Evaluating the accuracy of methods used for reverse engineering hand geometries for 3D printed splinting
Abstract
The COVID-19 pandemic has significantly accelerated the adoption of telehealth and remote rehabilitation methods, necessitating innovative solutions for medical treatments that traditionally require in-person interactions. This thesis investigates the accuracy of different reverse engineering techniques used to create 3D printed splints for hand therapy, focusing on the comparison between photogrammetry and mobile scanning methods. The study employs a custom-built photogrammetry scanner using 42 Raspberry Pi cameras and a mobile phone setup to capture hand geometries. Both methods' workflows are meticulously developed and assessed for their ability to produce precise 3D models.
The accuracy of the scanning methods is evaluated using uniform and non-uniform objects, including a wooden cube and 3D printed models of a hand, cat, and ship. Surface deviation mapping is employed to compare the scanned models with their original digital counterparts. The thesis also compares the accuracy and fit of 3D printed splints generated from these scans with traditionally made thermoplastic splints crafted by therapists.
Results indicate that the photogrammetry scanner provides higher accuracy and consistency than the mobile scanning method. The 3D printed splints, generated using specialized CAD-based software, exhibit significantly lower deviation and better fit compared to traditional splints. This research highlights the potential of 3D printing technology in orthotic fabrication, offering improved patient outcomes and streamlined processes for remote hand therapy.
The findings underscore the viability of photogrammetry and mobile scanning methods in telehealth applications, suggesting that while mobile scanning offers convenience, further enhancements are needed to match the precision of photogrammetry.