Spectral MD, Inc.

DeepView SnapShot® AI
Wound Imaging System

My responsibilities

From creating workflows, wireframes, prototypes, mockups and even commercial videos - you name it.

Our Custumers

Government

In mass casualty events like 911, or battlefield injuries, medical resources are stretched thin and every second matters.
💰Sponsor: Biomedical Advanced Research and Development Authority (BARDA) and Department of Defense (DoD)
🤔️Challenge: Limited medical resources need to be allocated reasonably
💡Solution: Rapid burn assessment prioritizes critical cases, ensuring faster care
📊Impact: Optimized resource allocation, reduced patient suffering

Commercial market‍

Hospitals and emergency rooms struggle with burn injury accuracy
🔍Research Insight:
     - 70% of patients are overestimated
      - Physician accuracy: 64-76%
💡Solution: Precise assessment reduces unnecessary treatments, wait times, and costs
📊Impact: Streamlined decisions and reduced medical expenses

Design Challenges


Problem

Our approach

• Complex clinical workflows resulted in long patient wait times
• Existing workflows were not adaptable for use in mass casualty events and emergency rooms

• Streamlined workflows to boost efficiency and adaptability
• Reduced the steps needed for results, cutting patient wait times
• Designed a system scalable for both government (mass casualty events) and commercial (burn centers, ERs) use cases

• Users struggled with touch interactions on cart-mounted screen because they need to raise arms for entering patient data
• Understanding AI-generated outputs was a challenge, as users found the information complex and difficult to interpret

• Redefined user interactions with screen, some classic UI components in mobile devices can not be used
• Simplified AI outputs with clear visual cues and explanations, reducing cognitive load and helping users quickly understand complex AI insights

• The absence of a unified design system led to inconsistencies in visual elements, impacting brand cohesion
• Color contrast issues made it difficult for users with visual impairments to navigate the interface

• Built a complete design system from the ground up in Figma, defining typography, colors, and reusable components to ensure visual consistency
• Improved color contrast to meet WCAG compliance, enhancing visibility and accessibility for visually impaired users

Project outcome

With the latest UI improvements, Spectral MD has achieved significant milestones:
‍Government Contract Success
• Initial BARDA contract fulfilled; extended research and development secured until 2028
• Software as a Medical Device (SaMD) is on track for FDA De Novo submission
‍
Positive Impact for healthcare professionals
• Expended 30+ hospitals across the UK and US participated in clinical studies
• Compared to old UI, we observed in usability testings of 31% reduction in time needed to estimate burn wounds for new users
• Easy to understood UI improved communication with patients and families

‍"The device helps us focus on critical cases faster while simplifying communication with families — a game-changer in high-pressure environments."
— Nurse Feedback

User research

Complex(and out-dated) electronic health records in hospitals
@Balor Scott & White, Dallas TX
Physicians tried to use our device
@University Medical Center, New Orleans LA
Physicians commenting on our device
@UNC Medical Center, Chapel Hill NC
We also talked to competitors to understand their products
@AAEM conference
Understanding complex clinical workflows across diverse hospital systems is no easy task. To ensure our device integrates seamlessly into existing workflows, we visited 6+ hospitals nationwide, conducting interviews with healthcare professionals, IT departments, and shadowing staff for full-day observations.Additionally, we attended major conferences — including AAEM, ABA, SRBC, and EMS — to gain deeper insights into market trends and user needs.

Data analysis

By synthesizing insights from user interviews and other sources — including internal engineers, user operation data, and historical unresolved issues — I created an affinity map to identify key areas for improvement in the next iteration.

This affinity map served as a critical input for defining user needs and shaping the project roadmap.

Every pink/orange stickers means a category. Like Visual Identity, Typing Experience Improvement, etc.

Redefine workflow

We visualized the current clinical workflow and burn wound treatment methods to ensure our system integrates seamlessly with existing processes. Our goal was to make the device a valuable tool, not an additional burden for nurses.
Key Improvements:
‍Reducing Repetitive Data Entry: User research revealed that time is precious in hospitals, and repeated data entry is frustrating. We now integrate with electronic health records (EHR) to automatically pull patient data.
‍Tailored for Emergency Scenarios: In emergency departments and mass casualty situations, Total Burn Surface Area (TBSA%) is critical for patient resuscitation. We prioritized making this information readily accessible for specific scenarios and user groups.
‍Fast Access in Urgent Situations: For time-sensitive cases, unnecessary paperwork can be skipped, allowing healthcare professionals to grab the device and capture wound images immediately.
Current burn management workflow
Software UI Architecture

Wound Profile

• Wound records in a glance
• Track wound size and healing probability change over time
• Quick start scan the current wound

PDF Report

• Detailed report for reference
• Can be took home with patients and shared with their families
• Available in both digital and paper version. Record shared between EHRs.

Gallery


• Display AI prediction results to help understand wound healing probability
• Organized by wounds for fast locating
• Show wound progression over the time

Field research

DFU is a new market for our company. We have experience working with hospital burn units and ERs, but for DFU, patients usually go to specialized DFU clinics periodically. Therefore the workflow is very different from our first-generation software for burn wounds.
For DFU wounds, physicians are more concerned with how the wound changes for their treatment, while patients are more concerned with how long it takes to heal. Therefore, we focus more on assisting physicians in obtaining information and making decisions and helping physicians provide patients with the confidence to cooperate with long-term treatment.

Outcome

I worked with the clinical team and usability experts and designed a digital experience that highlights the long-term wound care and assistance of clinical decisions. We ran validation tests with clinicians, nurses, and surgeons in compliance with FDA guidelines. Our goal was to ensure that the process of wound assessment of DeepView was crystal-clear and error-proof for clinicians.

Meeting
Web app

Aldo

Benevv

Social
Giving
Mobile App

Voice Assistant for Automobile

Voice User Iterface

CoC
Website

Georgia Tech Website Redeign

To the top