1.A cohort study of shift work type, physical labor intensity, and their joint association with type 2 diabetes
Haixia LU ; Lan LIN ; Changxue HUANG ; Yuxin JIN ; Yulong LIAN
Journal of Environmental and Occupational Medicine 2026;43(8):936-943
Background Shift work is increasingly common in modern industries as economic globalization continues to advance, and has been associated with type 2 diabetes mellitus (T2DM). Objective To investigate the independent and interactive effects of shift work types and physical labor intensity on the risk of T2DM within an occupational cohort of petroleum workers, aiming to provide a scientific basis for reducing the risk of T2DM among workers engaged in shift work and physical labor. Methods This cohort study enrolled petroleum workers who underwent occupational health examinations at Karamay Central Hospital in Xinjiang from 2013 to 2015. Using stratified sampling, employees aged 20–60 years who had been actively employed and had been working in their current positions for at least one year were selected. A total of 3720 participants completed the baseline questionnaire survey and health examination, and follow-up was conducted in 2023. T2DM was diagnosed according to the 2020 Chinese Guidelines for the Prevention and Treatment of Type 2 Diabetes. Shift work type was categorized into fixed day shift, two-shift rotation, three-team two-shift rotation, four-team three-shift rotation, and other shift schedules. Physical labor intensity was classified as light, moderate, heavy, and extremely heavy according to the Classification of Physical Labor Intensity for Petroleum Field Operations (SY/T 6358—2008). Poisson regression models were used to estimate the associations of shift work type, physical labor intensity, and their joint exposure categories with T2DM risk after adjustment for potential confounders. Results A total of 3473 participants completed follow-up. Poisson regression analysis showed that shift work type was associated with T2DM risk. After adjustment for confounders, compared with fixed day shift workers, those engaged in two-shift rotation, three-team two-shift rotation, four-team three-shift rotation, and other shift schedules had higher risks of T2DM, with RRs of 2.17 (95%CI: 1.44, 3.27), 1.86 (95%CI: 1.16, 2.97), 1.85 (95%CI: 1.17, 2.93), and 2.15 (95%CI: 1.20, 3.88), respectively. T2DM risk also differed across physical labor intensity groups, with the extremely heavy group showing the highest risk (RR=2.26, 95%CI: 1.44, 3.56), followed by the heavy group (RR=1.91, 95%CI: 1.29, 2.82). Compared with workers exposed to fixed day shift + light intensity, increased T2DM risks were observed in those exposed to fixed day shift + heavy intensity (RR=2.38, 95%CI: 1.16, 4.72) or extremely heavy intensity (RR=2.64, 95%CI: 1.13, 5.82); two-shift rotation + light (RR=2.71, 95%CI: 1.11, 6.14), moderate (RR=3.38, 95%CI: 1.43, 7.34), heavy (RR=3.61, 95%CI: 1.52, 7.84), or extremely heavy intensity (RR=4.93, 95%CI: 1.87, 11.12); three-shift two-team rotation + light (RR=2.75, 95%CI: 1.14, 6.19), moderate (RR=3.08, 95%CI: 1.14, 7.51), or heavy/extremely heavy intensity (RR=4.06, 95%CI: 1.77, 8.50); four-shift three-team rotation + heavy/extremely heavy intensity (RR=3.89, 95%CI: 1.78, 7.84); and other shift types + heavy/extremely heavy intensity (RR=4.57, 95%CI: 1.55, 11.30). Conclusion This study demonstrates that various types of shift work significantly increase the risk of T2DM, and that concurrent exposure to high physical labor intensity exacerbates this risk. Regular blood glucose monitoring and targeted health interventions are recommended for workers exposed to rotating shift schedules and high physical work intensity.
