Deep learning-assisted creation of a large-scale cone beam computed tomography dataset for the segmentation of impacted mandibular third molar

Cheng Niu, Bin Wang, Tingyuan Zeng, Jiangling Sun, Xubo Duan, Zhengfen Li, Meiyan Rong, Liming Chen, Wei Yang · 2026

Read Deep learning-assisted creation of a large-scale cone beam computed tomography dataset for the segmentation of impacted mandibular third molar by Cheng Niu, Bin Wang, Tingyuan Zeng, Jiangling Sun, Xubo Duan, Zhengfen Li, Meiyan Rong, Liming Chen, Wei Yang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13247881.txt.

Description

Cheng Niu and co-authors build in this 2026 paper a large scan library for planning wisdom tooth surgery. Impacted lower wisdom teeth sit close to a nerve, so surgeons outline both before cutting. Hand outlining of cone beam scans is slow, and learning systems need many marked examples. The team used deep learning itself to grow the library fast.

The set holds 1,262 scan volumes: 186 marked by experts and the rest drafted by a model and merged by a set strategy. Their attention-based network then topped rival models at outlining tooth, neighbor tooth, and nerve. Machine-drafted marks took about seventeen seconds per case and passed doctor review, a template for other scan libraries.

Catalog record

CREATORCheng Niu, Bin Wang, Tingyuan Zeng, Jiangling Sun, Xubo Duan, Zhengfen Li, Meiyan Rong, Liming Chen, Wei Yang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE31 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13247881

Browse the full library · About the Tower of Babel