Exploiting the interplay between cross-sectional and longitudinal data in Class III malocclusion patients

Cross-sectional data accurately model longitudinal growth in the craniofacial skeleton

Dense, longitudinal sampling represents the ideal for studying biological growth. However, longitudinal samples are not typically possible, due to limits of time, prohibitive cost, or health concerns of repeat radiologic imaging. In contrast, cross-sectional samples have few such drawbacks, but it is not known how well estimates of growth milestones can be obtained from cross-sectional samples. The Craniofacial Growth Consortium Study (CGCS) contains longitudinal growth data for approximately 2000 individuals. Single samples from the CGCS for individuals representing cross-sectional data were used to test the ability to predict growth parameters in linear trait measurements separately by sex. Testing across a range of cross-sectional sample sizes from 5 to the full sample, we found that means from repeated samples were able to approximate growth rates determined from the full longitudinal CGCS sample, with mean absolute differences below 1 mm at cross-sectional sample sizes greater than ~ 200 individuals. Our results show that growth parameters and milestones can be accurately estimated from cross-sectional data compared to population-level estimates from complete longitudinal data, underscoring the utility of such datasets in growth modeling. This method can be applied to other forms of growth (e.g., stature) and to cases in which repeated radiographs are not feasible (e.g., cone-beam CT).

The Validity of Machine Learning Procedures in Orthodontics: What Is Still Missing?

Artificial intelligence (AI) models and procedures hold remarkable predictive efficiency in the medical domain through their ability to discover hidden, non-obvious clinical patterns in data. However, due to the sparsity, noise, and time-dependency of medical data, AI procedures are raising unprecedented issues related to the mismatch between doctors’ mentalreasoning and the statistical answers provided by algorithms. Electronic systems can reproduce or even amplify noise hidden in the data, especially when the diagnosis of the subjects in the training data set is inaccurate or incomplete. In this paper we describe the conditions that need to be met for AI instruments to be truly useful in the orthodontic domain. We report some examples of computational procedures that are capable of extracting orthodontic knowledge through ever deeper patient representation. To have confidence in these procedures, orthodontic practitioners should recognize the benefits, shortcomings, and unintended consequences of AI models, as algorithms that learn from human decisions likewise learn mistakes and biases.

Complexity and data mining in dental research: A network medicine perspective on interceptive orthodontics

Procedures and models of computerized data analysis are becoming researchers' and practitioners' thinking partners by transforming the reasoning underlying biomedicine. Complexity theory, Network analysis and Artificial Intelligence are already approaching this discipline, intending to provide support for patient's diagnosis, prognosis and treatments. At the same time, due to the sparsity, noisiness and time-dependency of medical data, such procedures are raising many unprecedented problems related to the mismatch between the human mind's reasoning and the outputs of computational models. Thanks to these computational, non-anthropocentric models, a patient's clinical situation can be elucidated in the orthodontic discipline, and the growth outcome can be approximated. However, to have confidence in these procedures, orthodontists should be warned of the related benefits and risks. Here we want to present how these innovative approaches can derive better patients' characterization, also offering a different point of view about patient's classification, prognosis and treatment.

Sub-clustering in skeletal class III malocclusion phenotypes via principal component analysis in a southern European population

The main aim of this study was to generate an adequate sub-phenotypic clustering model of class III skeletal malocclusion in an adult population of southern European origin. The study design was conducted in two phases, a preliminary cross-sectional study and a subsequent discriminatory evaluation by main component and cluster analysis to identify differentiated skeletal sub-groups with differentiated phenotypic characteristics. Radiometric data from 699 adult patients of southern European origin were analyzed in 212 selected subjects affected by class III skeletal malocclusion. The varimax rotation was used with Kaiser normalization, to prevent variables with more explanatory capacity from affecting the rotation. A total of 21,624 radiographic measurements were obtained as part of the cluster model generation, using a total set of 55 skeletal variables for the subsequent analysis of the major component and cluster analyses. Ten main axes were generated representing 92.7% of the total variation. Three main components represented 58.5%, with particular sagittal and vertical variables acting as major descriptors. Post hoc phenotypic clustering retrieved six clusters: C1:9.9%, C2:18.9%, C3:33%, C4:3.77%, C5:16%, and C6:16%. In conclusion, phenotypic variation was found in the southern European skeletal class III population, demonstrating the existence of phenotypic variations between identified clusters in different ethnic groups.

A multilevel analysis of craniofacial growth in subjects with untreated Class III malocclusion

OBJECTIVE

To analyse the craniofacial growth of a long-term semi-longitudinal sample of Caucasian subjects with untreated Class III malocclusion.

SETTING AND SAMPLE POPULATION

A total of 144 Caucasian subjects (of North American and Italian origin) with untreated Class III malocclusion.

MATERIALS AND METHODS

Subjects aged 2 years and 9 months up to 21 years and 7 months were selected. A multilevel model was used to calculate growth curves for ten variables for both each individual subject and for the whole sample.

RESULTS

There was a statistically significant increase for total mandibular length (Co-Gn. T2-T1 = 8.4 mm), midfacial length (Co-A. T2-T1 = 3.4 mm) and lower anterior facial height (ANS-Me. T2-T1 = 3.8 mm). The multilevel analysis showed two points of acceleration of growth (about 3-5 years of age and 11-15 years of age) for seven out of ten variables. For Co-Gn and Co-A variables, males presented points of maximum growth delayed by 1 year in comparison with females, with a greater duration (1 year longer) and a greater total growth of about 5 mm. Active mandibular growth continued for a long time after the pubertal spurt: increases in mandibular length ended at about 17 years of age in females and at 21 years and 7 months in males.

CONCLUSIONS

Untreated Class III malocclusion showed a specific growth curve, especially for the mandible, whose excesses added up over time. In males, the amounts of mandibular and midfacial growth during the whole observation time were greater and lasted longer than in females.