2.Exploring Mechanism of Luoshi Neiyi Prescription in Treating Endometriosis Based on Ferroptosis and Serum Metabolomics
Haixia PAN ; Yingqiao ZHONG ; Ting MAO ; Ziyi DENG ; Meilin WU ; Lei HUANG ; Siyang CHEN ; Yong GUO ; Ying ZHOU
Chinese Journal of Experimental Traditional Medical Formulae 2026;32(20):201-212
ObjectiveThis study aimed to investigate the mechanism by which Luoshi Neiyi prescription treats endometriosis (EMs) through regulating ferroptosis, and to screen key metabolites and analyze their association with ferroptosis. MethodsClinical samples of normal endometrium from patients without EMs and eutopic and ectopic endometrium from EMs patients (10 cases each) were collected and divided into control group, eutopic group, and EMs group. Hematoxylin-eosin (HE) staining was performed to observe ectopic lesions of EMs. Immunohistochemistry was used to detect the expression of solute carrier family 7 member 11 (SLC7A11) and glutathione peroxidase 4 (GPX4). Enzyme-linked immunosorbent assay (ELISA) was adopted to determine the levels of malondialdehyde (MDA), ferrous ion (Fe2+), GPX4 and glutathione (GSH) in endometrial tissues, as well as serum levels of Fe2+, GPX4 and GSH. Real-time quantitative polymerase chain reaction (Real-time PCR) was used to detect the mRNA expression of SLC7A11, GPX4, transferrin receptor (TFR) and ferritin heavy chain 1 (FTH1). In the in vitro experiment, primary stromal cells were isolated from ectopic lesions of EMs patients. Cell counting kit-8 (CCK-8) was used to determine the optimal concentration of drug-containing serum for intervention. The level of reactive oxygen species (ROS) was measured, and Real-time PCR was applied to detect ferroptosis-related indicators. In the animal experiments, an EM rat model was established, and the rats were randomly assigned to the sham operation group, EMs group, low-dose Luoshi Neiyi Formula group (7.87 g·kg-1), high-dose Luoshi Neiyi prescription group (15.74 g·kg-1), and danazol group (42 mg·kg-1). Untargeted metabolomics detection and pathway enrichment analysis were conducted on serum samples from patients and rats. Spearman correlation analysis was performed to assess the relationship between differential metabolites and ferroptosis indicators. The correlations between differential metabolites in patient endometrium and serum and key ferroptosis indicators (GPX4, Fe2+, MDA, GSH) as well as ferroptosis-related mRNAs (GPX4, SLC7A11, FTH1) were analyzed, and correlation heatmaps were generated accordingly. ResultsCompared with normal eutopic endometrium, ectopic lesions in EMs patients showed glandular disorganization and stromal fibrosis. In ectopic endometrium, the contents of MDA, ROS, and Fe2+ decreased, while GPX4 level increased, and the mRNA expression of SLC7A11 and GPX4 was upregulated (P<0.05, P<0.01). In serum, the levels of GPX4 and Fe2+ were elevated, whereas the GSH level declined, suggesting abnormalities in ferroptosis-related pathways in ectopic lesions (P<0.05, P<0.01). After intervention with Luoshi Neiyi prescription-containing serum, the intracellular ROS level in ectopic endometrial stromal cells was elevated, the mRNA expression of SLC7A11 and GPX4 was downregulated, and TFR mRNA expression was upregulated (P<0.05, P<0.01). Metabolomics analysis revealed 1104 and 198 differential metabolites in EMs patients and EMs rats, respectively, compared with their corresponding control groups, and both low-dose and high-dose Luoshi Neiyi prescription were found to regulate this metabolic disturbance, with the core regulatory pathways mainly involving arginine and proline metabolism. Correlation analysis showed that the glycerophospholipids including PI(16∶0/17∶0) and PI[18∶2(9Z,12Z)] were negatively correlated with GPX4 and positively correlated with MDA, while 17α-hydroxyprogesterone was positively correlated with GPX4, SLC7A11, and FTH1 q<0.05). ConclusionLuoshi Neiyi prescription may systematically ameliorate disease-associated metabolic dysregulation via modulation of the serum arginine and proline metabolism pathway, and may regulate ferroptosis in ectopic lesions through a mechanism potentially linked to the serum glycerophospholipid and steroid metabolism pathways. Collectively, these findings provide experimental evidence for the clinical application of Luoshi Neiyi prescription.
