Objective:To assess common errors in head and neck CT angiography (CTA) image reports using four types of large language models (LLM), namely GPT-4, DeepSeek, ERNIE Bot and SparkDesk, and to assess the feasibility of using existing LLMs to support quality control of radiology reports in Chinese.Methods:The study was a cross-sectional study. Totally 1 000 head and neck CTA image reports were randomly selected using the simple random sampling method from Xuanwu Hospital, Capital Medical University in 2023, including 500 primary reports and 500 finalized reports. Two radiologists collaboratively identified six types of errors in the reports: description errors, writing errors, left-right confusion errors, diagnostic omissions, logical sequence errors, and other errors. The overall quality of the reports was assessed using a 5-point Likert scale. Subsequently, GPT-4, DeepSeek, ERNIE Bot and SparkDesk models were employed to detect the same six types of errors in the imaging reports and to provide overall scoring. The results from manual review were considered the gold standard for calculating F1 score to evaluate model performance. Intra-class correlation coefficients ( ICC) were used to assess the consistency between manual scores and the overall scores from the four LLMs. Results:In the primary imaging reports, the proportions of manually detected errors were as follows: descriptive errors 2.6% (13/500), writing errors 0.6% (3/500), left-right confusion errors 0, diagnostic omissions 6.4% (32/500), logical sequence errors 5.2% (26/500), and other errors 0. In the finalized imaging reports, the proportions of errors across the six categories were 0.2% (1/500), 0, 0, 0, 0, and 0.2% (1/500), respectively. For error detection in the primary imaging reports, the F1 scores of GPT-4 for the six error types were 0.992, 0.997, 0.997, 0.967, 0.980, and 0.992, respectively. DeepSeek achieved F1 scores of 0.980, 0.955, 0.981, 0.920, 0.995, and 0.960; ERNIE Bot scored 0.982, 0.990, 1.000, 0.956, 0.976, and 0.999; and SparkDesk achieved 0.985, 0.995, 1.000, 0.961, 0.982, and 1.000. In the detection of errors in finalized imaging reports, GPT-4′s F1 scores were 0.994, 0.995, 0.998, 0.973, 0.989, and 0.993; DeepSeek scored 0.968, 0.965, 0.985, 0.971, 0.991, and 0.983; ERNIE Bot achieved 0.996, 0.992, 1.000, 0.983, 0.999, and 0.997; and SparkDesk achieved 0.999, 0.999, 1.000, 1.000, 1.000, and 0.999. The consistency between GPT-4, DeepSeek, and SparkDesk models and human ratings was moderate, with ICC values of 0.514, 0.560, and 0.515 respectively (all P0.001); in contrast, the overall score of ERNIE Bot showed poor consistency with human ratings, with an ICC of 0.221 ( P0.001). Conclusion:LLMs demonstrate high accuracy in detecting errors in head and neck CTA imaging reports. The overall scoring of report quality shows moderate consistency with manual assessments, indicating a certain feasibility for automated quality control in reporting.

The acupoint name is a core term in traditional Chinese medicine and has its own mysterious and abstruse feature. Designated by the international organizations such as World Federation of Chinese Medicine Societies, World Health Organization, the phonetic translation method has been adopted for the standardization of acupuncture nomenclature. But this method neglects the cultural attributes of acupoint names. The liberal translation should be considered appropriately. English translation of acupoint name should be composed of two steps, intralingual translation (exegesis) and interlingual translation. During exegesis, the methods for discriminating phonetic loan character, selecting meanings and identifying character patterns should be sufficiently used. The interlingual translation is launched only after the fully understanding of acupoint names (based on intralingual translation).
Acupuncture Points ; Terminology as Topic ; Humans ; Translations ; Language ; Translating ; Medicine, Chinese Traditional

Pharmacovigilance impact research(PIR),as an important application field of pharmacoepidemiology,has attracted continuous attention in recent years from drug regulatory authorities,pharmaceutical manufacturers,and the academic community both domestically and internationally.This paper provides an interpretation of PIR based on the Guide for Methodology in Pharmacoepidemiologic Research(2nd edition).First,an overview of the implications of PIR will be provided,focusing on the pathways of pharmacovigilance activities and the significant importance of conducting PIR.Second,it reviews commonly used study designs and presents illustrative case examples.Building on this,the specific statistical considerations relevant to PIR were discussed.Finally,the challenges and prospects of conducting pharmacovigilance impact studies in a scientific and standardized manner are summarized.Compared with the previous edition,the 2nd edition has expanded the application scenarios of pharmacoepidemiology to include new areas such as PIR.Drawing on the guideline content and practical experience,this paper provides a detailed introduction and case analysis of PIR,serving as a reference for researchers engaged in this field.

