Lingyi Chi

Lingyi Chi


Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 June 2025 in JMIR Medical Informatics
Verified
Journal Article
Open access
ZYZhong Yao
Zhong Yao
LDLiantan Duan
Liantan Duan
SXShuo Xu
Shuo Xu
LCLingyi Chi
Lingyi Chi
DSDongfang Sheng
Dongfang Sheng
Background: Research on large language models (LLMs) in the medical field has predominantly focused on models trained with English-language corpora, evaluating their performance within English-speaking contexts. The performances of models trained with non–English language corpora and their performance in non-English contexts remain underexplored. Objective: This study aimed to evaluate the performances of LLMs trained on different languages corpora by using the Chinese National Medical Licensing Examination (CNMLE) as a benchmark and constructed analogous questions. Methods: Under different prompt settings, we sequentially posed questions to 7 LLMs: 2 primarily trained on English-language corpora and 5 primarily on Chinese-language corpora. The models’ responses were compared against standard answers to calculate the accuracy rate of each model. Further subgroup analyses were conducted by categorizing the questions based on various criteria. We also collected error sets to explore patterns of mistakes across different models. Results: Under the zero-shot setting, 6 out of 7 models exceeded the passing level, with the highest accuracy rate achieved by the Chinese LLM Baichuan (86.67%), followed by ChatGPT (83.83%). In the constructed questions, all 7 models exceeded the passing threshold, with Baichuan maintaining the highest accuracy rate (87.00%). In few-shot learning, all models exceeded the passing threshold. Baichuan, ChatGLM, and ChatGPT retained the highest accuracy. While Llama showed marked improvement over previous tests, the relative performance rankings of other models stayed similar to previous results. In subgroup analyses, English models demonstrated comparable or superior performance to Chinese models on questions related to ethics and policy. All models except Llama generally had higher accuracy rates for simple questions than for complex ones. The error set of ChatGPT was similar to those of other Chinese models. Multimodel cross-verification outperformed single model, particularly improving accuracy rate on simple questions. The implementation of dual-model and tri-model verification achieved accuracy rates of 94.17% and 96.33% respectively. Conclusions: At the current level, LLMs trained primarily on English corpora and those trained mainly on Chinese corpora perform similarly well in CNMLE, with Chinese models still outperforming. The performance difference between ChatGPT and other Chinese LLMs are not solely due to communication barriers but are more likely influenced by disparities in the training data. By using a method of cross-verification with multiple LLMs, excellent performance can be achieved in medical examinations.

1 January 2007 in Clinical Neurology and Neurosurgery
Verified
Journal Article
Closed access
XZXiuhe Zhao
Xiuhe Zhao
LCLingyi Chi
Lingyi Chi
YZYuying Zhao
Yuying Zhao
ZCZhaofu Chi
Zhaofu Chi

1 May 2021 in Neuropsychiatric Disease and Treatment
Verified
Journal Article
Open access
PDF file
TZTongxia Zhang
Tongxia Zhang
CYChuanzhu Yan
Chuanzhu Yan
YLYiming Liu
Yiming Liu
CLcao Lili
cao Lili
KJKunqian Ji
Kunqian Ji
YZYuying Zhao
Yuying Zhao
Purpose: Leukodystrophies are frequently regarded as childhood disorders, but they can occur at any age, and the clinical and imaging patterns of the adult-onset form are usually different from the better-known childhood variants. Several reports have shown that various late-onset leukodystrophies, such as X-linked adrenoleukodystrophy and Krabbe disease, may present as spastic paraplegia with the absence of the characteristic white matter lesions on neuroimaging; this can be easily misdiagnosed as hereditary spastic paraplegia. The objective of this study was to investigate the frequency of late-onset leukodystrophies in patients with spastic paraplegia. Patients and Methods: We performed genetic analysis using a custom-designed gene panel for leukodystrophies in 112 hereditary spastic paraplegia-like patients. Results: We identified pathogenic mutations in 13 out of 112 patients, including five patients with adrenomyeloneuropathy, three with Krabbe disease, three with Alexander disease, and two with cerebrotendinous xanthomatosis. In terms of clinical manifestations, in addition to spastic paraplegia, three adrenomyeloneuropathy probands also had adrenocortical insufficiency, two Alexander disease probands developed urinary retention, one CTX proband developed cataracts and chronic diarrhea and the other presented with chronic diarrhea and mild tendon xanthomatosis. None of the patients had evidence of diffuse leukodystrophy on neuroimaging. Conclusion: Patients with late-onset spastic paraplegia should be screened for underlying leukodystrophies, irrespective of the presence of additional complicating symptoms and neuroimaging abnormalities.

10 September 2024 in BMJ Open
Verified
Journal Article
Open access
PDF file
ZMZhiyong Ma
Zhiyong Ma
ZFZhijie Fu
Zhijie Fu
NLNa Li
Na Li
SHShanying Huang
Shanying Huang
LCLingyi Chi
Lingyi Chi
Introduction Heart failure (HF) is a complex syndrome that affects millions of people worldwide and leads to significant morbidity and mortality. Sacubitril/valsartan, a combination drug consisting of a neprilysin inhibitor and an angiotensin receptor blocker (ARB), has shown a greater improvement in the prognosis of HF than ACE inhibitors (ACEI) or ARB. Recent studies have found that ACEI/ARB or sacubitril/valsartan can increase flow-mediated dilation (FMD) and reduce pulse wave velocity (PWV), which are independent predictors of cardiovascular events and HF prognosis. The purpose of this study is to assess and compare the effect of sacubitril/valsartan and ACEI/ARB on FMD and PWV using meta-analysis and further provide a reference for the role of sacubitril/valsartan in the treatment of HF. Methods and analysis Clinical randomised controlled trials investigating the effect of sacubitril/valsartan and/or ACEI/ARB on FMD and PWV in patients with HF will be searched in the relevant database, including PubMed, Web of Science, Embase, Cochrane Library and China’s National Knowledge Infrastructure up to January 2024. The outcomes of interest are changes in endothelial function assessed by FMD and changes in arterial stiffness assessed by PWV. The risk of bias was evaluated using the revised Cochrane risk of bias tool for randomised trials (RoB2.0). Review Manager V.5.3 software is used for meta-analysis data synthesis, sensitivity analysis, meta-regression analysis, subgroup analysis and risk of bias assessment. The reporting bias of studies will be evaluated using the funnel plot, in which symmetry will be assessed by Begg’s and Egger’s tests. The evidence quality of the included studies will be evaluated by the Grading of Recommendations Assessment, Development, and Evaluation. Ethics and dissemination This study only analyses research data from the published literature and therefore does not require ethical approval. We will submit the systematic review to a peer-reviewed journal. PROSPERO registration number CRD42024538148.