Objective To construct machine learning models to predict the incidence of linezolid-induced thrombocytopenia(LIT).Methods A total of 198 patients treated with linezolid in a hospital between January 2020 and March 2024 were retrospectively included.Firstly,the patients were divided into LIT and non-LIT groups,and the basic characteristics of the two groups were compared.Then,the variables with significant differences between the two groups were selected as potential risk factors to construct models for predicting LIT,including Logistic regression,decision tree and random forest models,and the prediction performance of the models was evaluated and compared.Results There were 52(26.3％)patients developed LIT during the treatment.The univariate analysis showed significant differences in linezolid trough concentration(Cmin),baseline platelet counts and creatinine clearance,the incidence of cerebrovascular disease,acute respiratory distress syndrome,and abdominal infection in patients with and without LIT.Among the three models built based on these variables,the random forest model has the best predictive performance.The results of variable importance analysis based on random forest model showed that Cmin,baseline platelet count and combined with acute respiratory distress syndrome had higher importance scores.Conclusions The random forest model has high accuracy in predicting the occurrence of LIT,and the risk of LIT is higher in patients with higher levels of linezolid exposure and lower baseline platelets.

Objective To construct machine learning models to predict the incidence of linezolid-induced thrombocytopenia(LIT).Methods A total of 198 patients treated with linezolid in a hospital between January 2020 and March 2024 were retrospectively included.Firstly,the patients were divided into LIT and non-LIT groups,and the basic characteristics of the two groups were compared.Then,the variables with significant differences between the two groups were selected as potential risk factors to construct models for predicting LIT,including Logistic regression,decision tree and random forest models,and the prediction performance of the models was evaluated and compared.Results There were 52(26.3％)patients developed LIT during the treatment.The univariate analysis showed significant differences in linezolid trough concentration(Cmin),baseline platelet counts and creatinine clearance,the incidence of cerebrovascular disease,acute respiratory distress syndrome,and abdominal infection in patients with and without LIT.Among the three models built based on these variables,the random forest model has the best predictive performance.The results of variable importance analysis based on random forest model showed that Cmin,baseline platelet count and combined with acute respiratory distress syndrome had higher importance scores.Conclusions The random forest model has high accuracy in predicting the occurrence of LIT,and the risk of LIT is higher in patients with higher levels of linezolid exposure and lower baseline platelets.

ObjectiveTo explore the application law of compound prescriptions containing Coptidis Rhizoma to diabetes and its complications based on data mining. MethodA total of 288 records on formulas containing Coptidis Rhizoma in the treatment of diabetes and its complications and the corresponding syndromes and symptoms were retrieved from the medical records of modern physicians， with 218 on the treatment of diabetes and 70 on the treatment of diabetes complications. The Traditional Chinese Medicine Inheritance Computing Platform （V3.0） was used or data mining on the compositions of the formulas. Moreover， it was employed for the frequency statistics， formula analysis， symptom analysis， association rules analysis， network visualization， and cluster analysis. ResultThe 218 formulas containing Coptidis Rhizoma for the treatment of diabetes involved 190 Chinese medicinals. A total of 12 syndromes showed the occurrence frequency ≥ 6， and the one with the highest frequency was deficiency of both qi and yin， followed by the syndrome of excessive heat in lung and stomach， syndrome of Yin deficiency and effulgent fire， and syndrome of excessive heat and fluid consumption. The dose of Coptidis Rhizoma for the treatment of diabetes ranged from 1.5-30 g， and the most frequently used doses were 10， 6， and 5 g. The medicinal was frequently applied for the traditional Chinese medicine symptoms of dry stool， dry mouth and tongue， and swift digestion with rapid hungering. Moreover， 26 core medicinal pairs， 17 association rules， 3 diagrams on medicinal relationship， and 3 core combinations were yielded. The 70 formulas containing Coptidis Rhizoma for the treatment of diabetes complications involved 184 Chinese medicinals， and the top 3 complications were diabetic nephropathy， diabetic peripheral neuropathy， and diabetic gastroenteropathy. The dose of Coptidis Rhizoma in the treatment of diabetes complications fell in the range of 2-15 g， and 10， 6， and 5 g were most frequently used. ConclusionFormulas containing Coptidis Rhizoma for the treatment of diabetes and its complications mainly target the syndrome of dampness-heat transforming into fire and the syndrome of excessive heat damaging yin. This study can serve as a reference for standard use of Coptidis Rhizoma in the prevention and treatment of diabetes and its complications.

