# OrthoAI v2: From Single-Agent Segmentation to Dual-Agent Treatment Planning for Clear Aligners

arXiv: https://arxiv.org/abs/2603.15663 · HTML: https://mlsanity.com/paper/2603.15663/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 18 references that are in the map.
- First posted on arXiv 10 Mar 2026.

## Builds on (most cited first, 4 of 4)

- [Dynamic Graph CNN for Learning on Point Clouds](https://mlsanity.com/paper/1801.07829/) (2018)
- [Two-Stage Mesh Deep Learning for Automated Tooth Segmentation and Landmark Localization on 3D Intraoral Scans](https://mlsanity.com/paper/2109.11941/) (2021)
- [Teeth3DS+: An Extended Benchmark for Intraoral 3D Scans Analysis](https://mlsanity.com/paper/2210.06094/) (2022)
- [CHaRM: Conditioned Heatmap Regression Methodology for Accurate and Fast Dental Landmark Localization](https://mlsanity.com/paper/2501.13073/) (2025)

## Abstract (arXiv)

We present OrthoAI v2, the second iteration of our open-source pipeline for AI-assisted orthodontic treatment planning with clear aligners, substantially extending the single-agent framework previously introduced. The first version established a proof-of-concept based on Dynamic Graph Convolutional Neural Networks (dgcnn) for tooth segmentation but was limited to per-tooth centroid extraction, lacked landmark-level precision, and produced a scalar quality score without staging simulation. vtwo addresses all three limitations through three principal contributions: (i)~a second agent adopting the Conditioned Heatmap Regression Methodology (charm)~citerodriguez2025charm for direct, segmentation-free dental landmark detection, fused with Agent~1 via a confidence-weighted orchestrator in three modes (parallel, sequential, single-agent); (ii)~a composite six-category biomechanical scoring model (biomechanics × 0.30 + staging × 0.20 + attachments × 0.15 + IPR × 0.10 + occlusion × 0.10 + predictability × 0.15) replacing the binary pass/fail check of v1; (iii)~a multi-frame treatment simulator generating F = A × r temporally coherent 6-DoF tooth trajectories via SLERP interpolation and evidence-based staging rules, enabling ClinCheck 4D visualisation. On a synthetic benchmark of 200 crowding scenarios, the parallel ensemble of OrthoAI v2 reaches a planning quality score of 92.8 ± 4.1 vs. 76.4 ± 8.3 for OrthoAI v1, a +21% relative gain, while maintaining full CPU deployability (4.2 ± 0.8~s).