@inproceedings{song-etal-2026-escaping,
title = "Escaping the Frequent Finding Shortcut: Pathology-Aware Curriculum Learning 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 = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.688/",
pages = "14430--14442",
ISBN = "979-8-89176-462-0",
abstract = "Radiology Report Generation (RRG) has benefited from stronger encoder-decoder pipelines and particularly Large Language Model (LLM) report generators, but most methods still perform training through random mini-batches. This practice is poorly matched to chest radiology corpora, where common normal statements and frequent findings dominate the optimization signal, while clinically important pathological attributes appear in the long tail. We argue that random sampling pushes report generators toward a frequent-finding shortcut, where models quickly learn fluent report narratives and high-frequency disease phrases, then receive limited pressure to recover rare attributes that contribute little to aggregate cross-entropy. To address this problem, we propose ParaC, a pathology-aware replay curriculum for long-tail RRG. ParaC ranks training reports with pathology labels and structured medical knowledge, expands the training process from simple reports to complex long-tail cases, and replays prototype anchors to preserve basic cross-modal grounding. Extensive experiments on IU X-Ray and MIMIC-CXR confirm the superiority of ParaC on several existing methods, with further ablations analyzing the large gains brought by our pathology-aware curriculum to cover rare pathological findings."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="song-etal-2026-escaping">
<titleInfo>
<title>Escaping the Frequent Finding Shortcut: Pathology-Aware Curriculum Learning 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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Radiology Report Generation (RRG) has benefited from stronger encoder-decoder pipelines and particularly Large Language Model (LLM) report generators, but most methods still perform training through random mini-batches. This practice is poorly matched to chest radiology corpora, where common normal statements and frequent findings dominate the optimization signal, while clinically important pathological attributes appear in the long tail. We argue that random sampling pushes report generators toward a frequent-finding shortcut, where models quickly learn fluent report narratives and high-frequency disease phrases, then receive limited pressure to recover rare attributes that contribute little to aggregate cross-entropy. To address this problem, we propose ParaC, a pathology-aware replay curriculum for long-tail RRG. ParaC ranks training reports with pathology labels and structured medical knowledge, expands the training process from simple reports to complex long-tail cases, and replays prototype anchors to preserve basic cross-modal grounding. Extensive experiments on IU X-Ray and MIMIC-CXR confirm the superiority of ParaC on several existing methods, with further ablations analyzing the large gains brought by our pathology-aware curriculum to cover rare pathological findings.</abstract>
<identifier type="citekey">song-etal-2026-escaping</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.688/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>14430</start>
<end>14442</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Escaping the Frequent Finding Shortcut: Pathology-Aware Curriculum Learning 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 Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F song-etal-2026-escaping
%X Radiology Report Generation (RRG) has benefited from stronger encoder-decoder pipelines and particularly Large Language Model (LLM) report generators, but most methods still perform training through random mini-batches. This practice is poorly matched to chest radiology corpora, where common normal statements and frequent findings dominate the optimization signal, while clinically important pathological attributes appear in the long tail. We argue that random sampling pushes report generators toward a frequent-finding shortcut, where models quickly learn fluent report narratives and high-frequency disease phrases, then receive limited pressure to recover rare attributes that contribute little to aggregate cross-entropy. To address this problem, we propose ParaC, a pathology-aware replay curriculum for long-tail RRG. ParaC ranks training reports with pathology labels and structured medical knowledge, expands the training process from simple reports to complex long-tail cases, and replays prototype anchors to preserve basic cross-modal grounding. Extensive experiments on IU X-Ray and MIMIC-CXR confirm the superiority of ParaC on several existing methods, with further ablations analyzing the large gains brought by our pathology-aware curriculum to cover rare pathological findings.
%U https://aclanthology.org/2026.findings-emnlp.688/
%P 14430-14442
Markdown (Informal)
[Escaping the Frequent Finding Shortcut: Pathology-Aware Curriculum Learning for Radiology Report Generation](https://aclanthology.org/2026.findings-emnlp.688/) (Song et al., Findings 2026)
ACL