1.Applications and prospects of machine learning in perioperative transfusion medicine
Rui FAN ; Xiaoying ZHANG ; Weiwei SHANG ; Wenfei TANG ; Haimei MA
Chinese Journal of Blood Transfusion 2025;38(10):1450-1456
This paper systematically reviews the application progress of machine learning in perioperative transfusion medicine, focusing on its significant achievements in identifying transfusion risk factors, accurately predicting transfusion requirements, and enabling dynamic monitoring with real-time feedback. It also examines the methodologies, performance metrics, and clinical significance of constructing machine learning models across various surgical specialties, including orthopaedics, cardiac surgery, trauma, and obstetrics. The review further analyzes major challenges currently facing the field, including data bias, model overfitting and interpretability issues, alongside privacy and ethical concerns. Finally, it outlines future directions, highlighting how multimodal data fusion, deep learning applications, multicentre validation, and interdisciplinary collaboration are poised to significant potential for advancing the clinical translation of intelligent transfusion models, achieve personalized precision transfusion management, and enhance patient safety and therapeutic outcomes.
2.Preparation and Identification of High Immunogenic A/PR/8/34 Maternal Strain HA Protein for Influenza Virus Classical Reassortment.
Jing TANG ; Li XIN ; Junfeng GUO ; Wenfei ZHU ; Heyuan ZHANG ; Shaohui LANG ; Dayan WANG ; Yuelong SHU
Chinese Journal of Virology 2016;32(2):141-144
Preparation of maternal strain A/PR/8/34 HA antiserum for influenza virus classical reassortment. A/PR/8/34 virus was digested by bromelain after inactivation and purification. 5%-20% sucrose continuous density gradient centrifugation method was used to purify HA protein. SIRD method was used to select the target protein. SDS-PAGE method was used to identified HA protein. High Immunogenic A/PR/8/34 HA protein was successfully prepared and HI titer reached 10240. High purity HA antiserum was identified by SIRD method. The key reagent in the classical reassortment of influenza virus was prepared, and the complete set of technical methods were explored, which laid the foundation for the independent research and development of seasonal influenza vaccine strains of China.
Animals
;
Antibodies, Viral
;
immunology
;
Electrophoresis, Polyacrylamide Gel
;
Female
;
Hemagglutination Inhibition Tests
;
Hemagglutinin Glycoproteins, Influenza Virus
;
analysis
;
immunology
;
Humans
;
Influenza A Virus, H1N1 Subtype
;
genetics
;
immunology
;
Influenza, Human
;
immunology
;
virology
;
Rabbits
;
Reassortant Viruses
;
genetics
;
immunology
3.Determination of 10 Residual Organic Solvents in Norvancomycin Hydrochloride Raw Material by HS-GC
Wenfei GENG ; Minghao ZUO ; Xiaonan ZHANG ; Mi TANG ; Xuexia ZHANG ; Jie TONG
China Pharmacy 2015;(21):2977-2979
OBJECTIVE:To establish the method for determining 10 residual organic solvents in norvancomycin hydrochloride raw material. METHODS:Headspace gas chromatography was performed on the column of nitro modified polyethylene terephthal-ate glycol as stationary phase capillary column;the oven temperature program started at 40 ℃ for 3 min and increased at a rate of 8 ℃/min up to 150 ℃ for 10 min;the temperature was 200 ℃ with carrier gas of high-purity nitrogen gas,the constant flow rate was 5 ml/min with split ratio of 15∶1;the headspace vial equilibrium temperature was 85 ℃ with equilibrium time of 40 min,and the volume was 1 ml. RESULTS:The concentration of n-pentane,acetone,ethanol,benzene,acrylonitrile,toluene,xylene,chlo-robenzene,styrene,divinylbenzene had good linear relationship with its peak area values(r=0.995 7-0.999 9);the RSDs of preci-sion,repeatability tests was ≤6.6%;average recovery was in the range of 94.3%-106.6%(RSD=0.5%-4.5%,n=9). CONCLU-SIONS:The method is fast,sensitive and accurate,and can be used for the determination of residual organic solvents in norvanco-mycin hydrochloride raw material.
4.Prognostic Analysis of Skull-base Invasion of Nasopharyngeal Carcinoma Based on Magnetic Resonance Imaging
Lei CHEN ; Wenfei LI ; Lizhi LIU ; Yanping MAO ; Linglong TANG ; Ying SUN ; Aihua LIN ; Li LI ; Jun MA
Journal of Sun Yat-sen University(Medical Sciences) 2010;31(2):258-264
[Objective]To evaluate the prognostic value of skull-base invasion of nasopharyngeal carcinoma(NPC)based on magnetic resonance imaging(MRI).[Methods]A total of 924 patients who were diagnosed with NPC between 2003 and 2004,had undergone MRI scan and received mdiothempy as their primary treatment,and had no distant metastasis were included in this study.MRI images and medical records were analyzed retrospectively.All the 924 eases.patients who developed skull-base invasions based on MRI,315 patients with T3 disease and 227 patients with T2 disease were selected for analysis.The staging was according to the sixth edition of the American Joint Commission on Cancer(AJCC)staging system.[Results]Incidence of skullbase invasion according to MRI was 55.4%.Of 924 cases.skull-base invasion on MRI was not an independent prognostic factor for overall survival(OS)and distant metastasis-free survival(DMFS),but was a marginally significant independent prognostic factor for local relapse-free survival(LRFS),P=0.068.Grading of MRI-detected skull-base erosion according to the site of invasion was an independent prognostic factor for OS(P=0.002 and P=0.005)and DMFS(P=0.001 for both)in the 512 patients with skull-base invasions and 315 patients with T3 disease.Severe-grade of skull-base invasion on MRI was an independent prognostic factor for OS and DMFS in the 924 patients(P < 0.001 for both).No significant differences were observed on OS,LRFS,and DMFS between T2a patients and T3 patients with low-grade of MRI-deteeted skull-base involvement.[Conclusions]Skull-base invasion based on MRI is not an independent prognostic factor for NPC.However,severe-grade of invasion according to the site of involvement has positive prognostic value.

Result Analysis
Print
Save
E-mail