3.Exploring Mechanism of Luoshi Neiyi Prescription in Treating Endometriosis Based on Ferroptosis and Serum Metabolomics
Haixia PAN ; Yingqiao ZHONG ; Ting MAO ; Ziyi DENG ; Meilin WU ; Lei HUANG ; Siyang CHEN ; Yong GUO ; Ying ZHOU
Chinese Journal of Experimental Traditional Medical Formulae 2026;32(20):201-212
ObjectiveThis study aimed to investigate the mechanism by which Luoshi Neiyi prescription treats endometriosis (EMs) through regulating ferroptosis, and to screen key metabolites and analyze their association with ferroptosis. MethodsClinical samples of normal endometrium from patients without EMs and eutopic and ectopic endometrium from EMs patients (10 cases each) were collected and divided into control group, eutopic group, and EMs group. Hematoxylin-eosin (HE) staining was performed to observe ectopic lesions of EMs. Immunohistochemistry was used to detect the expression of solute carrier family 7 member 11 (SLC7A11) and glutathione peroxidase 4 (GPX4). Enzyme-linked immunosorbent assay (ELISA) was adopted to determine the levels of malondialdehyde (MDA), ferrous ion (Fe2+), GPX4 and glutathione (GSH) in endometrial tissues, as well as serum levels of Fe2+, GPX4 and GSH. Real-time quantitative polymerase chain reaction (Real-time PCR) was used to detect the mRNA expression of SLC7A11, GPX4, transferrin receptor (TFR) and ferritin heavy chain 1 (FTH1). In the in vitro experiment, primary stromal cells were isolated from ectopic lesions of EMs patients. Cell counting kit-8 (CCK-8) was used to determine the optimal concentration of drug-containing serum for intervention. The level of reactive oxygen species (ROS) was measured, and Real-time PCR was applied to detect ferroptosis-related indicators. In the animal experiments, an EM rat model was established, and the rats were randomly assigned to the sham operation group, EMs group, low-dose Luoshi Neiyi Formula group (7.87 g·kg-1), high-dose Luoshi Neiyi prescription group (15.74 g·kg-1), and danazol group (42 mg·kg-1). Untargeted metabolomics detection and pathway enrichment analysis were conducted on serum samples from patients and rats. Spearman correlation analysis was performed to assess the relationship between differential metabolites and ferroptosis indicators. The correlations between differential metabolites in patient endometrium and serum and key ferroptosis indicators (GPX4, Fe2+, MDA, GSH) as well as ferroptosis-related mRNAs (GPX4, SLC7A11, FTH1) were analyzed, and correlation heatmaps were generated accordingly. ResultsCompared with normal eutopic endometrium, ectopic lesions in EMs patients showed glandular disorganization and stromal fibrosis. In ectopic endometrium, the contents of MDA, ROS, and Fe2+ decreased, while GPX4 level increased, and the mRNA expression of SLC7A11 and GPX4 was upregulated (P<0.05, P<0.01). In serum, the levels of GPX4 and Fe2+ were elevated, whereas the GSH level declined, suggesting abnormalities in ferroptosis-related pathways in ectopic lesions (P<0.05, P<0.01). After intervention with Luoshi Neiyi prescription-containing serum, the intracellular ROS level in ectopic endometrial stromal cells was elevated, the mRNA expression of SLC7A11 and GPX4 was downregulated, and TFR mRNA expression was upregulated (P<0.05, P<0.01). Metabolomics analysis revealed 1104 and 198 differential metabolites in EMs patients and EMs rats, respectively, compared with their corresponding control groups, and both low-dose and high-dose Luoshi Neiyi prescription were found to regulate this metabolic disturbance, with the core regulatory pathways mainly involving arginine and proline metabolism. Correlation analysis showed that the glycerophospholipids including PI(16∶0/17∶0) and PI[18∶2(9Z,12Z)] were negatively correlated with GPX4 and positively correlated with MDA, while 17α-hydroxyprogesterone was positively correlated with GPX4, SLC7A11, and FTH1 q<0.05). ConclusionLuoshi Neiyi prescription may systematically ameliorate disease-associated metabolic dysregulation via modulation of the serum arginine and proline metabolism pathway, and may regulate ferroptosis in ectopic lesions through a mechanism potentially linked to the serum glycerophospholipid and steroid metabolism pathways. Collectively, these findings provide experimental evidence for the clinical application of Luoshi Neiyi prescription.
