【摘要】：膿毒癥是感染引起的全身炎癥反應(yīng)綜合征。是患者死亡的主要原因之一。由于嚴(yán)重膿毒癥和膿毒性休克來勢兇猛,病情發(fā)展迅速,給臨床救治工作帶來極大挑戰(zhàn)。因此,盡早明確膿毒癥的診斷,及時治療是改善其預(yù)后、降低病死率之關(guān)鍵。針對當(dāng)前膿毒癥臨床診斷困難和對醫(yī)生的臨床經(jīng)驗有著較高要求,以及隱藏在大量電子病歷中的信息沒有得到充分利用的現(xiàn)狀,依據(jù)膿毒癥臨床數(shù)據(jù)特點提出了一種基于神經(jīng)網(wǎng)絡(luò)和因子分解機的數(shù)據(jù)挖掘算法結(jié)構(gòu),能夠充分利用電子病歷信息,輔助醫(yī)生進行臨床診療服務(wù),快速準(zhǔn)確地進行膿毒癥診斷。
[Abstract]:Sepsis is a systemic inflammatory response syndrome caused by infection. Is one of the main causes of death. Because of severe sepsis and septic shock, the disease develops rapidly, which brings great challenge to clinical treatment. Therefore, early diagnosis of sepsis and timely treatment are the key to improve prognosis and reduce mortality. In view of the current difficulties in clinical diagnosis of sepsis and the high demand for doctors' clinical experience, as well as the fact that the information hidden in a large number of electronic medical records has not been fully utilized, According to the characteristics of clinical data of sepsis, a data mining algorithm structure based on neural network and factor decomposition machine is proposed, which can make full use of electronic medical record information, assist doctors in clinical diagnosis and treatment, and make rapid and accurate diagnosis of sepsis.
【作者單位】： 中國科學(xué)院上海技術(shù)物理研究所紅外探測與成像技術(shù)重點實驗室;中國科學(xué)院大學(xué);上�？萍即髮W(xué);上海交通大學(xué)醫(yī)學(xué)院附屬上海兒童醫(yī)學(xué)中心PICU;
【分類號】：R720.597;TP183
，

本文編號：2300334