Cheng Shen

Dr. Cheng Shen


Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

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

Representative Publications

View All Publications
11 March 2025 in Journal of Medical Internet Research
Verified
Journal Article
Open access
ZDZhang-Yi Dai
Zhang-Yi Dai
FWFu-Qiang Wang
Fu-Qiang Wang
CSCheng Shen
Cheng Shen
YJYan-Li Ji
Yan-Li Ji
ZLZhi-Yang Li
Zhi-Yang Li
QPQiang Pu
Qiang Pu
Background: Systematic reviews and meta-analyses rely on labor-intensive literature screening. While machine learning offers potential automation, its accuracy remains suboptimal. This raises the question of whether emerging large language models (LLMs) can provide a more accurate and efficient approach. Objective: This paper evaluates the sensitivity, specificity, and summary receiver operating characteristic (SROC) curve of LLM-assisted literature screening. Methods: We conducted a diagnostic study comparing the accuracy of LLM-assisted screening versus manual literature screening across 6 thoracic surgery meta-analyses. Manual screening by 2 investigators served as the reference standard. LLM-assisted screening was performed using ChatGPT-4o (OpenAI) and Claude-3.5 (Anthropic) sonnet, with discrepancies resolved by Gemini-1.5 pro (Google). In addition, 2 open-source, machine learning–based screening tools, ASReview (Utrecht University) and Abstrackr (Center for Evidence Synthesis in Health, Brown University School of Public Health), were also evaluated. We calculated sensitivity, specificity, and 95% CIs for the title and abstract, as well as full-text screening, generating pooled estimates and SROC curves. LLM prompts were revised based on a post hoc error analysis. Results: LLM-assisted full-text screening demonstrated high pooled sensitivity (0.87, 95% CI 0.77-0.99) and specificity (0.96, 95% CI 0.91-0.98), with the area under the curve (AUC) of 0.96 (95% CI 0.94-0.97). Title and abstract screening achieved a pooled sensitivity of 0.73 (95% CI 0.57-0.85) and specificity of 0.99 (95% CI 0.97-0.99), with an AUC of 0.97 (95% CI 0.96-0.99). Post hoc revisions improved sensitivity to 0.98 (95% CI 0.74-1.00) while maintaining high specificity (0.98, 95% CI 0.94-0.99). In comparison, the pooled sensitivity and specificity of ASReview tool-assisted screening were 0.58 (95% CI 0.53-0.64) and 0.97 (95% CI 0.91-0.99), respectively, with an AUC of 0.66 (95% CI 0.62-0.70). The pooled sensitivity and specificity of Abstrackr tool-assisted screening were 0.48 (95% CI 0.35-0.62) and 0.96 (95% CI 0.88-0.99), respectively, with an AUC of 0.78 (95% CI 0.74-0.82). A post hoc meta-analysis revealed comparable effect sizes between LLM-assisted and conventional screening. Conclusions: LLMs hold significant potential for streamlining literature screening in systematic reviews, reducing workload without sacrificing quality. Importantly, LLMs outperformed traditional machine learning-based tools (ASReview and Abstrackr) in both sensitivity and AUC values, suggesting that LLMs offer a more accurate and efficient approach to literature screening.

11 October 2023 in Biomedicines
Verified
Journal Article
Open access
PDF file
SSShangqi Song
Shangqi Song
CSCheng Shen
Cheng Shen
YHYang Hu
Yang Hu
YHYazhou He
Yazhou He
YYYong Yuan
Yong Yuan
YXYuyang Xu
Yuyang Xu
Surgery is a crucial treatment option for patients with resectable esophageal cancer. The emergence of minimally invasive esophageal techniques has led to the popularity of video-assisted thoracoscopic esophagectomy, which has proven to be more advantageous than traditional thoracotomy. However, some patients with esophageal cancer may not benefit from this procedure. Individualized treatment plans may be necessary for patients with varying conditions and tolerances to anesthesia, making conventional surgical methods unsuitable. Inflatable video-assisted mediastinoscopic transhiatal esophagectomy (IVMTE) has emerged as a promising treatment option for esophageal cancer because it does not require one-lung ventilation, reduces postoperative complications, and expands surgical indications. This technique also provides surgical opportunities for patients with impaired pulmonary function or thoracic lesions. It is crucial to have a comprehensive understanding of the advancements and limitations of IVMTE to tailor treatment plans and improve outcomes in patients with esophageal cancer. Understanding the advantages and limitations of this surgical method will help specific patients with esophageal cancer. We conducted a thorough review of the relevant literature to examine the importance of IVMTE for individualized treatment of this disease.