4.Decoding the immune microenvironment of secondary chronic myelomonocytic leukemia due to diffuse large B-cell lymphoma with CD19 CAR-T failure by single-cell RNA-sequencing.
Xudong LI ; Hong HUANG ; Fang WANG ; Mengjia LI ; Binglei ZHANG ; Jianxiang SHI ; Yuke LIU ; Mengya GAO ; Mingxia SUN ; Haixia CAO ; Danfeng ZHANG ; Na SHEN ; Weijie CAO ; Zhilei BIAN ; Haizhou XING ; Wei LI ; Linping XU ; Shiyu ZUO ; Yongping SONG
Chinese Medical Journal 2025;138(15):1866-1881
BACKGROUND:
Several studies have demonstrated the occurrence of secondary tumors as a rare but significant complication of chimeric antigen receptor T (CAR-T) cell therapy, underscoring the need for a detailed investigation. Given the limited variety of secondary tumor types reported to date, a comprehensive characterization of the various secondary tumors arising after CAR-T therapy is essential to understand the associated risks and to define the role of the immune microenvironment in malignant transformation. This study aims to characterize the immune microenvironment of a newly identified secondary tumor post-CAR-T therapy, to clarify its pathogenesis and potential therapeutic targets.
METHODS:
In this study, the bone marrow (BM) samples were collected by aspiration from the primary and secondary tumors before and after CD19 CAR-T treatment. The CD45 + BM cells were enriched with human CD45 microbeads. The CD45 + cells were then sent for 10× genomics single-cell RNA sequencing (scRNA-seq) to identify cell populations. The Cell Ranger pipeline and CellChat were used for detailed analysis.
RESULTS:
In this study, a rare type of secondary chronic myelomonocytic leukemia (CMML) were reported in a patient with diffuse large B-cell lymphoma (DLBCL) who had previously received CD19 CAR-T therapy. The scRNA-seq analysis revealed increased inflammatory cytokines, chemokines, and an immunosuppressive state of monocytes/macrophages, which may impair cytotoxic activity in both T and natural killer (NK) cells in secondary CMML before treatment. In contrast, their cytotoxicity was restored in secondary CMML after treatment.
CONCLUSIONS
This finding delineates a previously unrecognized type of secondary tumor, CMML, after CAR-T therapy and provide a framework for defining the immune microenvironment of secondary tumor occurrence after CAR-T therapy. In addition, the results provide a rationale for targeting macrophages to improve treatment strategies for CMML treatment.
