Confined to Kambin's prism: MRI-based segmentation and prediction of the need for superior articular process removal...

The endoscopic transforaminal lumbar approach offers reduced tissue disruption and faster recovery compared with other open and minimally invasive approaches but is limited by the narrow and variable anatomy of the transforaminal corridor. Degenerative changes such as superior articular process (SAP) hypertrophy can block safe access, forcing the surgeon to remove part of the SAP (foraminoplasty) either by reaming or drilling. This decision is often made subjectively or discovered intraoperatively. Recent advances in MRI-based nerve segmentation may provide an objective method to predict this need preoperatively.

The records of 23 patients who underwent endoscopic decompression and had high-resolution (≤ 2.4 mm) preoperative lumbar MR images were reviewed. The authors segmented neural and osseous structures using commercial planning software (Brainlab) and then placed a 6-mm virtual dilator along a simulated transforaminal trajectory to assess corridor patency. Operative reports and video confirmed whether SAP reaming was performed. Pre- and postoperative patient-reported outcomes were analyzed.

Segmentation showed an unobstructed 6-mm corridor in 22 of 23 patients (96%). The single patient with a predicted obstruction had SAP hypertrophy and required reaming. The simulation and the operative report agreed in every case. On average, patients improved significantly in every recorded clinical outcome within 2 years postoperatively. The size of the surgical corridor had a mild positive relationship to positive surgical outcomes, although not significant. The single patient with reaming had greater magnitude of clinical score improvement compared with the other patients for physical health, Oswestry Disability Index, and leg and back pain, although the difference was not significant.

MRI-based segmentation accurately identified foraminal anatomy that required foraminoplasty for access. This workflow turns a subjective intraoperative judgment into a reproducible, patient-specific geometrical assessment.