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

Abstract

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.

Evaluation of deep learning and convolutional neural network algorithms accuracy for detecting and predicting anatomical landmarks on 2D lateral cephalometric images: A systematic review and meta-analysis

“…Another study found that, compared to other radiographic techniques, cephalograms provide quantitative and qualitative results for anatomical landmark detection ( Bichu et al, 2021 , Joda and Pandis, 2021 , Liu et al, 2021 , Auconi et al, 2022 ). Skeletal landmark detection improves the accuracy of quantitative analyses as it identifies reference points.…”

Identification and verification of the temozolomide resistance feature gene DACH1 in gliomas

“…Machine learning is a multidisciplinary technique that combines statistics and computer science and uses a variety of strategies and algorithms to arrive at the best model ( 11 ). Compared with traditional statistical methods that concentrate on the causality of hypothesis testing and the significance of model features, machine learning focuses more on the downscaling of high-dimensional data and the predictive performance and generalization of models ( 12 , 13 ). As a result, machine learning is better suited for analyzing complex and large quantities of data (e.g., gene expression analysis, image feature extraction, drug sensitivity prediction, etc.).…”