UBAR - Embryological Analysis System
.png)
An AI-based system that assists embryologists with data analysis, case monitoring, and team coordination, optimizing their decision-making process.
B2B
AI-BASED SOFTWARE
HEALTH TECH
UX/UI PRODUCT DESIGN
UX RESEARCH
USER TESTING
My Role In The Project
Define the nature and scope of the product alongside the company's senior management team
Hands-on design of the entire UX lifecycle, including research, detailed design, testing, and delivery
Manage the UI design process, starting from recruiting the designers to adding the final touches
Closely collaborate with colleagues across departments, such as medical advisers, developers, QA, and writers
What Is IVF, And What Does UBAR Have To Do With It?
Background On The World of Fertility (IVF)
The IVF process helps people with fertility problems achieve pregnancy.
In this process, an embryologist fertilizes eggs with sperm outside the human body and allows the resulting embryos to grow for 3-6 days in an incubator, where they are recorded on video or stills. Then, based on the recordings, the embryologist selects the most suitable embryo to be transferred to the patient's uterus, where it will hopefully develop into a pregnancy.
The User
Embryologists are fertility experts who usually work in clinics or hospitals as part of a team, and share responsibilities and patients. Their daily schedule often involves handling tasks that arise on the spot based on the current medical situation in the lab. They manage multiple systems simultaneously, including digital and paper-based interfaces, which can complicate integrating an additional system into their workflow.
What UBAR Does
UBAR assists embryologists in making professional decisions using AI tools. The primary tool predicts the likelihood of each embryo resulting in a healthy pregnancy, enabling the embryologist to select the most suitable one for transfer.
How It Works
Here is a short video of a prototype we presented at a fertility conference that created quite a buzz...
The Problem I Was Assigned To Resolve
Although UBAR has proven itself from a medical point of view, embryologists do not use it or only do so very passively.
Workflow Stages
Initial Review
Getting to know the system, identify its strengths and weaknesses, and understand the pain points that need to be solved
Research
Laboratory visits, interviews with embryologists and doctors, and in-depth analysis of the existing system
UX Concept
Creating an interface that will meet the user's needs and engage him in active use of it
UI Concept
Designing a user-friendly and distinctive look and feel
Execution
Detailed design, design-system establishment, handoff, and ongoing adjustments
Analysis and Adjustments
Analysis of usage statistics, screen recordings, and guided user testing
Research
I conducted extensive research to understand the users' work habits and their needs. This in-depth understanding allowed me to create a system that specifically met their requirements.
My research included:
-
Observations conducted in embryological laboratories
-
Interviews with embryologists, laboratory managers, and doctors
-
Participation in social media groups of professionals and patients
-
Benchmarking existing systems and learning the industry-specific interface conventions
Me, making a fashion sacrifice for research
Key Challenges In The Redesign Process
Challenge #1
Adapting The System To The User Workflow
The original system, while offering impressive algorithmic tools, primarily reflected the thought process of its creators, who were algorithmists and doctors, rather than following the mindset of the embryologists who used it. In addition, the system's structure was significantly influenced by the chronology in which it was built over time and did not serve as a consistent and integrated tool.
Solutions
01
Shifting From Individual To Teamwork
We went from assigning each patient to a specific embryologist to a collaborative approach where everyone could access and edit all patient records. We introduced new tools to make it easier for users to seamlessly pick up where someone else left off. To enable this, we created a "profiles" feature and implemented an "activity log" that resolved regulatory issues associated with this new structure.
Adding a personal profile into the laboratory sign-in process facilitated the integration of collaboration and personalization
A glimpse at some of the screens from the 'profile' feature
This 'activity log' page allows embryologists to continue each other's work and discuss decisions. It also creatively resolves a regulatory issue related to medical liability, which initially hindered the addition of the 'profile' feature.
