Low-Field MRI for Dental Imaging in Pediatric Patients With Supernumerary and Ectopic Teeth: A Comparative Study of 0.55 T and Ultra-Low-Dose CT
Ines Willershausen,Stefania Evangeliou,Hans-Peter Fautz,Patrick Amarteifio,Matthias Stefan May,Armin Stroebel,Martin Zeilinger,Michael Uder,Lina Goelz,Markus Kopp
DOI: https://doi.org/10.1097/RLI.0000000000001129
2024-10-23
Abstract:Objectives: This study sought to elucidate the diagnostic performance of 0.55 T magnetic resonance imaging (MRI) for pediatric dental imaging, specifically in terms of the image quality (IQ) for detecting ectopic and/or supernumerary teeth, compared with routine ultra-low-dose computed tomography (ULD-CT) of the jaw. Materials and methods: A total of 16 pediatric patients (mean age: 12.4 ± 2.6 years, range: 9-17 years) with ectopic and/or supernumerary teeth screened from January 2023 to January 2024 were enrolled in this prospective, single-center study. All patients underwent ULD-CT as the clinical reference standard and 0.55 T MRI as the study scan on the same day. A 0.6-mm isotropic 3-dimensional T1w FLASH sequence was developed with a dedicated field of view of the upper and lower jaws. ULD-CT was performed using a new single-source computed tomography (CT) scanner equipped with a tin filter (Sn100, slice thickness: 1 mm, quality reference mAs: 24). The IQ for the tooth axis, the tooth length, the tooth root, root resorptions, cysts, the periodontal ligament space, and the mandibular canal was evaluated twice by 3 senior readers using a 5-point Likert scale (LS) (LS score of 1: insufficient, 3: reduced IQ but sufficient for clinical use, and 5: perfect) and compared between both methods. Subsequently, the results were dichotomized into nonvalid (LS score of ≤2) and valid (LS score of ≥3) for clinical use. Results: A total of 49 ectopic and/or supernumerary teeth in 16 pediatric patients were investigated using ULD-CT (CTDI: 0.43 ± 0.09 mGy) and 0.55 T MRI. The mean MRI acquisition time was 9:45 minutes. Motion artifacts were nonsignificantly different between 0.55 T MRI and ULD-CT (P = 0.126). The IQ for the tooth axis, the tooth root, root resorptions, and cysts was similar between the methods. The IQ for the periodontal ligament space and tooth length favored ULD-CT by 14% (confidence interval [CI]: 4.3%-24%) and 7.5% (CI: 1.8%-13%), respectively, whereas that for the mandibular canal favored 0.55 T MRI by -35% (CI: -54%-16%). Sufficient IQ was found especially for cystic lesions (CT: 100% sufficient, MRI: 95% sufficient), the tooth root (CT: 100%, MRI: 98%), root resorptions (CT: 94%; MRI: 85%), the tooth axis (CT: 100%; MRI: 98%), and the tooth length (CT: 99%; MRI: 91%). Conclusions: The findings indicate that 0.55 T MRI is a feasible, radiation-free technique for delineating ectopic and/or supernumerary teeth in pediatric patients. Nevertheless, to date, 0.55 T MRI has not yet been able to provide an optimal IQ for all anatomical tooth and jaw structures. In cases of advanced clinical indications that require optimal spatial resolution, high-resolution CT or cone-beam CT may still be necessary.