Objective:To summarize the clinical features, laboratory findings, treatment and prognosis of children confirmed as Mycoplasma pneumoniae-induced rash and mucositis (MIRM) in children. Methods:This retrospective study concluded 6 children diagnosed as MIRM in Department of Gastroenterology and Infectious Diseases, Shanghai Children′s Hospital, School of Medicine, Shanghai Jiao Tong University from August 2023 to April 2024. This paper described the characteristics of MIRM and analyzed the therapeutic strategy and prognosis.Results:A total of 6 children were diagnosed as MIRM including 2 boys and 4 girls with an age of onset was 6.4 (3.1, 7.5) years. Among the 6 patients, 4 patients had oral mucosal involvement among whom 2 showed crusting of the lips. Four patients had ocular involvement manifesting as conjunctival congestion and increased secretion. All patients presented with skin lesions, manifesting as target-shaped damage in 4 cases, herpes herpetiformis in 1 case and purpura-like rash in 1 case. Serological tests for Mycoplasma pneumoniae IgM and Mycoplasma pneumoniae nucleic acid test were positive in all 6 cases. Two cases received intravenous immunogloblin infusion combined with methylprednisolone, monotherapy of methylprednisolone in 4 cases. The course of glucocorticoids was 1-7 weeks, and the initial dose was 2-4 mg/(kg·d), which was gradually reduced according to the rash. The children were followed up for 3 to 9 months, no case suffered from long term ocular or cutaneous complications or recurrence of rash. All cases had good prognosis. Conclusions:Children diagnosed as MIRM present with mild symptoms and usually have good prognosis with early identification and appropriate intervention. Individualized therapy should be applied based on the severity of skin involvement.

Pharmacovigilance impact research(PIR),as an important application field of pharmacoepidemiology,has attracted continuous attention in recent years from drug regulatory authorities,pharmaceutical manufacturers,and the academic community both domestically and internationally.This paper provides an interpretation of PIR based on the Guide for Methodology in Pharmacoepidemiologic Research(2nd edition).First,an overview of the implications of PIR will be provided,focusing on the pathways of pharmacovigilance activities and the significant importance of conducting PIR.Second,it reviews commonly used study designs and presents illustrative case examples.Building on this,the specific statistical considerations relevant to PIR were discussed.Finally,the challenges and prospects of conducting pharmacovigilance impact studies in a scientific and standardized manner are summarized.Compared with the previous edition,the 2nd edition has expanded the application scenarios of pharmacoepidemiology to include new areas such as PIR.Drawing on the guideline content and practical experience,this paper provides a detailed introduction and case analysis of PIR,serving as a reference for researchers engaged in this field.

Objective:To assess common errors in head and neck CT angiography (CTA) image reports using four types of large language models (LLM), namely GPT-4, DeepSeek, ERNIE Bot and SparkDesk, and to assess the feasibility of using existing LLMs to support quality control of radiology reports in Chinese.Methods:The study was a cross-sectional study. Totally 1 000 head and neck CTA image reports were randomly selected using the simple random sampling method from Xuanwu Hospital, Capital Medical University in 2023, including 500 primary reports and 500 finalized reports. Two radiologists collaboratively identified six types of errors in the reports: description errors, writing errors, left-right confusion errors, diagnostic omissions, logical sequence errors, and other errors. The overall quality of the reports was assessed using a 5-point Likert scale. Subsequently, GPT-4, DeepSeek, ERNIE Bot and SparkDesk models were employed to detect the same six types of errors in the imaging reports and to provide overall scoring. The results from manual review were considered the gold standard for calculating F1 score to evaluate model performance. Intra-class correlation coefficients ( ICC) were used to assess the consistency between manual scores and the overall scores from the four LLMs. Results:In the primary imaging reports, the proportions of manually detected errors were as follows: descriptive errors 2.6% (13/500), writing errors 0.6% (3/500), left-right confusion errors 0, diagnostic omissions 6.4% (32/500), logical sequence errors 5.2% (26/500), and other errors 0. In the finalized imaging reports, the proportions of errors across the six categories were 0.2% (1/500), 0, 0, 0, 0, and 0.2% (1/500), respectively. For error detection in the primary imaging reports, the F1 scores of GPT-4 for the six error types were 0.992, 0.997, 0.997, 0.967, 0.980, and 0.992, respectively. DeepSeek achieved F1 scores of 0.980, 0.955, 0.981, 0.920, 0.995, and 0.960; ERNIE Bot scored 0.982, 0.990, 1.000, 0.956, 0.976, and 0.999; and SparkDesk achieved 0.985, 0.995, 1.000, 0.961, 0.982, and 1.000. In the detection of errors in finalized imaging reports, GPT-4′s F1 scores were 0.994, 0.995, 0.998, 0.973, 0.989, and 0.993; DeepSeek scored 0.968, 0.965, 0.985, 0.971, 0.991, and 0.983; ERNIE Bot achieved 0.996, 0.992, 1.000, 0.983, 0.999, and 0.997; and SparkDesk achieved 0.999, 0.999, 1.000, 1.000, 1.000, and 0.999. The consistency between GPT-4, DeepSeek, and SparkDesk models and human ratings was moderate, with ICC values of 0.514, 0.560, and 0.515 respectively (all P0.001); in contrast, the overall score of ERNIE Bot showed poor consistency with human ratings, with an ICC of 0.221 ( P0.001). Conclusion:LLMs demonstrate high accuracy in detecting errors in head and neck CTA imaging reports. The overall scoring of report quality shows moderate consistency with manual assessments, indicating a certain feasibility for automated quality control in reporting.

