BackgroundThe percutaneous transforaminal endoscopic lumbar interbody fusion (PETLIF) has many advantages as a new minimally invasive surgical technique, and its surgical approach passes through the Kambin's triangle. One of the greatest challenges in completing PETLIF is avoiding nerve root damage. In previous studies, the relevant anatomic data do not correspond well with current surgical techniques, and there is a paucity of studies based on magnetic resonance neurography (MRN), which is the clearest imaging method for nerve roots. The purpose of this study was to analyze the safety of the PETLIF at each lumbar segment based on measured results from the MRN imaging data and to propose a novel method of intraoperative positioning.MethodsThe coronal images with the clearest course of nerve roots were chosen for retrospective observation. During the PETLIF, the secure region of the operation was considered to be a trapezium. The following parameters were measured, respectively: trapezoid area, height, and median line length, as well as the relevant parameters of the positional relation between the point “O,” the most secure operating center point of the secure region, and each osseous anatomic landmark. And the data were compared with the size of the cage to obtain safety.ResultsAt L1-S1, with the downward motion of the target intervertebral space, the area increased from (67.94 ± 15.22) mm2 to (140.99 ± 26.06) mm2, and the height increased from (7.23 ± 1.17) mm to (12.59 ± 1.63) mm. At L1–L5, the length of the median line was increased from (9.42 ± 1.70) mm to (12.70 ± 1.88) mm. Even though it was reduced to (11.59 ± 1.99) mm at L5–S1, it was still longer than that at L3–L4. The safety obtained by the primary observational indicator was 34.52%, 33.33%, 53.57%, 96.43%, and 77.38%, respectively, at L1–S1. The safety obtained by the two secondary observational indicators was 77.38% and 95.24% at L3–L4 and 100% at L4–S1. There was no point “O” outside the anatomic mark line. The intraoperative positioning method of the point “O” was as follows: It was located medially and horizontally approximately 3/5 of the anatomic mark line at L1–L5; the horizontal distances were (0.48 ± 0.67) mm, (1.20 ± 0.89) mm, (2.72 ± 1.01) mm, and (3.69 ± 1.47) mm, respectively. In addition, it was necessary to locate (3.43 ± 1.41) mm inward at about 4/5 of the anatomic mark line at L5–S1.ConclusionsThe MRN allows clearer and more accurate visualization of the nerve roots, and the basic anatomic study of the Kambin's triangle based on this technology is of practical clinical significance. In the current study, it is believed that, during the PETLIF, cage implantation is the safest at L4–L5, followed by L5–S1; L1–L3 is more likely to cause nerve root injury, and L3–L4 is not less likely. To improve safety, a comprehensive individualized imaging assessment should be performed before surgery. This study also provides an easy method of intraoperative localization, which helps avoid nerve root injury.