AUTHOR=Deng Bo , Zhu Wenwen , Sun Xiaochuan , Xie Yanfeng , Dan Wei , Zhan Yan , Xia Yulong , Liang Xinyi , Li Jie , Shi Quanhong , Jiang Li 

TITLE=Development and Validation of an Automatic System for Intracerebral Hemorrhage Medical Text Recognition and Treatment Plan Output

JOURNAL=Frontiers in Aging Neuroscience

VOLUME=Volume 14 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/aging-neuroscience/articles/10.3389/fnagi.2022.798132

DOI=10.3389/fnagi.2022.798132

ISSN=1663-4365

ABSTRACT=The main purpose of the study was to explore a reliable way to automatically handle emergency case such as ICH. So an artificial intelligence system named H-system was designed, aiming to automatically recognizes medical text data of ICH patient and output the treatment plan. Furthermore, the efficiency and reliability of H-system were tested and analyzed. The H-system, which mainly based on a pretrained language model BERT (Bidirectional Encoder Representations from Transformers) and an expert module for logical judgment of extracted entities, was designed and founded by neurosurgeon and AI experts together. All emergency medical text data were from the neurosurgery emergency electronic medical record database (N-eEMRD) of the First Affiliated Hospital of Chongqing Medical University, Chongqing Emergency Medical Center, and Chongqing First people's Hospital, and the treatment plans of these ICH cases were divided into two types. A total of 1000 simulated ICH cases were randomly selected as training and validation set. After training and validating on simulated cases, real cases from three medical center were provided to test the efficiency of H-system. Doctors with 1 year and 5 years working experience on neurosurgery (Doctor-1Y and Doctor-5Y) were included to compare with H-system. Furthermore, the data of H-system, for instance sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), and the area under the receiver operating characteristics curve (AUC) were calculated and compared with Doctor-1Y and Doctor-5Y. In testing set, the accuracy of H-system's treatment plan was 88.55(88.16-88.94)%, the specificity was 85.71(84.99-86.43)%, and the sensitivity was 91.83(91.01-92.65)% (Table 2). The AUC value of H-system in testing set was 0.887(0.884-0.891). Furthermore, the time H-system spent on ICH case was significantly shorter than that of doctors with Doctor-1Y and Doctor-5Y. The accuracy and AUC of H-system were significantly higher than that of Doctor-1Y. In addition, the accuracy of H-system was more closed to that of Doctor-5Y. The H-system designed in the study can automatically recognize and analyze medical text data of ICH patient, and rapidly output accurate treatment plans with high efficiency. It may provide a reliable and novel way to automatically and rapidly handle emergency case such as ICH.