02
Moving From Working On A Single Embryo To Working In Batches
This change mimics the practical method of using trays in the laboratory, making it more intuitive for embryologists and, most importantly, dramatically expediting and optimizing the data uploading process to the system.
The new 'batch upload' process saves both time and effort.
Shifting to batch processing permitted us to add a summary at the end of the upload. This not only provides a clear overview but also allows for relevant action suggestions.
03
No More Dead Time Of Waiting
The system's analysis process can take a while, occasionally extending to a few minutes, which leads to a frustrating and broken user experience. Many users spent this time on tasks outside the system, often abandoning their session. Given the technical limitations in reducing the waiting time, we decided to address this issue from a UX perspective by enabling and encouraging parallel work.
.png)
This pop-up summarizes my UX approach to solving the waiting problem:
-
Provide transparency and manage expectations about waiting times
-
Encourage multitasking for better task management
-
Proactively suggest relevant tasks for an easier next step
.png)
As users switch to other tasks, a pinned progress bar allows them to track background processes and conveniently return to them.
These are some examples of progress bars of multiple processes
(collapsed and expanded stats).
Challenge #2
Encouraging Active Use
Before the redesign, users were primarily focused on a single data point – the embryo's chance of resulting in a healthy pregnancy. Thus, they would typically check this information and little else. The company aimed to encourage users to explore the new tools introduced, actively provide feedback, and enter data, fostering collaboration for mutual benefit. The expectation was that users would engage with data entry as part of the system's guiding process, and the company counted on this to consistently improve the system's predictive capabilities.
Solutions
01
Design That Calls For Action
We moved from displaying screens loaded with information to uncluttered screens highlighting the actions to be taken. We focused on ensuring that the user's next step would be accessible, easy, and clear. The most dramatic expression of the new approach was in the introduction of a new home page.
BEFORE
This is what the original home screen looked like: laden with information and unclear what actions could be taken.
AFTER
The new home screen is all about relevant tasks rather than dry information. It's welcoming and guides your focus on what needs to be done.
BEFORE
This "embryo page" contains important input fields and actions. However, its buttons are easy to miss, and their functionality is unclear.
Let's spotlight these elusive elements...
These numbers are actually clickable and are the most important input on the screen
Each plus sign is a button that leads to a multiple-choice question
AFTER
The new input component,
designed as a series of simple questions
All input fields are gathered in one place, presented as straightforward questions, already open and ready for user responses. These questions follow one another, making the task look easy and ensuring a smooth workflow.
02
Bite-Sized Tasks
To encourage the user to fill in more input fields, we've simplified forms by displaying one short question at a time. In addition, in a way that may be counter-intuitive, we've minimized mandatory fields, to make the process as pleasant as possible, letting users fill in what they prefer and skip the rest to prevent abandonment.
BEFORE
This questionnaire was displayed after selecting the embryo grade. it interrupted the fluency of the work and primarily served to improve the system algorithm (not directly benefiting the user).
AFTER
Instead of showing all questions each time, we now present only one question per case, randomly selected from the relevant options. This question remains constantly visible, without intrusive pop-ups, ensuring it's entirely dependent on the user's willingness and availability.
Not only is the question different in each case but so is our "thank you" response.
Another good example is the new "Transfer outcome" record feature
It’s a streamlined form that allows users to track the treatment's outcome. It's structured as a step-by-step process, with users providing input in bite-sized steps to reflect real-time progress.
Here’s a short demo of how it works, including setting a reminder for future follow-up.
.png)
This is how the ‘transfer outcome’ section looks like at the end of a healthy pregnancy.
Above is an example of a few of the forms that need to be completed as part of the outcome follow-up. We put much thought into making it as short and straightforward as possible.
The Redesign Impact
This redesign was showcased at a major global fertility conference, sparking significant interest and leading to a new collaboration with a large international clinic network. Furthermore, introducing the new system to clinics already familiar with the original version, even before its full development, proved highly successful, with positive and objective results and user feedback.