Jul 2026· Journal of Medical Internet Research· Vol 28· 0 citations· 32 references
Medicine
TL;DR
The hybrid pipeline integrating rule-based dictionary normalization with large language model (LLM)–based postprocessing significantly improved Korean-English code-switched medical ASR accuracy.
Abstract
Abstract Background Clinical documentation burden contributes significantly to physician burnout, with health care professionals spending much of their time on electronic health record interactions. Automatic speech recognition (ASR) systems offer a promising solution; however, their application in Korean medical settings faces unique challenges due to widespread Korean-English code-switching, where clinicians routinely alternate between Korean conversational language and English medical terminology within single utterances. Objective This study aimed to develop and evaluate a hybrid postprocessing approach combining medical terminology dictionary normalization with large language model (LLM)–based postprocessing to improve ASR accuracy for Korean-English code-switched medical speech. Methods We constructed a speech dataset from 23,652 nursing progress notes, with a linguistic composition of 67.73% (512,626/756,866) Korean, 23.54% (178,166/756,866) English, and 8.73% (66,074/756,866) numerals or special symbols. Four Korean nurses recorded the notes using 5 microphone types in an acoustically isolated environment. Speech recognition was performed using OpenAI’s gpt-4o-transcribe model. For postprocessing, a medical terminology dictionary containing 1070 mapping entries was constructed from 1000 nursing progress notes to normalize Korean phonetic renderings of English medical terms. Six LLMs (2 GPT and 4 Claude variants) were then evaluated across 5 temperature settings (0.0‐0.8). Performance was assessed using BERTScore (bidirectional encoder representations from transformers score; F1), Sentence-BERT cosine similarity, word error rate, and character error rate (CER), comparing postprocessed outputs against the original written notes. Statistical significance was assessed using paired Wilcoxon signed-rank tests with Holm correction (α=.05). Results Temperature optimization showed that all postprocessing models had small temperature-related effect sizes (all |Cohen dz| ≤0.15), with GPT-4o exhibiting the largest dependency and statistically significant improvement at temperature 0.6 (Holm-adjusted P<.001 for both BERTScore and CER) and the 4 Claude variants and GPT-4.1 exhibiting practically consistent performance across all settings. Baseline ASR achieved a BERTScore of 0.9131 and CER of 0.2336. Dictionary-based normalization performed 43,507 word-level substitutions in 70.8% (16,754/23,652) of transcribed sentences. LLM-only postprocessing reduced CER by 36.09% (Claude Sonnet 4) and 32.53% (GPT-4o) compared to baseline. The combined dictionary-LLM approach achieved the best performance: Claude Sonnet 4 attained a BERTScore of 0.9638 and CER of 0.0820, representing a 64.9% reduction in CER from baseline (P<.001). Conclusions The hybrid pipeline integrating rule-based dictionary normalization with LLM postprocessing significantly improved Korean-English code-switched medical ASR accuracy. Dictionary-based normalization yielded consistent additional improvements over LLM-only postprocessing for both GPT-4o and Claude Sonnet 4. As this modular framework requires no model retraining, it offers a practical means of mitigating multilingual challenges in medical ASR.
This study proposes a multi-stage intelligent documentation framework that integrates a 12-layer Conformer architecture, BERT-BiLSTM-CRF semantic modeling, and BART-based structured text generation that illustrates the effectiveness of deep time-frequency feature extraction and contextual sequence modeling for complex noisy signals.
Xian-Kun Zou, L. Wang, J. Sun et al.· Advanced Electromagnetics· 0 citations
Comparative analysis confirmed that the proposed bilingual NLP model outperforms existing monolingual and rule-based systems in linguistic inclusiveness and accessibility.
Background/Objectives: Clinical documentation places a significant time burden on healthcare professionals, including in the context of home care. Large language models (LLMs) offer potential for automated note generation, but current approaches rely on static prompt templates that fail to generalize across care settings, languages, and documentation formats. This study proposes and evaluates an adaptive retrieval-augmented generation (RAG) framework that uses retrieval as a format adaptation mechanism, enabling the generation of structured clinical notes from patient–provider transcripts across various documentation formats without model fine-tuning. Methods: The proposed framework retrieves dialogue–note pairs that demonstrate the structure of specific sections, allowing the transfer of formatting knowledge during inference. Experiments were conducted on three datasets covering two languages and various documentation formats: the Japanese Visiting Nurse corpus (JP-VN), MTS-Dialog, and ACI-BENCH. Six controlled conditions were evaluated: zero-shot prompting (C1), static few-shot prompting (C2), dense retrieval (C3), random retrieval (C4), sparse BM25 retrieval (C5), and hybrid retrieval using reciprocal rank fusion (RRF) (C6). Performance metrics include structural adherence to required section headings, content quality (ROUGE-1, BLEU, BERTScore), and the number of hallucinated clinical entities per generated record. Results: Structure compliance increased from 0–37% under static conditions (C1/C2) to 91–100% under all adaptive RAG conditions (C3–C6) across all datasets. On MTS-Dialog, dense retrieval achieved the highest content quality (ROUGE-1: 0.519 vs. 0.446–0.492 for C4–C6; p<0.001). Hallucinated entities in JP-VN decreased from 2.73–3.58 per note (C1/C2) to 1.15–1.30 (C3–C6), an approximately 55–56% reduction. Conclusions: Adaptive RAG can improve structure compliance and reduce hallucinations in multilingual clinical note generation without dataset-specific prompt engineering or model fine-tuning. These findings support retrieval-based format adaptation as a generalizable mechanism for diverse clinical documentation contexts.
