Unilateral Biportal Endoscopy (UBE) represents a novel, minimally invasive surgical approach in the management of lumbar spinal stenosis (LSS). It offers distinct advantages over conventional open surgery, including reduced tissue trauma, postoperative pain, and recovery time. However, despite its benefits, UBE is associated with rare complications such as dural tears, nerve root injuries, and pseudocyst formation. We report the case of a 73-year-old female who presented with severe radicular pain and reduced muscle strength in the left lower extremity following UBE for lumbar stenosis. Postoperatively, the patient experienced recurrent radicular pain. Magnetic resonance imaging (MRI) revealed the presence of a pseudocyst in the left lateral recess. Serum biomarkers showed no evidence of infection. Surgical exploration identified a cystic lesion containing air and serous fluid. The patient underwent open surgical laminectomy and pseudocyst drainage. Histopathological analysis of the lesion confirmed the presence of fibrinous connective tissue. Although pseudocyst formation is an infrequent complication of UBE, it can lead to significant postoperative morbidity. Management involves surgical intervention and detailed postoperative monitoring. Both UBE and open surgical techniques can be adopted considering its unique advantages in addressing certain postoperative complications. In this case, open surgery was selected to ensure adequate visualization and access to the lesion. Even though a more invasive approach was chosen following the intervention, the patient demonstrated marked clinical improvement post operative day 1, with resolution of the cystic lesion and associated symptoms. This case underscores the importance of considering pseudocyst formation in the differential diagnosis of postoperative complications following UBE. Prompt recognition and timely surgical management are essential to achieving favorable patient outcomes.

Background: Artificial intelligence (AI) has rapidly emerged as a major research domain in orthopaedics, driven by advances in computational methods and the widespread availability of digital clinical data. Although AI-related orthopaedic publications have increased substantially, the extent to which this growth reflects a balanced distribution of research activity across clinical subspecialties and AI modalities remains unclear. Purpose: The aim of this study was to evaluate temporal trends, subspecialty distribution, and methodological characteristics of AI-related orthopaedic research activity using a bibliometric approach. Methods: A bibliometric analysis was performed using publications indexed in two major scientific databases between 2016 and 2025. AI-related orthopaedic articles were identified using predefined search terms. Publications were analyzed according to annual publication trends, orthopaedic subspecialty classification, AI methodology, and proxy indicators of study design and evidence development. Subspecialty assignment was performed using a standardized taxonomy, and descriptive analyses were used to characterize research patterns. Results: AI-related orthopaedic publications increased exponentially during the study period. Research output was unevenly distributed across subspecialties, with adult reconstruction and spine surgery accounting for the largest share of publications, whereas foot and ankle surgery, pediatric orthopaedics, hand & wrist, and shoulder & elbow surgery remained underrepresented. Deep learning–based imaging studies predominated, while studies incorporating structured clinical data, longitudinal outcomes, and external validation were comparatively limited. Much of the literature remains exploratory, primarily focusing on early-phase model development. Conclusions: AI research in orthopaedics has expanded rapidly but remains methodologically heterogeneous, with research activity heavily concentrated in data-rich subspecialties such as adult reconstruction and spine surgery. The predominance of early-phase studies indicates that current publication volumes are largely driven by data accessibility rather than a balanced distribution across all orthopaedic domains. Future research should encourage multicenter collaboration, standardized reporting, and targeted studies in underrepresented subspecialties to ensure a more balanced development of orthopaedic AI tools.