Humans
;
Lymphoma, Large B-Cell, Diffuse/therapy*
;
Tumor Microenvironment/genetics*
;
Antigens, CD19/metabolism*
;
Leukemia, Myelomonocytic, Chronic/genetics*
;
Immunotherapy, Adoptive/adverse effects*
;
Male
;
Single-Cell Analysis/methods*
;
Female
;
Sequence Analysis, RNA/methods*
;
Receptors, Chimeric Antigen
;
Middle Aged
5.Allogeneic hematopoietic stem cell transplantation for the treatment of BCR::ABL-negative neutrophilic myeloid neoplasms: a clinical analysis of 12 cases
Tingting HAN ; Yun HE ; Jing LIU ; Yao CHEN ; Fengrong WANG ; Jingzhi WANG ; Yuhong CHEN ; Haixia FU ; Lanping XU ; Xiaohui ZHANG ; Xiaojun HUANG ; Yu WANG
Chinese Journal of Hematology 2025;46(9):827-832
Objective:To investigate the efficacy and safety of allogeneic hematopoietic stem cell transplantation (allo-HSCT) for the treatment of BCR::ABL-negative chronic neutrophilic leukemia (CNL) and MDS/MPN with neutrophilia.Methods:This study retrospectively analyzed 12 cases of CNL and MDS/MPN with neutrophilia that underwent allo-HSCT from March 2017 to June 2024, comprising 7 males and 5 females with a median age of 48 ( IQR: 28, 59) years. The 2-year overall survival (OS), disease-free survival (DFS), cumulative incidence of relapse (CIR), and transplantation-related mortality (TRM) rates were analyzed. Complications were also assessed. Results:Of the 12 patients, 6 received matched sibling HSCT and 6 received haploidentical HSCT. All patients had successful engraftment, and the median times of neutrophil and platelet engraftment were 17 ( IQR: 11, 24) days and 15 ( IQR: 9, 28) days, respectively. Grade Ⅱ–Ⅳ acute graft versus host disease (GVHD) and chronic GVHD occurred in 2 and 4 cases, respectively. The 2-year OS, DFS, CIR, and TRM rates were (65.6 ± 16.4) %, (41.7 ± 16.6) %, (47.2 ±18.2) %, and (11.1 ± 11.4) %, respectively, after a median follow-up time of 637 ( IQR: 330, 943) days. One patient died from treatment-related complications due to respiratory failure caused by coronavirus disease 2019. Two patients died due to relapse. Conclusion:Allo-HSCT can be applied as a safe and effective approach to treat CNL and MDS/MPN with neutrophilia.
6.CXCL17 methylation as a potential marker for papillary thyroid carcinoma
Chenxia JIANG ; Mengxia LI ; Haixia HUANG ; Junjie LI ; Yifei YIN ; Rongxi YANG
Chinese Journal of Clinical and Experimental Pathology 2025;41(5):627-633,638
Purpose To investigate the correlation between the methylation level at CpG sites of CXCL17 and the clinicopathological parameters of papillary thyroid carcinoma(PTC).Methods samples from 186 cases of PTC and 191 cases of benign thyroid nodule(BTN)were collected.Methylation levels of CXCL17 were semi-quantitatively as-sessed using mass spectrometry.Logistic regression analysis,which adjusted for age,gender and related hormones,was conducted to evaluate the correlation between CXCL17 methylation and PTC,and calculate the odds ratios(ORs)and 95%confidence intervals(CIs).Results Hypomethylation level of CXCL17_CpG_1.2 was significantly associat-ed with PTC(OR=1.36,95%CI:1.16-1.60,P<0.001)and early stage of PTC patients(Stage Ⅰ,OR=1.41,95%CI:1.19-1.67,P<0.001).Gender-based hierarchical management analysis showed that decreased methyla-tion level of CXCL17_CpG_1.2 was significantly associated with female PTC patients(OR=1.39,95%CI:1.15-1.67,P<0.001).In subgroups stratified by age(<50 and≥50 years old),hypomethylation at CXCL17_CpG_1.2 was significantly associated with PTC,with a stronger association in the younger subgroup(<50 years old:OR=1.42,95%CI:1.14-1.77,P<0.01;≥ 50 years old:OR=1.30,95%CI:1.03-1.64,P<0.05).Conclusion There was a significant difference in CXCL17 methylation levels between benign and malignant thyroid tumors.It was showed that hypomethylation of CXCL17 is closely associated with PTC,particularly in young women patient.Thus,CXCL17 methylation may serve as a biomarker for accurate differential diagnosis of thyroid nodule.
