How can we use AI to make healthcare better for humans?
STD Consulting Chatbot
STDs take a heavier toll on certain populations and communities, in particular, youth. However, social stigmas create obstacles for them seeking assistance.
Role / Contribution
User Experience Design
User Interface Design
Quo is an STDs Consulting Service Chatbot designed to tackle the issue of delayed medical support to young potential STDs patients. We offer an approachable, A.I. powered chatbot that can do an instant risk evaluation with Step by Step guidance on dealing with STDs.
Using the power of image recognition and chatbot consulting, we can provide diagnosis and medical referrals. Also, the Instructor provides medical resources such as a review system for clinics and putting patients in touch with an emotional support group. The product was developed with extreme privacy measures to ensure our user's privacy is safe and sound.
View Full Project ↓
Quo is an STDs Consulting Service Chatbot designed to tackle the issue of delayed medical support to young potential STDs patients. We offer an approachable A.I. powered chatbot that can do an instant risk evaluation with Step by Step guidance on dealing with STDs. Using the power of image recognition and chatbot consulting, we can provide diagnosis and medical referrals. Also, the Instructor provides medical resources such as a review system for clinics and putting patients in touch with an emotional support group. The product was developed with extreme privacy measures to ensure our user's privacy is safe and sound.
Brand new product/service pitch in 2 months.
DeepQ is the Healthcare department under HTC corporation. DeepQ is composed of cross-domain experts and engineers, such as computer science, software engineering, laws, user experience, design, etc. By developing artificial intelligence, blockchain, virtual reality / augmented reality technologies and products, DeepQ challenge to integrate advanced technologies in healthcare and clinical applications that progressively achieve precision medicine with humanity is woven in.
In 2019, our head of design decided to explore a new direction for medical products/services, which can utilize the strengths of DeepQ. He formed an independent UX task force with a UX researcher, a UX designer and a UI designer. I have the privilege to offer my capability as a UX designer. We had two months to establish a "Fresh New" product/service direction pitch.
In our first meeting, we went back to ground zero to clarify the strength of DeepQ and our positioning. We came up with this visual infographic of DeepQ strength fields. The 3 axises each represent the crucial cornerstone of DeepQ product, including Advanced technology, Medical Institution Co-operation Alliance and User Experience Developers. Our service space lies inside the area of this triangle.
Lacking certain types of strength might result in an avant-garde product. For instance, (i) is the case that immature technology and insufficient medical institution support. (ii) is the case in which our UX developers are inadequate to achieve the goal.
5-days intensive workshop
To kick off this project, we hold a 5 days-long intensive workshop. The objective of our workshop is to generate at least 3 actionable proposals for further development. We set some principles to guide the workshop's outcome.
(1) The topic should be in the medical/ healthcare domain
(2) The possible solution should be within the strengths of DeepQ
(3) Consider the time and resource we possess for this project
(4) Identify needs & insight with a few possible solutions.
(5) We won't dive into the design part deeply.
Proposals include project brief, context inquiry, secondary research, scholar researcher paper inquiry, persona, user journey map, Value Proposition Canvas, insights and opportunities, and potential solution.
Narrow 3 potential directions down into 1. Establish 5 factors to conduct evaluation. The STD topic wins the game.
(i) AIP: Artificial Intelligence Platform, an image recognition platform developed by DeepQ.
At the end of the 5 days intensive workshop, we present our 3 proposals to the whole UX design team. After review and critiques, we co-evaluate the directions with all the team members. Since all directions we propose are related to our strength in A.I. Image recognition, we evaluate these five directions based on
(1) Technical requirement feasibility
(2) Research Scope
(3) Data Collection Difficulty (A.I.)
(4) Annotate Difficulty (A.I.)
(5) Social Influence/Impact
After discussion and evaluation, we decide to further develop proposal 1: Sexually Transmitted Diseases (STDs) Helper. The proposal aims to provide a highly-private consulting chatbot for potential STD target audiences.
Based on the proposal, we reposition this project’s progress back to the research phase to complement the need for more structural foundation.
STDs & Youth
STDs take a heavier toll on certain populations and communities, in particular, youth.
STDs take a heavier toll on certain populations and communities, in particular, youth. According to the Centers for Disease Control Taiwan's research, the number of infected Gonorrhea patients has doubled in 5 years with the highest increase in age 10 to 19 (27%), and age 20 to 29 years (22%). Also, this trend happens not only in Taiwan but also around the world, according to The National Coalition of STD Directors (USA), Sexually active youth (aged 15-24) have the highest STD rates of any age group.