Objective:To summarize the clinical features, laboratory findings, treatment and prognosis of children confirmed as Mycoplasma pneumoniae-induced rash and mucositis (MIRM) in children. Methods:This retrospective study concluded 6 children diagnosed as MIRM in Department of Gastroenterology and Infectious Diseases, Shanghai Children′s Hospital, School of Medicine, Shanghai Jiao Tong University from August 2023 to April 2024. This paper described the characteristics of MIRM and analyzed the therapeutic strategy and prognosis.Results:A total of 6 children were diagnosed as MIRM including 2 boys and 4 girls with an age of onset was 6.4 (3.1, 7.5) years. Among the 6 patients, 4 patients had oral mucosal involvement among whom 2 showed crusting of the lips. Four patients had ocular involvement manifesting as conjunctival congestion and increased secretion. All patients presented with skin lesions, manifesting as target-shaped damage in 4 cases, herpes herpetiformis in 1 case and purpura-like rash in 1 case. Serological tests for Mycoplasma pneumoniae IgM and Mycoplasma pneumoniae nucleic acid test were positive in all 6 cases. Two cases received intravenous immunogloblin infusion combined with methylprednisolone, monotherapy of methylprednisolone in 4 cases. The course of glucocorticoids was 1-7 weeks, and the initial dose was 2-4 mg/(kg·d), which was gradually reduced according to the rash. The children were followed up for 3 to 9 months, no case suffered from long term ocular or cutaneous complications or recurrence of rash. All cases had good prognosis. Conclusions:Children diagnosed as MIRM present with mild symptoms and usually have good prognosis with early identification and appropriate intervention. Individualized therapy should be applied based on the severity of skin involvement.

Objective:To evaluate the efficacy and prognostic factors of allogeneic hematopoietic stem cell transplantation (allo-HSCT) in patients with secondary acute myeloid leukemia (sAML) .Methods:In this multicenter, retrospective clinical study, adult patients aged ≥18 years who underwent allo-HSCT for sAML at four centers of the Zhejiang Hematopoietic Stem Cell Transplantation Collaborative Group from January 2014 to November 2022 were included, and the efficacy and prognostic factors of allo-HSCT were analyzed.Results:A total of 95 patients were enrolled; 66 (69.5%) had myelodysplastic syndrome-acute myeloid leukemia (MDS-AML) , 4 (4.2%) had MDS/MPN-AML, and 25 (26.3%) had therapy-related AML (tAML) . The 3-year CIR, LFS, and overall survival (OS) rates were 18.6% (95% CI 10.2%-27.0%) , 70.6% (95% CI 60.8%-80.4%) , and 73.3% (95% CI 63.9%-82.7%) , respectively. The 3-year CIRs of the M-AML group (including MDS-AML and MDS/MPN-AML) and the tAML group were 20.0% and 16.4%, respectively ( P=0.430) . The 3-year LFSs were 68.3% and 75.4%, respectively ( P=0.176) . The 3-year OS rates were 69.7% and 75.4%, respectively ( P=0.233) . The 3-year CIRs of the groups with and without TP53 mutations were 60.0% and 13.7%, respectively ( P=0.003) ; the 3-year LFSs were 20.0% and 76.5%, respectively ( P=0.002) ; and the 3-year OS rates were 40.0% and 77.6%, respectively ( P=0.002) . According to European LeukmiaNet 2022 (ELN2022) risk stratification, the 3-year CIRs of patients in the low-, intermediate-, and high-risk groups were 8.3%, 17.8%, and 22.6%, respectively ( P=0.639) . The three-year LFSs were 91.7%, 69.5%, and 65.6%, respectively ( P=0.268) . The 3-year OS rates were 91.7%, 71.4%, and 70.1%, respectively ( P=0.314) . Multivariate analysis revealed that advanced disease at allo-HSCT and TP53 mutations were independent risk factors for CIR, LFS, and OS. Conclusion:There was no significant difference in the prognosis of patients who underwent allo-HSCT among the MDS-AML, MDS/MPN-AML, and tAML groups. Advanced disease at transplantation and TP53 mutations were poor prognostic factors. ELN2022 risk stratification had limited value for predicting the prognosis of patients with sAML following allo-HSCT.