7.Study on different test methods to evaluate the drying effect of flexible endoscope
Xianhuang LI ; Ziyi HUANG ; Liuji CHEN ; Jing ZHEN ; Haixia LIU ; Yanxia ZHANG ; Xun ZHANG ; Xi HUANG ; Jiuhong MA
China Journal of Endoscopy 2025;31(11):34-39
Objective To compare the effectiveness of cobalt chloride test papers and borescope in evaluating the drying of endoscopes,providing a reference for clinical selection of appropriate assessment tools.Methods 10 gastroscopes and 10 colonoscopes procedures were selected from November 2023 to March 2024 for circulation experiments.After endoscope reprocessing,they were randomly divided into five groups with different drying times(30 s,3 min,6 min,9 min,and 12 min)with 200 samples.Cobalt chloride test paper and borescope were used to evaluate the drying effectiveness.Results The qualified rate of gastrointestinal endoscope by borescope was significantly higher than that by cobalt chloride test paper,but cobalt chloride test paper was obviously better than that by borescope in convenience and cost-effectiveness.The borescope had the function of visualization,and the quantitative positioning observation of residual droplets showed that there were a large number of droplets left after gastrointestinal endoscope drying for 30 s,and the number of droplets was obviously reduced after drying for 3 min.Among them,the gastroscope was mainly light(1~5 drops/strip)and none(0 drops/strip),and the colonoscope was mainly moderate(6~10 drops/strip)and light droplets.Compared with different drying times,there was a significant statistical difference in the number of residual droplets in the endoscope(P<0.01),and the localized droplets mainly remained in the distal bending part and the proximal bifurcation part of the lumen.Conclusion Both cobalt chloride test papers and borescope are important methods for assessing endoscope drying efficacy,each with its own advantages and limitations.Clinicians can use both methods in combination to comprehensively evaluate the overall dryness of endoscopes,thereby ensuring the safety and effectiveness of the endoscope reprocessing process.
8.Combined screening of two primary immunodeficiency diseases and spinal muscular atrophy in neonates by multiplex real-time fluorescence quantitative PCR
Chao ZHANG ; Jianbin YANG ; Shiqiang SHANG ; Chi CHEN ; Huaqing MAO ; Xiaolei HUANG ; Fang HONG ; Haixia MIAO ; Hanyi ZHAO ; Rulai YANG
Chinese Journal of Laboratory Medicine 2025;48(2):249-257
Objective:To explore the feasibility of joint screening of the two primary immunodeficiency diseases [severe combined immunodeficiency (SCID) and X-linked agammaglobulinemia(XLA)] and spinal muscular atrophy(SMA) in newborns by multiplex real-time quantitative PCR technology, and to provide evidence for early screening, diagnosis and treatment of children.Methods:Cross-sectional study. From July 2021 to January 2023, a total of 103 240 dry blood spots samples of newborns were collected which were delivered to Neonatal Disease Screening Center of Zhejiang by cold chain transportation. The concentrations of the T cell receptor excision ring (TREC), Kappa deletion of the recombinant excision loop (KREC), and exon 7 deletion of Survival Motor Neuron 1 (SMN1) gene in dry blood spots were simultaneously detected by multiplex real-time fluorescence quantitative PCR, taken ribonuclease P/MRP 30 000 subunits (RPP30) as an internal reference gene. The positive newborns were further diagnosed by other laboratory tests and gene sequencing was taken as gold standard. Children samples from 1 case of SCID, 3 cases of XLA and 2 cases of SMA were used for positive verification. The correlation between detected concentration of TREC/KREC and basic information in newborns were analyzed. The differences among groups for each factor were analyzed.Results:One case of SCID, 2 cases of XLA, 9 cases of SMA and 7 cases of other genetic diseases (4 cases of DiGeorge syndrome, 1 case of trisomy 21 syndrome, 1 case of Noonan syndrome and 1 case of super male syndrome) were identified by multiplex real-time fluorescence quantitative PCR. The positive predictive values of screening neonatal SCID, XLA and SMA were 2.44% (1/41), 2.78% (2/72) and 9/9 respectively. Taking the samples from clinically diagnosed 1 case of SCID, 3 cases of XLA and 2 cases of SMA as positive validation samples, which were all identified. The detected results of TREC/KREC correlated with time of blood collection, sex, weight, gestational age and delivery mode of newborns, whose r values were 0.162/0.187, 0.066/0.032, 0.045/0.042, ?0.015/?0.088 and 0.014/0.068 respectively (all P<0.05). Conclusions:Relying on current neonatal screening platform in Zhejiang, it is feasible to screen jointly two kinds of primary immunodeficiency diseases and spinal muscular atrophy in newborns by multiple real-time fluorescence quantitative PCR technology.
