@inproceedings{song-etal-2026-condistill,
title = "{C}on{D}istill: Asymmetric Context Self-Distillation for Radiology Report Generation",
author = "Song, Ziming and
Shen, Mingyang and
Vong, Chi Man",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.937/",
pages = "21617--21630",
ISBN = "979-8-89176-459-0",
abstract = "Building fine-grained cross-modal alignment via pure radiograph-report pairs has been a long-standing challenge in the task of Radiology Report Generation (RRG). An intuitive solution towards this issue lies at introducing extra contexts to assist the report generation process, yet clinical contexts are normally unavailable in most real-world scenarios. In this paper, we present an asymmetric context self-distillation framework for RRG, namely ConDistill. To facilitate radiograph-only report generation, it refines the basic radiograph-report alignment, by progressively injecting additional contexts via multi-stage curriculum. Since these contexts are not available in inference, we develop a self-distillation strategy to transfer the bridge states from the teacher branches with extra contexts to the student one with solely radiograph inputs. For evaluation, we conduct extensive experiments on both radiograph-only and multi-context benchmarks, i.e., MIMIC-CXR and MIMIC-RG4, respectively. Experimental results validate our promising performance and further ablations offer potential insights for the context learning in LLM-based RRG frameworks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="song-etal-2026-condistill">
<titleInfo>
<title>ConDistill: Asymmetric Context Self-Distillation for Radiology Report Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ziming</namePart>
<namePart type="family">Song</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mingyang</namePart>
<namePart type="family">Shen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chi</namePart>
<namePart type="given">Man</namePart>
<namePart type="family">Vong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Building fine-grained cross-modal alignment via pure radiograph-report pairs has been a long-standing challenge in the task of Radiology Report Generation (RRG). An intuitive solution towards this issue lies at introducing extra contexts to assist the report generation process, yet clinical contexts are normally unavailable in most real-world scenarios. In this paper, we present an asymmetric context self-distillation framework for RRG, namely ConDistill. To facilitate radiograph-only report generation, it refines the basic radiograph-report alignment, by progressively injecting additional contexts via multi-stage curriculum. Since these contexts are not available in inference, we develop a self-distillation strategy to transfer the bridge states from the teacher branches with extra contexts to the student one with solely radiograph inputs. For evaluation, we conduct extensive experiments on both radiograph-only and multi-context benchmarks, i.e., MIMIC-CXR and MIMIC-RG4, respectively. Experimental results validate our promising performance and further ablations offer potential insights for the context learning in LLM-based RRG frameworks.</abstract>
<identifier type="citekey">song-etal-2026-condistill</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.937/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>21617</start>
<end>21630</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ConDistill: Asymmetric Context Self-Distillation for Radiology Report Generation
%A Song, Ziming
%A Shen, Mingyang
%A Vong, Chi Man
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F song-etal-2026-condistill
%X Building fine-grained cross-modal alignment via pure radiograph-report pairs has been a long-standing challenge in the task of Radiology Report Generation (RRG). An intuitive solution towards this issue lies at introducing extra contexts to assist the report generation process, yet clinical contexts are normally unavailable in most real-world scenarios. In this paper, we present an asymmetric context self-distillation framework for RRG, namely ConDistill. To facilitate radiograph-only report generation, it refines the basic radiograph-report alignment, by progressively injecting additional contexts via multi-stage curriculum. Since these contexts are not available in inference, we develop a self-distillation strategy to transfer the bridge states from the teacher branches with extra contexts to the student one with solely radiograph inputs. For evaluation, we conduct extensive experiments on both radiograph-only and multi-context benchmarks, i.e., MIMIC-CXR and MIMIC-RG4, respectively. Experimental results validate our promising performance and further ablations offer potential insights for the context learning in LLM-based RRG frameworks.
%U https://aclanthology.org/2026.emnlp-main.937/
%P 21617-21630
Markdown (Informal)
[ConDistill: Asymmetric Context Self-Distillation for Radiology Report Generation](https://aclanthology.org/2026.emnlp-main.937/) (Song et al., EMNLP 2026)
ACL