Objective: The present study was conducted to investigate the response of peroxiredoxin 3 (PRX3) to oxidative stress induced by high-risk human papillomavirus (HPV). Methods: Sixty patients with cervical cancer were included and sixty patients with hysteromyoma were assigned as controls. Serum PRX3 was detected by enzyme-linked immunosorbent assay. The expression of PRX3 and oncoprotein E6 of HPV16 or HPV18 was examined in cervical cancer tissues by immunohistochemistry and in cervical cancer cells by Western blotting respectively. Results: Patients with invasive squamous cervical cancer showed higher level of serum PRX3 than control subjects with hysteromyoma. PRX3 expression was up-regulated and was positively associated with that of E6 of HPV16 or HPV18 in cervical cancer tissues. The correlation was confirmed in HPV-containing cervical cancer cell lines including CaSki, and HeLa. Conclusions Our result indicated a positive response of PRX3 to HPV-induced oxidative stress. Serum PRX3 might be a potential indicator of active amplification of high-risk HPV in cervical cancer cells.

Objective To summarize the features of multislice spiral computed tomography (MSCT) examination of gastrointestinal stromal tumors (GISTs),and investigate the relationship between predictors and risk of MSCT examination for GISTs.Methods The clinical data of 110 patients with GISTs who were admitted to the Tianjin Medical University Cancer Institute and Hospital from July 2011 to February 2014 were retrospectively analyzed.All the patients received 64-slices spiral CT (64S-SCT) or 16-slices spiral CT (16S-SCT) scan,and the data were transported to the PACS work station for multiplanar reconstruction.All the tumor samples were collected during operation and diagnosed by morphological manifestation and immunohistochemistry of tumors.Very low,low,and medium risk of GISTs were regarded as lower risk grade,and high risk of GISTs as high risk grade.The univariate analysis and multivariate analysis about features of imaging and risk were done by chi-square test and multivariate logistic regression model.Results Tumors located at the stomach in 81 cases,small intestines in 26 cases and colorectum in 3 cases.Diameter of tumors was 0.8-25.0 cm.Smaller tumors were in round or oval shape with well demarcated boundary,and larger tumors were irregular with unclear boundary.Endo-luminal growth of lessions was detected in 25 cases,duplex growth in 35 cases and extra-luminal growth in 50 cases.Enhanced CT scan showed that most of tumors in 105 patients demostrated moderate and high enhancement,heterogeneous enhancement in 74 cases,low density sacvariable necrosis area without enhancement in 60 cases and superficial,cracked-like and deep ulcer without calcification,metastasis and ascites in 23 cases.According to the features of GISTs by MSCT examination,location of tumor,diameter,shape,boundary,growth,enhancement,cystic necrosis,ulcer and metastasis were risk factors affecting risk classification of tumors by univariate analysis (x2=7.442,49.966,31.513,46.038,13.836,16.626,23.489,8.280,6.811,P ＜0.05).Diameter of tumor more than 10 cm and ulcer were independent risk factors affecting risk classification of tumors by multivariate analysis (OR =9.927,0.070 ; 95％ confidence intewal:1.888-52.180,0.012-0.398,P ＜ 0.05).Conclusion There is a characterization in the location,diameter,shape,boundary of tumor,growth,enhancement,cystic necrosis,ulcer and metastasis,and diameter of tumor more than 10cm and ulcer are independent risk factors affecting the risk classification of GISTs.