Correction to "Automatic Quality Assessment of Echocardiograms Using Convolutional Neural Networks: Feasibility on the Apical Four-Chamber View".

In the above paper [1], the first footnote should have indicated the following information: A. H. Abdi and C. Luong are joint first authors.

Registration of serial maxillary models via the weighted rugae superimposition method.

OBJECTIVES: We introduce a weighted method for superimposition of serial digital maxillary models based on the variable stability of rugae points. SETTING AND SAMPLE POPULATION: Plaster maxillary models of 24 randomly selected 12-year-olds as well as their models at the age of 14 were obtained and scanned using a benchtop structured-light 3D scanner. METHODS: The models were registered twice, once via the unweighted and again via the proposed weighted rugae superimposition method based on 12 rugae landmarks. For each superimposition, distances between the corresponding rugae points were measured and compared with reported displacements of rugae points in literature. RESULTS: The unweighted superimposition produced no meaningful differences in terms of total displacements of registration landmarks, whereas the weighted method recognized the medial points of the third ruga as the most stable landmarks. Results of the weighted method also demonstrated statistically significant smaller changes for medial rugae points in almost every dimension compared to the lateral rugae points. These results comply with the growth patterns of maxilla and rugae point displacements reported in similar studies. CONCLUSION: Considering the variable stability of rugae points during growth, the weighted rugae superimposition method results in more promising registrations on serial models. This method prioritizes registration landmarks based on clinical criteria of choice and is suitable for analysis of other structures such as tooth movements.