After we research the context of why they experience higher rates of STDs and still increasing, we synthesized our findings into this diagram.
Since the development of communication technology and the rise of mobile service, the young generation has easy access to adult content and flourishing dating app service which increases the chance of having sexual intercourse. However, the absence of proper sex education in schools with a lack of communication on STDs with health care providers and parents often leaves young people with little accurate information on STDs and how to protect themselves. Create a perfect storm resulting in the increasing trend of STD transmission among youth.
STDs are just infections that usually transmit through sex. They have no inherent moral or immoral component. They infect people regardless of race, gender, religion or sexual orientation. However, these social stigmas create obstacles for STDs patients seeking assistance, especially among youth. Young people face unique barriers to STD support services, including concerns about confidentiality, transportation, financial issues, shame, and embarrassment the impact of STDs in different populations.
...these social stigmas create obstacles for STDs patients seeking assistance ...
... concerns about confidentiality, transportation, financial issues, shame, and embarrassment ...
Many STDs are said to be socially stigmatized. In other words, people are judged or condemned for being infected with STDs. Other types of diseases may also carry some level of social stigma. However, the STD stigma is often particularly strong.
When someone is diagnosed with an STD and lets people know about it, there is often a very negative reaction. They may find that people think of them as dirty or slutty. They may lose partners or experience intimate partner alienation. Moreover, enter the vicious self-stigmatized circle.
Thoroughly understand our audiences' unpleasant journey
After conducting numerous interviews and secondary research, we concluded our findings to formulate a detailed persona and user journey mapping. We discovered that in comparison to the older population groups, the younger generation is warier of the social stigmas that come with STDs. As a result, they delay in addressing their symptoms and seeking medical assistance, which worsens their condition.
An A.I.Powered STDs Instructor, with symptom evaluation and medical consulting.
We proposed an A.I. powered STDs Instructor. Using the power of Artificial Intelligence image recognition and NLP consulting, we offer instant risk evaluation and medical referrals. Also, the Instructor provides medical resources such as a review system for clinics and putting patients in touch with an emotional support group.
Not just a product, but a holistic service design.
We used the value proposition canvas to explore the insight that we gained from the research phase into possible solutions. We synthesized our knowledge of the target audience into the customer segment. We laid out the data and pinpointed the following:-
The job that needed to be done
The pain points of the process
The gains, which is the additional expectation from the audience.
We then brainstormed the ideas we can offer.
In severe times, what is the best medium for our target audience?
How can we humanize technology that enables patients to share their concerns?
After thorough discussion and consideration, we proposed a chatbot approach. A chatbot that acts as an exclusive consultant that possesses the power of artificial intelligence with knowledge and database of STDs. Besides, according to the research conducted at MIT, people tend to be more honest with a machine than compared to a real person. Therefore, we created a space for potential patients of STDs to express their concerns and gain tailor-made assistance and advice.
We decided to use the KANO method to evaluate our ideas. The Kano model is a theory for product development and satisfaction developed in the 1980s by Professor Noriaki Kano. According to this theory, services can be classified into 5 categories, prioritized by features and development sequence.
We reach out to the CDC (Center for Disease Control) Taiwan and NGO communities that support STDs associated services for doing the evaluation.
#1 A.I. Risk Evaluation
#2 Mentor Guidance
#3 Stress Relief Game
#4 Patient Support/ Sharing
#5 Medical Resources Review
#6 Conversation Template
#7 Extreme Privacy Measure
#8 Follow-up Care
We used our DeepQ Artificial Intelligence Platform to do the pilot testing of our concept, which inputs around 500 images of STDs and non-STDs symptoms to train our A.I. model. Here, we used two different annotation methods, one is classification, and the other is object detection.
The result is promising. We have an 84% accuracy in classification, and 26% in object detection just from 500 images, which proves that the concept is workable.
After refining the essential features of our product, we consult with our developer to establish our service blueprint. This was to understand how our exterior and interior service interacts with each other. Moving ahead with a more holistic view of what our service should offer, we narrowed down the detailed UX flow of what we should develop. At the beginning of our journey, we need to know the status of our users in order to provide appropriate services. Therefore, triage flow is crucial for our service.
We began by generating the wireframe. We wanted to keep it as simple and intuitive as possible. Therefore most of our services are present in the chatroom with artificial intelligence A.I. chatbot. We focussed more on creating a template and UI set for different user scenarios to fit every need. However, we still need a status user interface to demonstrate the toolkit and service we provide. Therefore, the staus page serves as an action panel to trigger service.