Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hamamci, Ibrahim Ethem, Er, Sezgin, Shit, Suprosanna, Reynaud, Hadrien, Yang, Dong, Guo, Pengfei, Edgar, Marc, Xu, Daguang, Kainz, Bernhard, Menze, Bjoern
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908607569723392
author Hamamci, Ibrahim Ethem
Er, Sezgin
Shit, Suprosanna
Reynaud, Hadrien
Yang, Dong
Guo, Pengfei
Edgar, Marc
Xu, Daguang
Kainz, Bernhard
Menze, Bjoern
author_facet Hamamci, Ibrahim Ethem
Er, Sezgin
Shit, Suprosanna
Reynaud, Hadrien
Yang, Dong
Guo, Pengfei
Edgar, Marc
Xu, Daguang
Kainz, Bernhard
Menze, Bjoern
contents Recent progress in vision-language modeling for 3D medical imaging has been fueled by large-scale computed tomography (CT) corpora with paired free-text reports, stronger architectures, and powerful pretrained models. This has enabled applications such as automated report generation and text-conditioned 3D image synthesis. Yet, current approaches struggle with high-resolution, long-sequence volumes: contrastive pretraining often yields vision encoders that are misaligned with clinical language, and slice-wise tokenization blurs fine anatomy, reducing diagnostic performance on downstream tasks. We introduce BTB3D (Better Tokens for Better 3D), a causal convolutional encoder-decoder that unifies 2D and 3D training and inference while producing compact, frequency-aware volumetric tokens. A three-stage training curriculum enables (i) local reconstruction, (ii) overlapping-window tiling, and (iii) long-context decoder refinement, during which the model learns from short slice excerpts yet generalizes to scans exceeding 300 slices without additional memory overhead. BTB3D sets a new state-of-the-art on two key tasks: it improves BLEU scores and increases clinical F1 by 40% over CT2Rep, CT-CHAT, and Merlin for report generation; and it reduces FID by 75% and halves FVD compared to GenerateCT and MedSyn for text-to-CT synthesis, producing anatomically consistent 512*512*241 volumes. These results confirm that precise three-dimensional tokenization, rather than larger language backbones alone, is essential for scalable vision-language modeling in 3D medical imaging. The codebase is available at: https://github.com/ibrahimethemhamamci/BTB3D
format Preprint
id arxiv_https___arxiv_org_abs_2510_20639
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging
Hamamci, Ibrahim Ethem
Er, Sezgin
Shit, Suprosanna
Reynaud, Hadrien
Yang, Dong
Guo, Pengfei
Edgar, Marc
Xu, Daguang
Kainz, Bernhard
Menze, Bjoern
Computer Vision and Pattern Recognition
Recent progress in vision-language modeling for 3D medical imaging has been fueled by large-scale computed tomography (CT) corpora with paired free-text reports, stronger architectures, and powerful pretrained models. This has enabled applications such as automated report generation and text-conditioned 3D image synthesis. Yet, current approaches struggle with high-resolution, long-sequence volumes: contrastive pretraining often yields vision encoders that are misaligned with clinical language, and slice-wise tokenization blurs fine anatomy, reducing diagnostic performance on downstream tasks. We introduce BTB3D (Better Tokens for Better 3D), a causal convolutional encoder-decoder that unifies 2D and 3D training and inference while producing compact, frequency-aware volumetric tokens. A three-stage training curriculum enables (i) local reconstruction, (ii) overlapping-window tiling, and (iii) long-context decoder refinement, during which the model learns from short slice excerpts yet generalizes to scans exceeding 300 slices without additional memory overhead. BTB3D sets a new state-of-the-art on two key tasks: it improves BLEU scores and increases clinical F1 by 40% over CT2Rep, CT-CHAT, and Merlin for report generation; and it reduces FID by 75% and halves FVD compared to GenerateCT and MedSyn for text-to-CT synthesis, producing anatomically consistent 512*512*241 volumes. These results confirm that precise three-dimensional tokenization, rather than larger language backbones alone, is essential for scalable vision-language modeling in 3D medical imaging. The codebase is available at: https://github.com/ibrahimethemhamamci/BTB3D
title Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.20639