9.Building of an intelligent DRG grouping audit system in a hospital
Juan ZHANG ; Yang PU ; Wen LIU ; Yingpeng WANG ; Lianhua KONG ; Yaxin HUANG ; Bin WAN ; Haixia DING
Chinese Journal of Hospital Administration 2025;41(8):614-618
Diagnosis-related groups (DRG) payment is an important component of deepening the reform of medical insurance payment methods. In June 2023, a tertiary hospital launched an intelligent DRG grouping audit system to enhance grouping accuracy. By establishing a multi departmental collaborative organizational structure, building a standardized knowledge base and a rule base covering five categories (diagnosis, fees, testing, nursing, and pathology), and integrating electronic medical records, medical orders, testing, and imaging data throughout the entire diagnosis and treatment process, the intelligent DRG grouping audit system with data collection, identification, extraction, comparison, and output modules was constructed to achieve intelligent audit. At the same time, it was formed a closed-loop management system for pre reporting quality control, in-process group entry control, and post data analysis and assessment, which would prevent the risk of differentiated behaviors such as high coding and high sets, and ensure the reasonable use of medical insurance funds. By January 2024, the system had covered 89 ADRG groups, and improved the efficiency and quality of DRG grouping audit. Compared with February to May 2023, the monthly average rejection rate of medical records on the first page decreased by 9.4% after the system was put into operation (June to December 2023), and core medical indicators such as the number of DRG groups, medical insurance settlement cases, and time consumption index continued to improve. The practical experience could provide reference and inspiration for other hospitals in China.
10.Research and application of a new deep learning based strategy for platelet histogram review
Enming ZHANG ; Chao YANG ; Xianchun CHEN ; Yan LIN ; Taixue AN ; Haixia LI ; Yongjian HE ; Zhiwei LIU ; Limei FENG ; Wanying LIN ; Tie XIONG ; Kai QIU ; Ya GAO ; Lizhu HUANG ; Jing HE ; Chunyan WANG ; Dehua SUN ; Bo SITU ; Lei ZHENG
Chinese Journal of Laboratory Medicine 2025;48(9):1201-1206
Objective:To develop an artificial intelligence (AI)-based platelet review strategy to identify abnormal platelet histograms with no significant difference between initial impedance platelet count (PLT-I) and PLT-F results.Methods:This study included 5 119 routine blood analysis in Nanfang Hospital of Southern Medical University and its Ganzhou branch from July 2023 and March 2024. Specimens exhibiting abnormal platelet histograms and an initial platelet count >40×10?/L underwent review using the fluorescent platelet count (PLT-F) channel. Consistency of the results was defined as a difference between impedance platelet count (PLT-I) and PLT-F less than ±20% of the PLT-F results. A deep learning model was developed using platelet and red blood cell histogram data from a training set of 3 807 specimens. The model′s diagnostic performance was evaluated on an independent external validation set ( n=805) using receiver operating characteristic (ROC) curve analysis. Changes in the number of reviewed samples and sample turnaround time were analyzed to assess its clinical utility. Results:The deep learning model based on platelet and red blood cell histograms achieved an area under the ROC curve (AUC) of 0.854 in the training set. At a cutoff value of 0.1, the sensitivity was 0.954 and specificity was 0.358. The model could reduce review by 16.80% (190/1 131). In the validation set, the AUC was 0.805, with a sensitivity of 0.955 and specificity of 0.307, corresponding to a reduction of 17.41% (47/270) in reviewed specimens.Conclusion:The platelet review prediction model developed based on deep learning technology can efficiently identify samples with consistent results before and after review, reducing unnecessary reviews and shortening specimen testing time, thereby improving the efficiency of platelet test.

Result Analysis
Print
Save
E-mail