Rheumatology & AI
ARTIX: AI for Raynaud's Phenomenon Quantification
This study introduces ARTIX, an AI-driven mobile phone-based tool designed to objectively quantify Raynaud's phenomenon (RP) severity. It aims to provide a patient-centered, image-based assessment method, complementing traditional self-reported measures.
Di Battista et al. Arthritis Research & Therapy (2025) 27:120
Executive Impact: Revolutionizing Remote RP Monitoring
ARTIX leverages mobile phone photography and machine learning to offer a novel, objective method for assessing Raynaud's phenomenon. This technology promises enhanced patient self-monitoring and improved clinical evaluation through readily available consumer devices, democratizing access to advanced diagnostic support.
0
Sensitivity in RP Detection
0
Peak AUC in Cold Challenge
0
RP Patients Returning to Baseline ARTIX
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
AI Algorithm Development
Validation & Performance
Clinical Implications
AI Algorithm Development
ARTIX employs a multi-step computational pipeline for hand image analysis, including UNet-based hand segmentation, MediaPipe landmark detection, and custom colour analysis algorithms. This ensures robust and accurate quantification of finger redness, a key indicator for Raynaud's phenomenon.
Validation & Performance
Validated against healthy controls and thermography, ARTIX consistently showed significant differences between RP patients and controls. It demonstrated high sensitivity (over 90%) and strong AUC values (up to 0.983) in discriminating RP, particularly 2 minutes post-cold challenge, and maintained performance across seasonal variations.
Clinical Implications
ARTIX offers a non-invasive, accessible tool for objective RP quantification. Its ability to perform well without a 'cold stimulation' step makes it suitable for home use, potentially transforming patient self-assessment and remote monitoring, thereby complementing clinical evaluation and patient-reported outcomes.
0.983
Peak AUC for ARTIX (May-July, 2 min post-challenge)
ARTIX achieved its highest discriminative accuracy (AUC 0.983) during the warmer months (May to July), specifically 2 minutes after cold challenge. This indicates exceptional performance in detecting the rewarming process, a critical phase in Raynaud's phenomenon.
Enterprise Process Flow
Original Image Capture
→
Hand Segmentation (UNet)
→
Key Points Identification (MediaPipe)
→
Finger Segmentation
→
Red Channel Analysis
→
Gaussian Filtering
→
ARTIX Score Calculation
| Feature | ARTIX (AI-based Mobile Tool) | Thermography (Traditional Gold Standard) |
|---|---|---|
| Technology | Mobile phone photography, Machine Learning | Specialized thermal camera |
| Accessibility | High (smartphone required) | Limited (specialized equipment, clinical setting) |
| Home Monitoring | Highly feasible, patient-centered | Not feasible for remote, frequent use |
| Performance (RP vs. HC) | Consistent significant difference across all timepoints, even in colder months | Occasional failure to detect difference at early timepoints in colder months |
| Data Type | Color intensity distribution (redness) | Surface temperature (°C) |
| Cost | Low (leveraging existing devices) | High (equipment purchase & maintenance) |
Impact of ARTIX on Patient Monitoring
In a cohort of 45 RP patients, ARTIX revealed that 71.1% returned to baseline values within 10 minutes post-cold challenge, significantly outperforming thermography (37.8% return rate). This demonstrates ARTIX's potential to provide more nuanced and consistent insights into vascular reactivity, enabling better longitudinal tracking of individual patient responses to cold and treatment efficacy. Patients on vasoactive therapy showed higher ARTIX values, indicating ARTIX's ability to reflect clinical status and treatment impact.
Calculate Your Potential ROI with AI Automation
Estimate the significant time and cost savings your enterprise could achieve by implementing AI solutions in your operational workflows.
Estimated Annual Cost Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A structured approach to integrating ARTIX and similar AI solutions into your healthcare or research operations.
Phase 1: Discovery & Strategy
Initial consultation to understand your specific needs, data landscape, and define clear objectives for AI integration. This includes assessing current Raynaud's monitoring practices and identifying key performance indicators for ARTIX.
Phase 2: Pilot & Customization
Deployment of ARTIX in a controlled pilot environment. Customization of the algorithm and user interface to fit your specific patient cohorts, mobile devices, and reporting requirements, ensuring seamless integration with existing systems.
Phase 3: Validation & Training
Conducting internal validation studies to confirm ARTIX's performance within your specific context. Comprehensive training for clinicians and patients on using the mobile application and interpreting ARTIX scores.
Phase 4: Full-Scale Deployment & Monitoring
Roll-out of ARTIX across your target patient population. Continuous monitoring of performance, user feedback, and clinical outcomes. Iterative improvements based on real-world data and emerging research.
Phase 5: Advanced Integration & Expansion
Exploring integration with electronic health records (EHR) and other digital health platforms. Potential expansion of ARTIX capabilities to include other dermatological or vascular conditions, leveraging the core AI infrastructure.
Ready to Transform Your Enterprise with AI?
Connect with our AI specialists to explore how ARTIX or other custom AI solutions can be tailored to meet your organization's unique challenges and opportunities.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat