AI-powered segmentation system boosts dental imaging accuracy by up to 12 %

A recently published research paper introduced a multimodal contrastive learning system, ToothMCL, for combining CBCT scans and intraoral scans (IOS) to improve tooth segmentation in digital workflows. Tested on a dataset of 3,867 patients, it increased segmentation accuracy by ~12% (CBCT) and ~8% (IOS) compared to previous methods. As digital dentistry continues to grow—such as in diagnostic imaging, aligner planning, and CAD/CAM workflows—tools like this highlight how AI is refining every step of the clinical process.