.png)
"She always serves customers with the best UX design. She is really cheerful and a happy person." - Mr. Mishra, Bajaj Sr Manager
"She always serves customers with the best UX design. She is really cheerful and a happy person." - Mr. Mishra, Bajaj Sr Manager
DIRECT - MICROSOFT AI FOR GOOD
Diabetic retinopathy using computer technology
Problem :
​
Diabetes is prevalent among over 4 million patients and is a leading cause of blindness among the senior adults in US and developing countries, especially India. Ophthalmologists suggest that a low-cost, easy to use solution that can be used by non-medical professionals will help solve the actual problem of early diagnosis of DR.
My role :
​
UX Research & UX Design Lead , Product Manager
​
​
Team :
​
In collaboration with Microsoft sponsors , my team, mentors, and medical experts.
​
​
​
Microsoft's goal :
To construct a low-cost, easy-to-use, high-image-quality “AI Fundus Camera” with “point-and-shoot” operation .
who
where
what
why
​
How might we encourage diabetic patients from United States and developing nations to early detect diabetic retinopathy in order to save time, cost and to avoid being more vulnerable to vision loss?
​
​
Problem Statement :
This affinity diagram was very useful in understanding how we look at this problem. Also, it helped identifying the potential "jobs to be done" for short term and long term.After having some understanding about the problem space, it was time to know more about our stakeholder, for which conducting stakeholder analysis helped in identifying stakeholders, their relationship with the problem, and qualitative research methods applicable to major stakeholders.
Insights
1
Interview template
The solution - in progress
And
A low-cost, high resolution fundus camera
A supporting AI-based web and mobile app
The Process
1
2
2
3
3
4
4
Requirement gathering
Research
Validate research
Ideation
Iterate
Test
Design
7
6
5
5
Requirement gathering
Understanding the business
Before diving deeper into the actual problem is, it was important to know what the business thinks about the problem space. To understand the same, we brainstormed, discussed and came up with following insights about MIcrosoft's vision to address the problem.
Vision
Help people & businesses throughout the world realize their full potential.
Mission
Infuse data science and AI to solve world's greatest challenges.
Goal
To construct a low-cost, easy-to-use, high-image-quality “AI Fundus Camera” with “point-and-shoot” operation .
Research
User Research
1
Secondary Research
2
Primary Research
Triangulation
3
Team DIRECT was all set to start the user research. But, we're systematic, so we needed a plan and hence with came up with this standard sequence of research.
Secondary research -> find gaps
Primary research -> fill gaps from secondary research.
Triangulation -> Triangulate the results by comparing attitudinal and behavioral data and see if we find any interesting insights ;)
Diabetic Retinopathy is prevalent in the global market, but more so in developing countries, specially in India.
-
-
The Global Diabetic Retinopathy Market is projected to grow at a significant growth rate of 7.3% over the forecast period 2020-2026, as concluded in a research report.
-
-
DR could be treated if patients get diagnosis and therapy early.
-
Ophthalmologists are short-handed in developing countries.
-
Our competitors (D-EYE and oDocs) should be used by medical experts and rely on their diagnosis. Also have low stability and simple functionality in companion apps.
-
COVID increases uncertainties and challenges for our project.
1
Key findings
secondary research
Feedback time
Before we put in cost into primary research, it was important to understand if our team and clients were on the same page. While we believed everything’s on the right track, we learned from feedback that the immediate goals aren’t getting attention because of major focus on the long-term goals. Hence, it was time to take a step back and shift the focus to achieve goals of immediate priority. For the same, I created mental map inspired from the discussion during feedback where I divided the problem into two halves, phase-1 and phase 2 or short term and long term. We further met in team to divided each phase into subsections that gave more clarity on specifics like what are the immediate goals, or what are the immediate jobs to be done ? etc. In short, we converged using mental map which looked like shown below :
_edited.png)
Primary Research :
The mental map directed us towards the primary research plan.According to the plan, doing qualitative research made most sense because we were trying to get answers for "Whats" and "whys" as opposed to "how many" or "how much". Breaking this down further, the information we needed to know was highly likely to be attitudinal and behavioral, and hence we chose methods like expert interviews, observations & contextual analysis and diary study. Stay tuned for the glimpse of each method implementation :
​
1. Expert Interview insights
2. Observation & contextual enquiry
3. Diary study.