subjective and stressful diagnosis.
Teammates: (all MS Engineering Design Innovation)
Katrina Shah
Sally Park
Hua Chin
Sheila Lo
Personal Contribution:
Created discussion guides and led user interviews
Drafted design hypothesizes (our assumptions around feature/function needs)
Created low-fidelity paper prototypes to test hypothesizes and user experience
Led user testing of interface and physical designs
Synthesized research and testing into design requirements
- Leveraged UI principles to created the final design of the image processing screen. Narrated the final product video.
Needs Finding and User Research
The goal of our research was to understand the features and limitations of existing retinal diagnostic systems, how ERs currently triage dizzy patients, and the key features and information needed by healthcare providers to enable them to make quick, accurate decisions about a patients care.
- Researched existing retinal, CT, and MRI imaging software systems to understand the what features users are accustomed to and how the information is communicated in a digital system.
- Interviewed an ER nurse, medical student, Neurologist, Ophthalmology technician, and an X-ray technician to understand the roles different stakeholders play throughout a patients journey and their needs/goals.
- Interviewed lead physician and biomedical engineer from VertigoMetric to understand their needs from a technology perspective.
Teammate Hua getting a retinal scan to better understand current workflows
2x2 of Stakeholders throughout patients journey vs. their level of decision making for patients care
Opportunity Identification
Through our research, we identified several opportunity areas in which we could work with VertigoMetric but scoped our project to focus on two areas so that we could make the biggest impact in the 10 weeks we had:
1. Creating the digital interface used to process the retinal images
2. Designing the visual communication of the retinal analysis for the physician in both digital and PDF formats
Key Observations:
- The ER is a very complex, intense, emotional environment where stakeholders have specific roles and responsibilities. Dizziness adds to that complexity in that it is very difficult to diagnose.
- Verto has a specific use and ER clinicians are not specifically trained in retinal anatomy or pathologies.
- Current methods for diagnosing the cause of dizziness are primarily qualitative. Physicians want objective, quantitative results so they will be able to more confidently make a decision on the next steps in caring for a patient
Design Imperatives and Prototyping
1st round of paper prototypes - testing what features to include
2nd round of paper prototypes - testing features and how best to communicate information
1st round of digital prototypes - testing format, design, and functionality
At the beginning of each design sprint we defined what we wanted to learn and created prototypes that would allow us to test different design hypotheses (see images above). We conducted user testing, first with paper and physical prototypes, and then with higher fidelity digital prototypes. At the end of the quarter, we settled on a set of design imperatives that would guide our final design concept.
User testing with neurologist
High-Level Design Imperatives
Quickly integrate - Verto hardware and software must complement or improve the current ER workflow and align with stakeholders roles and responsibilities
Minimize learnability and maximize efficiency - interface must be designed so that non-specialized/infrequent users can quickly go through the examination process, get accurate results, and save diagnosis time.
- Save time and improve confidence - Provide physicians with objective information to allow them to make an informed decision quickly.
Final Deliverable & Impact
We created a digital interface that enables the user to quickly and accurately identify and mark the optic disc and fovea and analyze their positions relative to each other. We also created a results page that provides the physician with quantitative data so that they can more confidently make a decision on how to best treat their patient. Our final deliverable consisted of a video that walked through the entire digital interface from user login in to image analysis and creating the results page. Video created by Katrina Shah, narrated by Erica Isaacs.
Image Process Page - on this screen the user (physician or nurse) can adjust the position of the crosshairs to ensure they are in the correct place before the software calculates the retinal skew or angle between the optic disc and fovea. The user can use the mouse to move the crosshair manually or use the sliding bars on the corresponding tomogram (cross-section) to change the position. The dark spots in the 3D image and the dips in the tomograms correspond to the Macula (fovea) and Optic Nerve (disc), respectively.