Ingeniería Biomédica · 2026
Phantom craneal pediátrico para la identificación de fracturas y suturas por ultrasonido
Solo ficha. El repositorio marca este trabajo como «Acceso abierto»: el PDF no es público, así que aquí no hay texto para leer. Ver la ficha en el repositorio.
Resumen
The identification of cranial fractures and sutures in pediatric patients by ultrasound is a technique that is rarely used due to the constant changes occurring in the anatomy of the developing skull. Ultrasound is presented as a safe and accessible alternative in contrast with the widely used computed tomography (CT); however, its adoption has been limited by the lack of practice in identifying anatomical structures and the absence of adequate simulators that allow training for healthcare professionals. For this reason, this work aimed to develop a pediatric cranial phantom that is durable and resistant to degradation, allowing the simulation of structures such as sutures and fractures for their identification by ultrasound. A 3D model was designed based on a neonatal skull, into which simulated sutures and fractures were incorporated. Through physical–acoustic tests, different material mixtures were evaluated, leading to the selection of silicone for soft tissue and type IV extra-hard plaster combined with additives for bone tissue. Three prototypes were manufactured and validated through ultrasound imaging with the assistance of a physician. The final model made it possible to obtain images comparable to real cases, including records from The POCUS Atlas, a living pediatric patient, and a pediatric cadaver. Its usefulness as an educational tool was demonstrated through a computational analysis in MATLAB, which produced acceptable structural similarity values. The sagittal suture reached 97%, the coronal 95%, and the metopic 76%, indicating that their trajectories and characteristic patterns were clearly reproduced in the phantom. The lambdoid suture presented a value of 64%, a figure consistent with the variability this suture usually shows in real images. In the case of fractures, the non-displaced linear fracture obtained 79% geometric similarity and the depressed fracture 73%, reflecting that their contours and directions can be recognized without difficulty. This demonstrates an approach in similarity in the functional representation of relevant anatomical structures of the phantom for ultrasound training.