8 June 2024 in BMC Pediatrics
Verified
Journal Article
Open access
PDF file
XHXiaolu Hu
Xiaolu Hu
CSCheng Shen
Cheng Shen
Background: A case-control study was conducted to analyze the role of cerebrospinal fluid immunoglobulin in the differential diagnosis of autoimmune encephalitis and viral encephalitis in children. Methods: One hundred and twenty patients with autoimmune encephalitis (AE) treated in our hospital from February 2021 to February 2022 were included as the observation group (AE group). 100 patients with viral encephalitis (VE group) were selected as the control group. The clinical data of all patients were collected and analyzed retrospectively. Immunoglobulin G (IgG) and immunoglobulin A (IgA)in cerebrospinal fluid of the two patients were measured by immune turbidimetry. Immunoglobulin M (IgM), and the diagnostic value of immunoglobulin in cerebrospinal fluid (CSF) in patients with AE was analyzed by receiver working curve (ROC). Results: The level of IgG in the cerebrospinal fluid of the AE group was higher than that of the VE group, and the level of IgM was lower than that of the VE group, and the difference was statistically significant (P < 0.05). There was no significant difference in IgA levels between the two groups (P > 0.05). In terms of Magnetic Resonance (MR) features, the paraventricular, hippocampal, occipital and parietal lobes were more involved in AE patients, frontal and temporal lobes were more involved in VE patients, and paraventricular and occipital lobes were involved in MS. The proportion of bilateral extensive lesions in both groups was significantly higher than 50%. The proportions of patients in the AE group involving the lateral ventricle, insula, and parietal lobes were significantly higher than those in the VE group, and the proportions involving the basal ganglia, temporal lobes, and frontal lobes were significantly lower than those in the VE group, and the differences were statistically significant (All P < 0.05). The Area Under Curve (AUC) of IgG, IgA and IgM alone in the diagnosis of AE were 0.795(0.587–0.762), 0.602(0.502–0.631) and 0.627(0.534–0.708), respectively with the sensitivity values of 81.24% and 65.608, respectively and the specificity values of 65.08%, 57.54% and 75.01% respectively. The AUC of IgA + IgM in the diagnosis of AE was 0.733(0.617–0.849), and the sensitivity and specificity are 62.58% and 75.07% respectively. The AUC of IgA + IgG in the diagnosis of AE was 0.823(0.730–0.917), and the sensitivity and specificity were 81.24% and 67.54% respectively. The AUC of IgG + IgM in the diagnosis of AE was 0.886(0.814 ~ 0.958), and the sensitivity and specificity were 84.48% and 77.59% respectively. The AUC of IgA + IgM + IgG in the diagnosis of AE was 0.924 (0.868–0.981) with the sensitivity of 93.82%, and the specificity of 77.56%. Conclusion: The level of immunoglobulin in cerebrospinal fluid can be used as an effective reference index for the diagnosis of AE. The combined detection of IgA, IgM and IgG can improve the accuracy, sensitivity and specificity of AE.

24 July 2023 in Thoracic Cancer
Verified
Journal Article
Open access
XCXi Chen
Xi Chen
YWYun Wang
Yun Wang
GCGuowei Che
Guowei Che
CSCheng Shen
Cheng Shen
Epithelioid emangioendothelioma (EHE) is a rare tumor originating from vascular endothelial cells, which belongs to malignant vascular tumors with low to moderate differentiation and potential metastatic ability and its incidence rate is less than one in a million. We herein report a rare case of EHE of pulmonary origin and a review of the literature concerning the clinical and pathological features of this disease. The patient underwent left upper lobectomy to completely extirpate the lesion by video‐assisted thoracic surgery. Our findings suggest the difficulty of making a diagnosis before surgery and that more cases need to be reported in order to facilitate the preoperative diagnosis of such a rare tumor.

9 November 2023 in Journal of Translational Medicine
Verified
Journal Article
Open access
PDF file
CSCheng Shen
Cheng Shen

13 September 2023 in International Journal of Surgery
Verified
Journal Article
Open access
ZDZhangyi Dai
Zhangyi Dai
XZXiaolong Zhang
Xiaolong Zhang
CSCheng Shen
Cheng Shen
YXYikai Xing
Yikai Xing
QPQiang Pu
Qiang Pu
An abstract is unavailable. This article is available as a PDF only.