Showcase
Philips in association with the NHS asked for an AI based solution for the medical industry that would aid in the prevention and treatment of chronic diseases.
There has been a number of attempts to use AI to aid the medical industry though to date few have seen success.
In this project an AI enhanced system was designed to aid the medical industry in a different way. The system does not try to perform a medical task, rather it has been designed to aid the medical professionals who do.
Team
Robert Cade
UX Lead, Strategy, Project Manager.
Anne-Marie Kroupova
UI Lead
Lois Hunt
Graphic design
My Role
Product Design
Discovery Research
Strategy
Ideation
Interactive Prototypes
Team Management
The Challenge
Brief – How might we use AI to support people to reach a happy, meaningful and productive one hundred year life?
Background – Global life expectancy has risen by more than seven years since 1990. We are living longer than ever before but chronic diseases are now the leading cause of death worldwide. The World Health Organization states that 60 percent of all deaths are due to chronic illnesses, such as heart disease, stroke, cancer, chronic respiratory diseases and diabetes. – These chronic illnesses are a growing issue
across the world; they require ongoing management over a period of years, decades even, and threaten to overwhelm existing healthcare systems, societies and economies. Furthermore, according to the Association of American Medical Colleges the US will face a shortage of more than 100,000 doctors by 2030.
Minimal Design
Lifestyle Design
Medical Assistance
Creative Methods
Considered and planned from end to end.
As Lead on the project, my team owned all aspects of the research, strategy and creative vision. Here are a few of the highlights I set up to help this very talented group of individuals to thrive.
Industry Review
Industry review facilitates an understanding of a project’s position relative to other similar products or services. Understanding the forces at work in the industry is an important component of effective strategic planning.
User Testing
the process of collecting information about usability and overall user experience from actual users during the design process. This is done through a variety of user testing methods.
User Journey Mapping
A user journey map (also known as a customer journey map) is a diagram that visually illustrates the user flow through an experience, starting with initial contact or discovery, and continuing through the process of engagement into long-term involvement.
User research & Analysis
Helps us identify and prove or disprove our assumptions, find commonalities across our target audience members, and recognize their needs, goals, and mental models.
Product strategy
A product strategy is a high-level plan describing what a business hopes to accomplish with its product and how it plans to do so. The strategy should answer key questions such as who the product will serve (personas), how it will benefit those personas and the company’s goals for the product throughout its life cycle
Hi-fidelity Prototypes
Industry review
Microsoft made big headlines when they announced their new chatbot. Writing with the slang-laden voice of a teenager, Tay could automatically reply to people and engage in “casual and playful conversation” on Twitter.
Tay grew from Microsoft’s efforts to improve their “conversational understanding”. To that end, Tay used machine learning and AI. As more people talked with Tay, Microsoft claimed, the chatbot would learn how to write more naturally and hold better conversations.
By flooding the bot with a deluge of racist, misogynistic, and anti-semitic tweets, Twitter users turned Tay – a chatbot that the Verge described as “a robot parrot with an internet connection” – into a mouthpiece for a terrifying ideology.
After a cursory effort to clean up Tay’s timeline, Microsoft pulled the plug on their unfortunate AI chatbot.
No AI project captures the “moonshot” attitude of big tech companies quite like Watson for Oncology. In 2013, IBM partnered with The University of Texas MD Anderson Cancer Center to develop a new “Oncology Expert Advisor” system. The goal? Nothing less than to cure cancer.
The first line of the press release boldly declares, “MD Anderson is using the IBM Watson cognitive computing system for its mission to eradicate cancer.” IBM’s role was to enable clinicians to “uncover valuable insights from the cancer center’s rich patient and research databases.”
So, how’d that go?
Medical specialists and customers identified “multiple examples of unsafe and incorrect treatment recommendations,” including one case where Watson suggested that doctors give a cancer patient with severe bleeding a drug that could worsen the bleeding.
The medical industry has been impacted by the advancement in AI though not in a positive way. Like any industry that becomes early adopters of technology, they have been subject to failures. The worry is that as lives depend on the medical industry they may be slow to adopt AI driven products in the future.
The research suggests that some of the projects may have been overly ambitious, seeking to revolutionise rather than support. This observation influenced the design principles used here and ultimately lead to a versatile product designed to enhance and support.
Strategy
Research showed that the AI systems of the past have often gotten in the way. Users were mistrustful of the systems preferring instead to be treated by a human. With this in mind this system was designed to be much less intrusive.
This system was designed to have access to cloud-based data collected via patient authorised apps and services. The patient would use a mechanism similar to ‘login in with Facebook’ and data would be uploaded to a cloud. The AI would be able to analyse the data and make suggestions and predictions based on user data. These suggestions and statistics would then be available to medical professionals via a desktop widget. Below is an explanation of how the system would perform
Traditionally patients speak to a medical professional, most often their GP when they have a health issue. When not feeling their best research suggests that patients prefer to interact with another human.
While with their GP a patient describes symptoms and is examined. From there the GP will look at the patients history, identify the symptoms and order further investigation. This whole process takes time. Eventually the GP will form a hypothesis and prescribe treatment which may or may not help the patient. In the latter case the process must be repeated in order to eventually find the correct diagnosis.
The proposed system uses the same traditional process but ties in at various points using behaviours that patients have already adopted. Patients have already gotten used to wearable smart technology as well as using institutions such as Facebook and Google to securely sign into systems and services.
The system would use information collected in this way to help in the diagnosis process. Through this data, the AI can suggest possible causes to the medical professional conducting an examination.
There are two distinct user-facing parts of the system. One deals with the patient and is designed to be as unobtrusive as possible.
The other is designed to aid a medical professional. It has been designed to be an aid to their workflow.
This video gives a high level overview of the system as a whole.
DESIGN + PROTOTYPING
Patient Interface
Here is a more detailed look at what the patient interface might look like with a device like a Fitbit. It uses the same mechanism as sign in with Facebook which is familiar to the user.
A Closer look
DESIGN + PROTOTYPING
Medical Professional Interface
This is a visual of what the medical interface might look like as used in a consultation. It uses familiar mechanics and visuals to be as intuitive as possible.
A Closer look
The Benefits
- Integrates with patients’ current lifestyle.
- Integrates with medical professionals’ current workflow.
- Could track and trace infections in a pandemic.
- Would cut down on waiting times and workloads for medical labs.
- Impossible to misdiagnose patients as there is no direct diagnosis.
- Scalable to integrate new features and technologies