AI_LectureNote is a historical, readability-oriented post-ASR workflow for Korean-English medical lectures. It rewrites speech-to-text output into study transcripts while restoring Latin-script medical terms rather than Korean phonetic transliterations. We retrospectively evaluate the workflow on four author-recorded lectures across five conditions. In this pilot, post-processing raised the macro English-script rendering rate from 0.39 to 0.71 on the whisper-1 path and from 0.26 to 0.65 when applied to 3-minute chunked gpt-4o-transcribe output. However, English-script rendering did not imply semantic faithfulness: the two post-processed conditions showed semantic drift in 34 and 36 of 282 reference sentences and polarity failures in 11 and 13 of 101 polarity-cue rows. A descriptive cross-input comparison suggested different candidate failure patterns: polarity-failure sets overlapped more strongly across front-ends (Jaccard 0.60; 9 shared of 15 unioned failures) than general semantic-drift sets (Jaccard 0.23; 13 shared of 57 unioned drifts). This single-annotator pilot documents concrete failure modes rather than population rates and supports evaluating surface accuracy, term-script rendering, chunk-level script consistency, and medical-meaning preservation separately.
Kyeongeon Lee, Donghoon Chang, Seungryeol Baek et al.· 0 citations
BACKGROUND
Effective patient education requires accurate communication aligned with patients' emotional and semantical needs. Text-based large language models (LLMs) lack access to non-verbal cues, which may contribute to misaligned responses.
METHODS
We evaluated emotional and semantic misalignment in a text-based LLM using 64,200 utterances from 16,583 patient education cases across six departments and three centers. Dolphin was developed integrating text and audio cues and evaluated through emotion recognition, semantic consistency assessment, branch-level ablations, and a double-blinded randomized trial against a matched text-based LLM comparator (Chinese Clinical Trial Registry: (ChiCTR2500095933).
FINDINGS
The text-based LLM showed emotional misalignment in 36.7% of responses and semantic misalignment in 28.3% of cases, with higher misalignment under greater burden. Dolphin outperformed the text-based LLM in emotion recognition accuracy (0.886 vs. 0.713) and semantic consistency (84.9% vs. 82.1%; both adjusted p < 0.001). Ablations supported contribution of audio branches. Dolphin received higher expert ratings than the text-based LLM and human educators (all p < 0.001). In 555 patients, Dolphin was associated with greater patient satisfaction (98.6% vs. 93.8%), suggestion acceptance (76.1% vs. 58.9%; p < 0.001), proactive disclosure (44.6% vs. 26.5%; p < 0.001), and fewer 7-day unplanned recontact (12.9% vs. 22.9%; p = 0.002). No unsafe recommendations or safety events were identified.
CONCLUSIONS
Compared with text-based LLM, Dolphin improved emotional-semantic alignment and patient-education outcomes, supporting bimodal alignment as a strategy for reducing misalignment-driven communication failures.
FUNDING
National Natural Science Foundation of China, State Key Laboratory Special Fund, and Chinese Academy of Medical Sciences Innovation Fund.
Peixing Wan, Zigeng Huang, Haoquan Huang et al.· i Medicina· 0 citations
This study evaluates six AI medical translation systems using a mixed-methods approach, integrating BLEU scores, user surveys (N=775), and behavioral data. A standardized bilingual corpus was constructed from authoritative sources including the WHO and NMPA, while an AHP-BLEU hybrid model was developed to combine subjective user evaluations with objective scores across word, sentence, and paragraph levels in both Chinese-English and English-Chinese tasks. Results show Atman and Youdao outperform others in overall quality, with DeepL excelling in terminology. Spearman correlation analysis confirms a strong positive association (ρ=0.943, p=0.005) between BLEU scores and user satisfaction, validating the model. Despite rapid advances, current AI medical translation tools still struggle with term accuracy, context adaptation, and document complexity. The proposed AHP-BLEU framework helps align evaluation with user priorities, offering a more balanced view of performance. Future improvements should include semantic-aware metrics and human-verified baselines to better support multilingual medicine applications, from Traditional Chinese Medicine globalization to virtual consultations.
Xiangyu Wu, Yongqi Zeng, Qing Wang et al.· Advances in Engineering Tech...· 0 citations