Marketplace - B2B Radiology AI Platform

B2B Platform Medical Imaging Clinical Workflow SaaS Workflow Optimization

Marketplace is a web-based B2B radiology platform that connects imaging facilities and radiologists. Imaging facilities upload and track medical imaging exams (such as CT scans, MRIs, and X-rays), while radiologists review the exams and provide diagnostic reports.

Project Scope

  • Led the design of two new user profiles (Clinician and Screening Provider) within the Marketplace platform to support emerging clinical use cases and enable seamless integration of a radiology AI service into existing clinical workflows.
Research
UX Design
Wireframes
UI Design
Style Guide
Testing
Clinical Management
Operation Management
Product Managers
Engineers Team
01

Discover

Understand the business goals, user needs, existing product, and key challenges through research and analysis.

02

Define

Translate insights into clear problems, requirements, and opportunities that establish the design direction.

03

Explore

Develop and evaluate different ideas, user flows, and interaction patterns to identify the most effective solution.

04

Design

Turn concepts into intuitive experiences through wireframes, prototypes, visual design, and design systems.

05

Validate

Test the solution with users and stakeholders, identify issues, and iterate to improve usability and effectiveness.

06

Hand Off

Prepare designs, specs, and docs, and collaborate with developers for accurate implementation.


The Main Goals

  • Support simplified clinical use cases compared to traditional radiology workflows.
  • Reduce complexity by minimizing steps, statuses, and information overload.
  • Simplify the submission and assignment process for medical imaging studies.
  • Adapt actions and tools to each user type and workflow.
  • Integrate AI/algorithmic analysis services as an additional diagnostic offering within the radiology workflow.
  • Ensure patient reports reflect validated radiologist decisions accurately.
  • Enable radiologist validation of AI results where required.

Identify the Problems and Suggest Solutions

Research sessions with stakeholders, review of operational workflows, and clinical requirements analysis revealed several areas requiring adaptation to support the new use cases.

Problem
1

The existing Marketplace experience was built around radiology workflows and terminology. The new user groups (Clinicians and Screening Provider staff) have different objectives and expectations, making many existing tools and actions irrelevant to their needs.

Solution
1

Designed two dedicated user profiles with tailored workflows, actions, and navigation patterns, ensuring each user type only sees the tools and information relevant to their tasks.

Problem
2

The traditional workflow contained unnecessary complexity for the new use cases, included multiple steps, statuses, and operational actions required for collaboration between Imaging Facilities and Radiologists.

Solution
2

Redesigned the workflow architecture by reducing statuses, simplifying process steps, and removing actions that were not applicable to the new user journeys.

Problem
3

The platform contained clinical and operational information designed for radiologists. Presenting the same level of detail to clinicians and screening provider users could make key information harder to find and understand.

Solution
3

Prioritized relevant information for each profile, simplifying screens and displaying only the data required to complete the user's task at each stage of the workflow.

Problem
4

Supporting the new AI service offering required capabilities that did not exist in the Marketplace platform. The platform lacked a dedicated process for radiologist validation and an automated way to generate patient-facing reports from approved results.

Solution
4

Introduced a radiologist validation feature and a report generation workflow that transforms validated results into a structured patient report.

Use cases

The clinician uploads the patient’s exam to the Marketplace and reviews the AI results during the consultation, using them to support clinical evaluation. Results are delivered directly to the clinician, without radiologist involvement.


Final Design

Exams Grid

The two-panel layout keeps the exam list visible while reviewing an exam, reducing navigation and preserving context. The content and actions adapt to each user’s role and workflow, keeping the interface focused and relevant.

Exam Upload

To upload an exam, the user selects a folder from their device. The system automatically detects and lists the studies, evaluating each one against the configured AI algorithms.

Patient Report validation

Patient Report validation consists of verifying and aligning the radiologist’s report written in PACS with the system-generated version to ensure consistency between both.


Prototypes


Reflection & Impact

The metrics below reflect the first six months following launch, compared against the platform's performance prior to the redesign.

100%

Faster image upload

0%

Increase in AI service adoption

100%

Reduction in manual validation effort

0+

Exams processed through AI workflows

0+

Customers actively using AI workflows

60 min

Saved per AI validation

Design System

Colors

Primary

Black0D2024

Blue4E59A4

White—

Secondary

Light violetC2C5DD

Light grayd8d8d8

Medium gray757575

Light greenD5F0D4

light blueF1F6FF

OrangeFF7201

YellowFFD200

Green1A932E

Red ValidationD24B4B

Icons

Buttons