Protected Health Information Recognition by Fine-Tuning a Pre-training Transformer Model / 대한의료정보학회지

Objectives@#De-identifying protected health information (PHI) in medical documents is important, and a prerequisite to deidentification is the identification of PHI entity names in clinical documents. This study aimed to compare the performance of three pre-training models that have recently attracted significant attention and to determine which model is more suitable for PHI recognition. @*Methods@#We compared the PHI recognition performance of deep learning models using the i2b2 2014 dataset. We used the three pre-training models—namely, bidirectional encoder representations from transformers (BERT), robustly optimized BERT pre-training approach (RoBERTa), and XLNet (model built based on Transformer-XL)—to detect PHI. After the dataset was tokenized, it was processed using an inside-outside-beginning tagging scheme and WordPiecetokenized to place it into these models. Further, the PHI recognition performance was investigated using BERT, RoBERTa, and XLNet. @*Results@#Comparing the PHI recognition performance of the three models, it was confirmed that XLNet had a superior F1-score of 96.29%. In addition, when checking PHI entity performance evaluation, RoBERTa and XLNet showed a 30% improvement in performance compared to BERT. @*Conclusions@#Among the pre-training models used in this study, XLNet exhibited superior performance because word embedding was well constructed using the two-stream self-attention method. In addition, compared to BERT, RoBERTa and XLNet showed superior performance, indicating that they were more effective in grasping the context.

ANNO: A General Annotation Tool for Bilingual Clinical Note Information Extraction / 대한의료정보학회지

Objectives@#This study was conducted to develop a generalizable annotation tool for bilingual complex clinical text annotation, which led to the design and development of a clinical text annotation tool, ANNO. @*Methods@#We designed ANNO to enable human annotators to support the annotation of information in clinical documents efficiently and accurately. First, annotations for different classes (word or phrase types) can be tagged according to the type of word using the dictionary function. In addition, it is possible to evaluate and reconcile differences by comparing annotation results between human annotators. Moreover, if the regular expression set for each class is updated during annotation, it is automatically reflected in the new document. The regular expression set created by human annotators is designed such that a word tagged once is automatically labeled in new documents. @*Results@#Because ANNO is a Docker-based web application, users can use it freely without being subjected to dependency issues. Human annotators can share their annotation markups as regular expression sets with a dictionary structure, and they can cross-check their annotated corpora with each other. The dictionary-based regular expression sharing function, cross-check function for each annotator, and standardized input (Microsoft Excel) and output (extensible markup language [XML]) formats are the main features of ANNO. @*Conclusions@#With the growing need for massively annotated clinical data to support the development of machine learning models, we expect ANNO to be helpful to many researchers.

Stacking Ensemble Technique for Classifying Breast Cancer / 대한의료정보학회지

OBJECTIVES: Breast cancer is the second most common cancer among Korean women. Because breast cancer is strongly associated with negative emotional and physical changes, early detection and treatment of breast cancer are very important. As a supporting tool for classifying breast cancer, we tried to identify the best meta-learner model in a stacking ensemble when the same machine learning models for the base learner and meta-learner are used. METHODS: We used machine learning models, such as the gradient boosted model, distributed random forest, generalized linear model, and deep neural network in a stacking ensemble. These models were used to construct a base learner, and each of them was used as a meta-learner again. Then, we compared the performance of machine learning models in the meta-learner to determine the best meta-learner model in the stacking ensemble. RESULTS: Experimental results showed that using the GBM as a meta-learner led to higher accuracy than that achieved with any other model for breast cancer data and using the GLM as a meta learner led to low root-mean-squared error for both sets of breast cancer data. CONCLUSIONS: We compared the performance of every meta-learner model in a stacking ensemble as a supporting tool for classifying breast cancer. The study showed that using specific models as a metalearner resulted in better performance than single classifiers, and using GBM and GLM as a meta-learner is appropriate as a supporting tool for classifying breast cancer data.

