(1) Background: In the present study, we evaluated the efficacy of a 3D-printed, patient-specific polycaprolactone/beta tricalcium phosphate (PCL/β-TCP) scaffold in the treatment of complex zygomatico-maxillary defects. (2) Methods: We evaluated eight patients who underwent immediate or delayed maxillary reconstruction with patient-specific PCL implants between December 2019 and June 2021. The efficacy of these techniques was assessed using the volume and density analysis of computed tomography data obtained before surgery and six months after surgery. (3) Results: Patients underwent maxillary reconstruction with the 3D-printed PCL/β-TCP scaffold based on various reconstructive techniques, including bone graft, fasciocutaneous free flaps, and fat graft. In the volume analysis, satisfactory volume conformity was achieved between the preoperative simulation and actual implant volume with a mean volume conformity of 79.71%, ranging from 70.89% to 86.31%. The ratio of de novo bone formation to total implant volume (bone volume fraction) was satisfactory with a mean bone fraction volume of 23.34%, ranging from 7.81% to 66.21%. Mean tissue density in the region of interest was 188.84 HU, ranging from 151.48 HU to 291.74 HU. (4) Conclusions: The combined use of the PCL/β-TCP scaffold with virtual surgical simulation and 3D printing techniques may replace traditional non-absorbable implants in the future owing to its accuracy and biocompatible properties.

Aim: To evaluate the clinical outcomes of double-strut fibula onlay bone graft combined with a vascularized fibula free flap for mandibular reconstruction, focusing on long-term graft volume changes and viability. Materials and Methods: This retrospective review included 9 patients (7 males, 2 females) who underwent simultaneous reconstruction with a fibula free flap and fibula onlay bone graft between 2017 and 2023. The mean follow-up period was 2.19±1.69 years. Clinical and radiologic data were collected to assess bone volume, speech, dietary function, and postoperative complications. Volumetric analysis was conducted using 3D imaging, and the Mann-Whitney U test was used to compare bone volume changes between the graft and flap groups. Results: The mean fibula onlay graft length was 5.9±1.7 cm (range: 2.8 to 7.6 cm). Postoperative imaging showed higher bone volumes in the fibula flap group at all time points. The onlay graft group exhibited a higher median percentage bone volume reduction (43.48%) compared with the fibula flap group (22.37%), although this difference did not reach statistical significance ( P =0.136), and the series was too small to exclude a clinically meaningful difference. Bone union was achieved in 88.9% of cases, with 1 graft failure (11.1%) and 1 wound infection (11.1%). Conclusion: In this small, uncontrolled series, the combination of a periosteal-preserving fibula onlay bone graft with a vascularized fibula free flap was feasible and achieved vertical augmentation with acceptable graft viability and a low complication rate. Larger controlled studies are required before the technique can be recommended over established alternatives.

Computed tomography angiography (CTA)-based segmentation of the superficial temporal arteries (STAs) and facial vessels (FVs) is important for neurosurgical and reconstructive planning. Nevertheless, segmentation of STAs and FVs remains challenging because of their small caliber, tortuous courses, and proximity to high-intensity bony structures. This study aims to develop a deep learning framework for accurate automated segmentation of these craniofacial vessels. A single-input 3D nnU-Net v2 model was trained using raw CTA volumes, while a Fusion-based Vesselness Map (FVM) was constructed from multiscale vessel-enhancement filters to emphasize small vascular structures and suppress irrelevant regions such as the skull and skin. Instead of being used as an additional input channel, the FVM was incorporated into the loss function as a spatial prior to guide the network toward vessel boundaries and distal branches. In 72 clinical cases, the FVM-guided model improved segmentation accuracy compared with a baseline model trained with Dice Focal Loss, particularly in boundary delineation. For the STAs, the Average Symmetric Surface Distance decreased from 6.543 mm to 2.941 mm. Qualitative evaluation further showed reduced segmentation noise and fewer false positives near bone and distal branches. These findings suggest that integrating classical vessel enhancement into deep learning supervision can improve morphologically consistent craniofacial vessel segmentation and support preoperative surgical planning.