@inproceedings{wang-etal-2026-cptcoder,
title = "{CPTC}oder: A Reliable {LLM} System for Medical Procedure Code Prediction",
author = "Wang, Benlu and
Shangguan, Ziyao and
Tegtmeyer, Kyle and
Zhang, Zhenyu and
Chheang, Sophie and
Cohan, Arman",
editor = "Durrett, Greg and
Jian, Ping",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 3: System Demonstrations)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-demo.60/",
pages = "605--614",
ISBN = "979-8-89176-392-0",
abstract = "We present **CPTCoder**, a human-in-the-loop system that predicts standardized medical procedure codes from clinical text. Clinical procedure coding is an extreme multi-label classification problem over a long-tailed space of short numeric identifiers, where a single-digit difference denotes an entirely different procedure. CPTCoder adapts an instruction-tuned LLM with a code-aware vocabulary and constrained decoding that guarantees all outputs are valid codes. To support human review, we derive per-code posterior inclusion probabilities from n-best reweighting, producing interpretable confidence scores that rank predictions and flag uncertain cases. A post-decoding constraint repair step enforces mutual-exclusion rules between conflicting codes. To enable reproducible research in this underexplored setting, we release **MIMIC-CPT**, a PhysioNet-accessible benchmark of 37,885 expert-cleaned report{--}code pairs with a deliberately hardened test split: 88{\%} of test examples contain label combinations unseen during training, and over a third include codes with five or fewer training occurrences. We additionally provide 413,085 weakly aligned pairs and evaluate on a separate live dataset from a hospital, which includes out-of-domain radiology reports with billing-expert-verified labels. CPTCoder achieves 0.61 and 0.51 micro-F1 on the hardened MIMIC split and Hospital-298 respectively, outperforming the strongest baseline by 12 and 5 absolute points while reducing digit-level near-miss errors."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-cptcoder">
<titleInfo>
<title>CPTCoder: A Reliable LLM System for Medical Procedure Code Prediction</title>
</titleInfo>
<name type="personal">
<namePart type="given">Benlu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ziyao</namePart>
<namePart type="family">Shangguan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kyle</namePart>
<namePart type="family">Tegtmeyer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhenyu</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sophie</namePart>
<namePart type="family">Chheang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Arman</namePart>
<namePart type="family">Cohan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Greg</namePart>
<namePart type="family">Durrett</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ping</namePart>
<namePart type="family">Jian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-392-0</identifier>
</relatedItem>
<abstract>We present **CPTCoder**, a human-in-the-loop system that predicts standardized medical procedure codes from clinical text. Clinical procedure coding is an extreme multi-label classification problem over a long-tailed space of short numeric identifiers, where a single-digit difference denotes an entirely different procedure. CPTCoder adapts an instruction-tuned LLM with a code-aware vocabulary and constrained decoding that guarantees all outputs are valid codes. To support human review, we derive per-code posterior inclusion probabilities from n-best reweighting, producing interpretable confidence scores that rank predictions and flag uncertain cases. A post-decoding constraint repair step enforces mutual-exclusion rules between conflicting codes. To enable reproducible research in this underexplored setting, we release **MIMIC-CPT**, a PhysioNet-accessible benchmark of 37,885 expert-cleaned report–code pairs with a deliberately hardened test split: 88% of test examples contain label combinations unseen during training, and over a third include codes with five or fewer training occurrences. We additionally provide 413,085 weakly aligned pairs and evaluate on a separate live dataset from a hospital, which includes out-of-domain radiology reports with billing-expert-verified labels. CPTCoder achieves 0.61 and 0.51 micro-F1 on the hardened MIMIC split and Hospital-298 respectively, outperforming the strongest baseline by 12 and 5 absolute points while reducing digit-level near-miss errors.</abstract>
<identifier type="citekey">wang-etal-2026-cptcoder</identifier>
<location>
<url>https://aclanthology.org/2026.acl-demo.60/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>605</start>
<end>614</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CPTCoder: A Reliable LLM System for Medical Procedure Code Prediction
%A Wang, Benlu
%A Shangguan, Ziyao
%A Tegtmeyer, Kyle
%A Zhang, Zhenyu
%A Chheang, Sophie
%A Cohan, Arman
%Y Durrett, Greg
%Y Jian, Ping
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-392-0
%F wang-etal-2026-cptcoder
%X We present **CPTCoder**, a human-in-the-loop system that predicts standardized medical procedure codes from clinical text. Clinical procedure coding is an extreme multi-label classification problem over a long-tailed space of short numeric identifiers, where a single-digit difference denotes an entirely different procedure. CPTCoder adapts an instruction-tuned LLM with a code-aware vocabulary and constrained decoding that guarantees all outputs are valid codes. To support human review, we derive per-code posterior inclusion probabilities from n-best reweighting, producing interpretable confidence scores that rank predictions and flag uncertain cases. A post-decoding constraint repair step enforces mutual-exclusion rules between conflicting codes. To enable reproducible research in this underexplored setting, we release **MIMIC-CPT**, a PhysioNet-accessible benchmark of 37,885 expert-cleaned report–code pairs with a deliberately hardened test split: 88% of test examples contain label combinations unseen during training, and over a third include codes with five or fewer training occurrences. We additionally provide 413,085 weakly aligned pairs and evaluate on a separate live dataset from a hospital, which includes out-of-domain radiology reports with billing-expert-verified labels. CPTCoder achieves 0.61 and 0.51 micro-F1 on the hardened MIMIC split and Hospital-298 respectively, outperforming the strongest baseline by 12 and 5 absolute points while reducing digit-level near-miss errors.
%U https://aclanthology.org/2026.acl-demo.60/
%P 605-614
Markdown (Informal)
[CPTCoder: A Reliable LLM System for Medical Procedure Code Prediction](https://aclanthology.org/2026.acl-demo.60/) (Wang et al., ACL 2026)
ACL
- Benlu Wang, Ziyao Shangguan, Kyle Tegtmeyer, Zhenyu Zhang, Sophie Chheang, and Arman Cohan. 2026. CPTCoder: A Reliable LLM System for Medical Procedure Code Prediction. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 605–614, San Diego, California, United States. Association for Computational Linguistics.