Statistics and Deep Belief Network-Based Cardiovascular Risk Prediction / 대한의료정보학회지

OBJECTIVES: Cardiovascular predictions are related to patients' quality of life and health. Therefore, a risk prediction model for cardiovascular conditions is needed. METHODS: In this paper, we propose a cardiovascular disease prediction model using the sixth Korea National Health and Nutrition Examination Survey (KNHANES-VI) 2013 dataset to analyze cardiovascular-related health data. First, statistical analysis was performed to find variables related to cardiovascular disease using health data related to cardiovascular disease. Second, a model of cardiovascular risk prediction by learning based on the deep belief network (DBN) was developed. RESULTS: The proposed statistical DBN-based prediction model showed accuracy and an ROC curve of 83.9% and 0.790, respectively. Thus, the proposed statistical DBN performed better than other prediction algorithms. CONCLUSIONS: The DBN proposed in this study appears to be effective in predicting cardiovascular risk and, in particular, is expected to be applicable to the prediction of cardiovascular disease in Koreans.

Development of Child-Teen Obesity Treatment Service Platform / 대한의료정보학회지

OBJECTIVES: This study aimed to develop an effective and efficient obesity treatment and management service platform for obese children/teenagers. METHODS: The integrated smart platform was planned and established through cooperation with service providers such as hospitals and public health centers, obese children/teenagers who constitute the service's user base, and IT development and policy institutions and companies focusing on child-teen obesity management and treatment. RESULTS: Based on guidelines on intervention strategies to manage child-teen obesity, we developed two patient/parent mobile applications, one web-monitoring service for medical staff, one mobile application for food-craving endurance, and one mobile application for medical examinations. CONCLUSIONS: The establishment of the integrated service platform was successfully completed; however, this study was restrictively to the hospital where the pilot program took place. The effectiveness of the proposed platform will be verified in the future in tests involving other organizations.

Data-Mining-Based Coronary Heart Disease Risk Prediction Model Using Fuzzy Logic and Decision Tree / 대한의료정보학회지

OBJECTIVES: The importance of the prediction of coronary heart disease (CHD) has been recognized in Korea; however, few studies have been conducted in this area. Therefore, it is necessary to develop a method for the prediction and classification of CHD in Koreans. METHODS: A model for CHD prediction must be designed according to rule-based guidelines. In this study, a fuzzy logic and decision tree (classification and regression tree [CART])-driven CHD prediction model was developed for Koreans. Datasets derived from the Korean National Health and Nutrition Examination Survey VI (KNHANES-VI) were utilized to generate the proposed model. RESULTS: The rules were generated using a decision tree technique, and fuzzy logic was applied to overcome problems associated with uncertainty in CHD prediction. CONCLUSIONS: The accuracy and receiver operating characteristic (ROC) curve values of the propose systems were 69.51% and 0.594, proving that the proposed methods were more efficient than other models.

Korean Anaphora Recognition System to Develop Healthcare Dialogue-Type Agent / 대한의료정보학회지

OBJECTIVES: Anaphora recognition is a process to identify exactly which noun has been used previously and relates to a pronoun that is included in a specific sentence later. Therefore, anaphora recognition is an essential element of a dialogue agent system. In the current study, all the merits of rule-based, machine learning-based, semantic-based anaphora recognition systems were combined to design and realize a new hybrid-type anaphora recognition system with an optimum capacity. METHODS: Anaphora recognition rules were encoded on the basis of the internal traits of referred expressions and adjacent contexts to realize a rule-based system and to serve as a baseline. A semantic database, related to predicate instances of sentences including referred expressions, was constructed to identify semantic co-relationships between the referent candidates (to which semantic tags were attached) and the semantic information of predicates. This approach would upgrade the anaphora recognition system by reducing the number of referent candidates. Additionally, to realize a machine learning-based system, an anaphora recognition model was developed on the basis of training data, which indicated referred expressions and referents. The three methods were further combined to develop a new single hybrid-based anaphora recognition system. RESULTS: The precision rate of the rule-based systems was 54.9%. However, the precision rate of the hybrid-based system was 63.7%, proving it to be the most efficient method. CONCLUSIONS: The hybrid-based method, developed by the combination of rule-based and machine learning-based methods, represents a new system with enhanced functional capabilities as compared to other pre-existing individual methods.

