Apollo AI
A multiuser platform for cancer care powered by AI, aimed to significantly improve workflows for nurses, empower oncologists with personalized care tools, and provide cancer patients with a more supportive and streamlined treatment experience.
Overview
My approach combined storyboards and strategic planning to innovate the development of an AI product. Visualizing the cancer care journey from start to finish using storyboards similarly done in films, ensured the designs were user-centric and effective while iterating based on user testing. I was able to create a thoughtful product that resonated with three different users intertwined closely.
Problem
Existing cancer care systems face numerous challenges, including inefficient coordination, complex decision-making processes, and limited patient engagement. These issues contribute to suboptimal outcomes for patients and providers alike, highlighting the urgent need for innovative solutions.
USERS:
• Oncologist
• Oncology Nurse
• Cancer Patient
Design Objectives
Product Design
Design a product leveraging AI models trained in medical data and provide a proof of concept that addresses real-world problems in cancer care, suitable for an MVP 1 release.
As an HCI AI product designer
My focus was on designing AI-powered systems that prioritized user needs, human-AI interaction, ethical considerations, and accessibility principles.
Protect patient privacy.
Be transparent about how AI works.
Prevent biases in AI algorithms.
Get patient consent.
Establish accountability and governance.
Empower patients to understand and participate.
Team Structure
My Role & Responsibilities
Design Strategy
UX/UI Lead
Visual Design
Teammates & Collaborators
Product Manager
3 Engineers
4 Subject Matter Experts
1 Business Development Manager
Timeline + Process
We had 12 weeks to prove the technology’s capabilities and improve the user experience, prepping for engineer hand off to launch an MVP.
Design Strategy kicked off the product design
The design strategy for this product began with the innovative use of storyboarding, a method inspired by Hollywood practices.
Storyboard method instead of user journey map
By visualizing the cancer care journey through storyboards, we were able to identify key pain points, user interactions, and opportunities for AI integration. This approach allowed us to kick off the product design process with a deep understanding of user needs and a clear direction for leveraging AI technology to address them.
Read the full Design Strategy here
Other parts of the process include:
Research + Discovery i.e. design thinking workshop to identify users and cancer care journey
3 user testing along with 3 iteration
Visual design
Prep for development handoff
Provide design support
User Research Insights
Consistent feedback:
Difficulty coordinating care between providers causes problems because it leads to confusion, delays, and inefficiencies in treatment. Patients may not get the care they need on time, leading to higher costs and dissatisfaction. It can also make it harder for healthcare teams to work together, risking patient safety and quality of care.
Other key insights
Long waits for appointments, tests or treatments.
Difficulty coordinating care between different providers.
Financial strain from medical expenses and insurance.
Patients are too weak to log their progress will need a caregiver.
Uncertainty throughout treatment, side effects impacting quality of life.
Challenges navigating the healthcare system and understanding options.
Preliminary wireframes used for the first user testing
Other user feedback
Oncologist:
"I’m a big supporter of personalized care, because the reality is our human bodies are not alike. Even with the same type of diagnosis, same age or comorbidities… not all cancer patients will respond similarly with the same standard cancer care treatment."
Nurse:
"Always about communication breakdowns since a patient’s information are scattered across different systems, making it difficult to track and coordinate care effectively. It would be nice to just see what tests are missing or where to get their imaging."
Cancer Patient (Ovarian cancer survivor):
"I was too weak to keep track of how much water I was supposed to be drinking after radiology, there’s no way I’d be able to log on and track how I feel during treatment on this platform."
Prioritization
Product and I narrowed down the scope of what to solve since we couldn’t solve every problem identified.
Insights from the user testing prompted us to refine the coordination flows closer to an MVP:
Making workflows smoother by automating tasks
Improving communication among healthcare teams.
Personalizing treatments based on patient data.
Better coordinating care and updates.
Empowering patients with information and support.
Enhancing the overall quality of care.
Areas in blue indicates the AI presence in assisting with coordination
Coordinated Experience
Oncologist’s Dashboard:
Providing oncologists with comprehensive patient information and AI-generated insights, the dashboard enables more informed decision-making and personalized care planning during the first consultation for diagnosis.
Patient demographics and medical history
Diagnostic test results (e.g., imaging scans, biopsy reports)
Symptoms and presenting complaints
Family history of cancer
Risk factors and comorbidities
Initial assessment and provisional diagnosis
Recommended next steps or further investigations
Treatment options and considerations
Decision support tools based on AI analysis
Appointment scheduling and follow-up plans
Nurse view:
With AI's help, nurses can track patient treatment more effectively. The dashboard shows predictive analytics to prevent issues, personalized care recommendations, and alerts for changes in the patient's condition. It also optimizes care coordination and offers decision support tools for better care delivery. Overall, AI enhances patient care and outcomes.
Patient Overview: Basic patient information and current treatment plan.
Treatment Progress: Timeline of appointments, procedures, and medication administration.
Vital Signs: Monitoring of vital signs and symptoms with alerts for abnormalities.
Medication Log: Record of medications administered.
Test Results: Access to lab results and diagnostic findings.
Care Plan: Detailed care plan and protocols.
Communication Tools: Integration with communication platforms for collaboration.
Patient Support: Access to educational resources and support materials.
Patient or Caregiver’s view:
The patient dashboard provides a comprehensive overview of the patient's medical information and treatment journey. The AI plays a crucial role in monitoring for adverse reactions or side effects by analyzing data inputs from the patient's symptoms, treatment regimen, and medical history. It helps by flagging potential issues early, allowing healthcare providers to intervene promptly and adjust treatment plans as necessary. This proactive approach enhances patient safety and contributes to better health outcomes.
Personal information and medical history
Current medications and treatments
Adverse reactions or side effects experienced
Symptoms and changes in health status
Treatment schedule and adherence
Communication tools for contacting healthcare providers
Access to educational resources and support services
Alerts for potential adverse reactions detected by AI
Reminders for appointments and medication refills
Option to report symptoms or concerns to healthcare team
Adverse reactions detected:
When an adverse reaction occurs and the AI detects abnormalities entered by the patient, the alert to the nurse on mobile are clear and actionable. This streamlined communication ensures that healthcare providers can respond promptly and effectively to adverse events, improving patient safety and care quality.
The Ai’s ability to include details such as:
Type of abnormality detected and its severity level
Relevant patient information (such as medical history and current treatment)
Any recommended actions or next steps
Alert provide prompts or options for the nurse to acknowledge receipt of the alert
Nurse can escalate the issue to other care teams if necessary
Document any interventions or follow-up actions taken.
Constraints & Evolution
Despite the potential of AI, we faced constraints such as the limited availability of AI models tailored specifically for cancer care.
This impacted the user experience where I was only able to design 50% of the user experience.
However, we planned to overcome these challenges by evolving our platform over time to incorporate additional AI capabilities and features. This iterative approach will allow us to adapt to changing needs and continuously improve the platform.
Conclusions + Learnings
By combining the insights gained from storyboarding with the capabilities of AI technology, healthcare product designers can create innovative solutions that have the potential to revolutionize cancer care. Despite constraints, the journey of designing and developing such a platform is one of continuous learning, adaptation, and a relentless commitment to improving patient outcomes.
Storyboarding drives innovation by visualizing the cancer care journey.
Available AI models enhance care by streamlining workflows and empowering patients.
Adaptability is crucial in design to work with available resources.
Continuous learning integrates new AI advancements for improved solutions.
User-centric design ensures the platform meets user needs effectively.