Objective To investigate the effects and mechanism of Zhachong Shisanwei Pills on rats with cerebral ischemia.Methods Totally 75 rats were randomly divided into sham-operation group,model group,positive drug group(Ginaton,21.6 mg/kg),and Zhachong Shisanwei Pills low-,medium-,and high-dosage groups(81,162,324 mg/kg).Each treatment group was given the corresponding drug by gavage for 5 days.On the 6th day,a cerebral ischemia rat model was prepared by suture method.After 24 hours of modeling,the drugs were given in the same manner for 2 days.Neurological function scoring,horizontal beam walking scoring,and grip strength testing were performed on rats.TTC staining was used to detect the cerebral infarction rate,HE staining and Nissl staining were used to observe the morphology of brain tissue.TUNEL staining was used to detect the apoptosis rate of brain tissue cells.Differential genes in the treatment of cerebral ischemia using Zhachong Shisanwei Pills were screened by transcriptomics,and RT-qPCR,immunohistochemistry and Western blot were used to detect differential gene mRNA and protein expression.Results Compared with the sham-operation group,the model group rats showed a decrease in neurological function scores,horizontal beam walking scores,grip strength,an increase in cerebral infarction rate,neuronal nucleus condensation,vacuolar changes,widened intercellular spaces,the number of Nissl bodies reduced,and the apoptosis rate increased(P<0.01,P<0.001);compared with the model group,the Zhachong Shisanwei Pills medium-dosage group showed an increase in neurological function score,horizontal beam walking score,and grip strength in rats,a decrease in cerebral infarction rate,a lower degree of neuronal damage,an increase in the number of Nissl bodies,and a decrease in cell apoptosis rate(P<0.05,P<0.01).Transcriptome and bioinformatics analysis screened the Hippo signaling pathway related to the anti-cerebral ischemia effect of Zhachong Shisanwei Pills.The key genes of this pathway,mammalian sterile line 20 like kinase(MST1)1,Yes related protein(YAP)1,large tumor suppressor kinase(LATS)1,and TEA domain family member(TEAD)1 were detected.The results showed that the expression of MST1 mRNA and protein in brain tissue of model rats significantly increased,while the expressions of YAP1,LATS1,TEAD1 mRNA and protein significantly decreased;Zhachong Shisanwei Pills could down-regulate the expression of MST1 in brain tissue of model rats,and up-regulate the expressions of YAP1,LATS1 and TEAD1.Conclusion Zhachong Shisanwei Pills may exert anti-cerebral ischemia effects through the Hippo signaling pathway.

Objective:To explore the clinical efficacy of 125I seeds implantation in treating cancer-induced pain and motor dysfunction caused by brachial plexus compression through neurophysiological monitoring. Methods:A retrospective study was conducted on 8 patients (4 males, 4 females; age 58-63 years) who underwent 125I seeds therapy for cancer-induced brachial plexus injury at Hebei Provincial People′s Hospital from January 2021 to August 2023. Pain severity was assessed by using the numerical rating scale (NRS) and motor function was evaluated by using the Fugl-Meyer (F-M) assessment. Electrophysiological monitoring was used to assess changes in sensory and motor branch conduction velocity (CV) of the musculocutaneous nerve, axillary nerve, median nerve, ulnar nerve, and radial nerve before and 3 months after treatment. Paired t-test was used for data analysis. Results:All 8 patients had moderate to severe pain (6 had motor dysfunction). The preoperative and postoperative NRS scores was 5.9±1.0 and 3.3±1.7, respectively ( t=4.93, P=0.002), while F-M scores was 44.8±7.6 and 54.8±5.7, respectively ( t=-3.52, P=0.017). Electrophysiological results showed that 7 patients had lesion involvement in the lower trunk of the brachial plexus, and 1 patient had involvement in the upper trunk. The preoperative and postoperative motor branch CV of the ulnar nerve was (47.2±2.6) and (59.7±8.2) m/s, respectively ( t=-3.17, P=0.034), while the sensory branch CV was (41.8±1.2) and (56.0±5.7) m/s, respectively ( t=-5.82, P=0.001). The nerve CV increased compared to the preoperative ones. Conclusions:125I seeds implantation has good clinical efficacy in treating cancer-related pain and motor dysfunction caused by brachial plexus compression. Changes in electrophysiology can quantitatively monitor the recovery of sensory and motor functions of the brachial plexus.