Performance of the Malmgren Index for Assessing Root Resorption on 2D vs. 3D Radiographs: A Pilot Study.

Hanne MichielsensJulie DecreusGiacomo BegnoniAnna VerdonckReinhilde JacobsGuy WillemsMaria Cadenas de Llano-Pérula

Published in: Healthcare (Basel, Switzerland) (2023)

The original Malmgren index is not suited for 3D images, especially axial, where using dichotomized values (resorption yes/no) leads to overestimation of RR. A low-dose CBCT of the upper incisors could detect RR with high diagnostic accuracy in the early stages of orthodontic treatment, especially in patients with dental trauma or familial RR history.

Keyphrases

low dose
deep learning
high dose
convolutional neural network
early onset
bone loss
optical coherence tomography
oral health
machine learning
image quality
magnetic resonance
computed tomography