Forensic 3D avatars - AI-assisted injury detection and interactive digital-twin visualization

Area of Science:

• Forensic Medicine
• Medical Imaging
• Artificial Intelligence

Background:

• Forensic documentation of injuries is crucial for legal proceedings.
• Current methods can be time-consuming and subjective.
• Advancements in 3D scanning and AI offer potential for improved objectivity and efficiency.

Purpose of the Study:

• To assess the feasibility of using a dermatology-based 3D full-body scanning system (VECTRA WB360) with AI for forensic documentation.
• To evaluate the system's capability in automated detection, localization, and measurement of simulated injuries.
• To explore the potential benefits and limitations of this technology in forensic medicine.

Main Methods:

• Utilized the VECTRA WB360 3D full-body scanner to create high-resolution, color-accurate 3D avatars.
• Employed artificial intelligence (AI) software for automated analysis of simulated cutaneous findings.
• Assessed the system's performance in identifying and localizing injuries, including in challenging anatomical areas.
• Developed an interactive forensic dashboard for data visualization.

Main Results:

• The 3D scanning system rapidly generated detailed 3D avatars.
• AI-assisted analysis successfully detected and localized simulated injuries with high accuracy.
• The system demonstrated effectiveness even in anatomically complex regions.
• Identified advantages such as speed, standardization, and precise anatomical localization.

Conclusions:

• The integrated 3D scanning and AI system shows promise for enhancing forensic documentation efficiency and objectivity.
• Limitations including dermatology-focused training data, motion artifacts, and cost require further investigation.
• Future development of forensic-specific AI modules and validation are necessary for clinical and legal adoption.