Aliado

Designing Collaborative
AI for Surgeons

Aliado was part of a research project exploring how AI-Systems could help surgeons in cancer treatment, which is characterized by information overload. The challenge was to understand limitations of AI-Models in clinical usage and align our and stakeholders' thinking to find suitable solutions.

Futher Facts

Timeframe:
12 Weeks, 2020/21

Skills: 


AI-UX, Shadowing, User Research, Co-Creation Workshop Facilitation

Team
Frederic Myers (Designer)
Sven Hornburg (Designer)
Johanna Brandenburg (PhD Student)
Michael Haselbeck-Köbler (PhD Student)
André Schulze (PhD Student)

Cooperation Partners: 


UKHD, NCT




Awards:
UXDA 2021, aed neuland 2021

My Role

I researched AI-Technology, prepared the AI & Data Scientist Expert Interviews.

Conducted user research and shadowing together with my fellow students to gain insights.

I was developing the final concept for the surgeons side, worked on wireframes and the prototye.


THE Challenge

Surgeons are running short on
time in the fight against cancer

Surgeons are running short on time in the fight against cancer

+47%

rise in cancer patients
until 2040

rise in cancer patients

until 2040

+36%

+36%

more medical data per year until 2025

more medical data

per year until 2025

44%

44%

of clinicians time is spent on paperwork

of clinicians time is

spent on paperwork

THE Outcome

AI-System supporting surgeons 


during decision-making process

AI-System supporting surgeons during decision-making

↑ Watch the video to see how the solution looks like

01 - Structuring medical findings
to make them more accessible

Medical findings are hard to access, even when digitally stored taking up to two minutes to find. They are also not machine readable. The Aliado System scans digital documents using specific search algorithms and highlights important parameters for the doctor to investigate before meeting with the patient. He can annotate these to give the AI-system feedback.

↑ Analizing the findings and highlighting parameters

02 - Summarizing Information

 to
support decision-making

After gathering data, surgeons have to prepare a case overview to serve as a base for the decision-making. The Aliado AI-System saves time by automatically generating a case overview out of the patient data. This improves decision-making and creates machine-readable data to train further ML Models.

↑ Glanceable overview of all important patient data

↑ Glanceable Patient Overview

03 - Surfacing Information to
augment surgeons experience

Since a patients condition can change quickly surgeons often have to revisit decisions during surgery. Out of the cases database, Aliado can extract similar patients for surgeons to reference, allowing them to leverage the collective experience and expertise of other surgeons for their decision.

Since a patients condition can change quickly surgeons often have to revisit decisions during surgery. Out of the cases database, Aliado can extract similar patients for surgeons to reference, allowing them to leverage the collective experience and expertise of other surgeons for their decision.

↑ Comparing past patient cases in OR

↑ Comparing similar patient cases in OR

THE Process

Co-Creating a realistic vision
for AI in surgical applications

Co-Creating a realistic vision for AI in surgical applications

01 - ResearcH

Getting up to speed and into the nitty-gritty of clinical processes

More

5

Days spent in hospital

16

Expert Interviews

02 - Synthesis

Validating and complementing insights, identifying surgeons' pains

More

Piecing together insights with medical students in 2 workshops

Investigating into problems we found in the first phase deeper

03 - Ideation

Brainstorming ideas on how AI could help surgeons during treatment

More

Multiple co-creation sessions with surgeons from various experience levels

Key wishes were a brief overview of relevant patient data & support on decisions

04 - Turning Point

Pivoting the concept towards structuring medical data to make it suitable for AI-Applications

More

We sat together with Data Scientists to get feedback on our ideas

The lack of suitable data made our concept unrealistic

05 - Design

Tailoring the solutions to surgeons needs through co-creation

More

3

Co-Creation Days

Refinining Information Architecture and views

Impact

Learning about the role of Design
and getting intrigued by the potential of AI

Learning about the role of Design and getting intrigued by the potential of AI

01

Set up a sufficient data infrastructure

Set up a sufficient data
infrastructure

02

Protect privacy & invest in data security

03

Incorporate means to give Feedback

04

Account for system evolution

Account for system evolution

05

Combine strenghts to eliminate weaknesses

Combine strenghts to eliminate

weaknesses

05

Make Decision Processes transparent

05

Respect the users emotional needs

08

Keep the Human in control at all times

↑ Our (preliminary) principles for designing AI-Experiences. I am totally up for a chat discussing AI & Design

↑ Our (preliminary) principles for designing AI-Experiences. I'd love to chat about AI & Design

Looking back two years later, the influence of this project on my thinking was immense. But there have been 2 Learnings that proved to be insightful:

The importance of Designers as mediators.

