Three-Dimensional Wound Measurement: A Low-Budget Approach using Smartphones
3D Reconstruction, Chronic Wounds, Structure from Motion (SfM),Deep Neural Networks, Smartphones, Computer Vision, Accessible Healthcare Technologies
Chronic wounds are a significant public health challenge, impacting the quality of life of millions of people and requiring accurate measurements for effective monitoring and treatment. Traditional methods, such as manual measurements, often fail on uneven surfaces, while modern 3D solutions, although more accurate, rely on expensive equipment or external servers, limiting their accessibility. This work proposes a cost-effective approach for 3D reconstruction and measurement of wounds using only a smartphone. Combining Structure from Motion (SfM) and deep neural networks for depth estimation, the system aims to provide accurate measurements without the need for specialized hardware or remote processing. The objectives include the optimization of the SfM pipeline for smartphones, the evaluation of the accuracy of monocular depth maps, and the comparison of both methodologies in terms of accuracy, execution time, and computational feasibility. Clinical and nonclinical validations demonstrate the robustness of the system in various scenarios and imaging conditions. Previous work points to a measurement error below 5\%, however, it requires reasonable computational power. By eliminating dependencies on expensive equipment and external infrastructure, the proposal democratizes access to advanced wound assessment technologies, making them viable in resource-limited settings and remote regions.