A Shape-Aware Total Body Photography System for In-Focus Surface Coverage Optimization

Area of Science:

• Medical Imaging
• Dermatology
• Computer Vision

Background:

• Total Body Photography (TBP) is crucial for skin cancer screening in high-risk patients.
• Current TBP systems require enhancement in automatic lesion detection and analysis.
• Image resolution and sharpness are key limitations in existing TBP technologies.

Purpose of the Study:

• To propose a novel shape-aware TBP system for high-fidelity full-body image capture.
• To optimize image quality, focusing on resolution and sharpness across the entire body surface.
• To improve automated analysis of skin lesions for enhanced skin cancer screening.

Main Methods:

• A 360-degree rotary beam system with depth and RGB cameras.
• 3D body shape estimation to understand body geometry.
• An in-focus surface optimization method to dynamically adjust focus distance.
• Calibrated camera poses to ensure optimal focus over complex 3D surfaces.

Main Results:

• The system achieves high resolution (0.068 mm/pixel and 0.0566 mm/pixel).
• Approximately 85-95% of the body surface area is captured in-focus.
• The shape-aware focus method surpasses traditional auto-focus protocols.
• Demonstrated effectiveness on diverse body shapes, poses, and mannequin scans.

Conclusions:

• The proposed shape-aware TBP system significantly enhances image fidelity for full-body imaging.
• High-quality imaging is expected to advance automated skin lesion analysis.
• This technology holds promise for improving early skin cancer detection rates.