Our basic understanding of both disciplines enabled us to bridge the gap between Data Scientists and Clinicians. This lead us to a realistic concept that still catered to the visions and needs of our stakeholders ultimately aligning both disciplines behind a common goal. Especially in complex projects (like AI for Medical Purposes) interdisciplinary work is required in which such a generalist/mediator skillset is beneficial.



The game-changing potential of AI and its Implications on Design




My initial interest in AI became an obsession after our thesis. Why? Because Human-AI-Collaboration is a genuinely new paradigm with enormous potential. Especially for the field of Interaction Design which has been in a one-way-street towards maximizing screentime this is great news: More intelligent technology opens up the potential to interact with it in more natural and humane ways.

But to successfully design such experiences, new skills are required: Understanding Limitations and Strengths of AI-Technology, defining guidelines that ensure the great power that AI-Systems bring is also utilized with the right level of responsibility.

Seeing that there haven’t been much best practices out there so far, we decided to distill our learnings into design principles. We presented them already at some talks, opening up discussions I believe are really necessary as we transition from the Information Age into the Age of Intelligence.

The importance of Designers as mediators.

Our basic understanding of both disciplines enabled us to bridge the gap between Data Scientists and Clinicians. This lead us to a realistic concept that still catered to the visions and needs of our stakeholders ultimately aligning both disciplines behind a common goal. Especially in complex projects (like AI for Medical Purposes) interdisciplinary work is required in which such a generalist/mediator skillset is beneficial.



The game-changing potential of AI and its Implications on Design




My initial interest in AI became an obsession after our thesis. Why? Because Human-AI-Collaboration is a genuinely new paradigm with enormous potential. Especially for the field of Interaction Design which has been in a one-way-street towards maximizing screentime this is great news: More intelligent technology opens up the potential to interact with it in more natural and humane ways.

But to successfully design such experiences, new skills are required: Understanding Limitations and Strengths of AI-Technology, defining guidelines that ensure the great power that AI-Systems bring is also utilized with the right level of responsibility.

Seeing that there haven’t been much best practices out there so far, we decided to distill our learnings into design principles. We presented them already at some talks, opening up discussions I believe are really necessary as we transition from the Information Age into the Age of Intelligence.

The importance of Designers as mediators.

Our basic understanding of both disciplines enabled us to bridge the gap between Data Scientists and Clinicians. This lead us to a realistic concept that still catered to the visions and needs of our stakeholders ultimately aligning both disciplines behind a common goal. Especially in complex projects (like AI for Medical Purposes) interdisciplinary work is required in which such a generalist/mediator skillset is beneficial.




The game-changing potential of AI and its Implications on Design




My initial interest in AI became an obsession after our thesis. Why? Because Human-AI-Collaboration is a genuinely new paradigm with enormous potential. Especially for the field of Interaction Design which has been in a one-way-street towards maximizing screentime this is great news: More intelligent technology opens up the potential to interact with it in more natural and humane ways.

But to successfully design such experiences, new skills are required: Understanding Limitations and Strengths of AI-Technology, defining guidelines that ensure the great power that AI-Systems bring is also utilized with the right level of responsibility.

Seeing that there haven’t been much best practices out there so far, we decided to distill our learnings into design principles. We presented them already at some talks, opening up discussions I believe are really necessary as we transition from the Information Age into the Age of Intelligence.

About me

I am a passionate & curious mind.
Feel free to connect on LinkedIn

© 2022 Hannes Maurer

Created with 💙 in Esslingen

About me

I am a passionate & curious mind.
Feel free to connect on LinkedIn

© 2022 Hannes Maurer

Created with 💙 in Esslingen

About me

I am a passionate & curious mind.
Feel free to connect on LinkedIn

© 2022 Hannes Maurer

Created with 💙 in Esslingen