A Case of Systemic Lupus Erythematosus Presenting as a Large Rectosigmoid Ulcer / 대한내과학회지

Systemic lupus erythematosus (SLE) is a multi-system inflammatory disorder that has many symptoms. Gastrointestinal symptoms are common, while colonic involvement in the form of ischemic colitis or a colonic ulcer is rare in SLE. The differential diagnosis of ischemic proctitis with ulceration includes an infected ulcer, ulcerative colitis, Crohn's disease, solitary rectal ulcer colitis, malignant tumor, and lupus colitis. Here, we report a 22-year-old male with abdominal pain and diarrhea, who had a huge rectal ulcer that nearly obstructed the rectosigmoid junction. This turned out to be a rare gastrointestinal manifestation of lupus. He recovered after being treated with high-dose oral steroids. Our case demonstrates that a rectal ulcer is a rare, but important, complication of SLE and can be the initial clinical manifestation of the disease.

A Case of Systemic Lupus Erythematosus Presenting as a Large Rectosigmoid Ulcer / 대한내과학회지

Systemic lupus erythematosus (SLE) is a multi-system inflammatory disorder that has many symptoms. Gastrointestinal symptoms are common, while colonic involvement in the form of ischemic colitis or a colonic ulcer is rare in SLE. The differential diagnosis of ischemic proctitis with ulceration includes an infected ulcer, ulcerative colitis, Crohn's disease, solitary rectal ulcer colitis, malignant tumor, and lupus colitis. Here, we report a 22-year-old male with abdominal pain and diarrhea, who had a huge rectal ulcer that nearly obstructed the rectosigmoid junction. This turned out to be a rare gastrointestinal manifestation of lupus. He recovered after being treated with high-dose oral steroids. Our case demonstrates that a rectal ulcer is a rare, but important, complication of SLE and can be the initial clinical manifestation of the disease.

Lysophosphatidylcholine Increases Ca2+ Current via Activation of Protein Kinase C in Rabbit Portal Vein Smooth Muscle Cells

Lysophosphatidylcholine (LPC), a metabolite of membrane phospholipids by phospholipase A(2), has been considered responsible for the development of abnormal vascular reactivity during atherosclerosis. Ca2+ influx was shown to be augmented in atherosclerotic artery which might be responsible for abnormal vascular reactivity. However, the mechanism underlying Ca2+ influx change in atherosclerotic artery remains undetermined. The purpose of the present study was to examine the effects of LPC on L-type Ca2+ current (ICa(L)) activity and to elucidate the mechanism of LPC-induced change of ICa(L) in rabbit portal vein smooth muscle cells using whole cell patch clamp. Extracellular application of LPC increased ICa(L) through whole test potentials, and this effect was readily reversed by washout. Steady state voltage dependency of activation or inactivation properties of ICa(L) was not significantly changed by LPC. Staurosporine (100 nanometer) or chelerythrine (3 micrometer, which is a potent inhibitor of PKC, significantly decreased basal ICa(L), and LPC-induced increase of ICa(L) was significantly suppressed in the presence of PKC inhibitors. On the other hand, application of PMA, an activator of PKC, increased basal ICa(L) significantly, and LPC-induced enhancement of ICa(L) was abolished by pretreatment of the cells with PMA. These findings suggest that LPC increased ICa(L) in vascular smooth muscle cells by a pathway that involves PKC, and that LPC-induced increase of ICa(L) might be, at least in part, responsible for increased Ca2+ influx in atherosclerotic artery.

Elevated Contractile Responses to Acetylcholine in Organ Cultured Rabbit Carotid Artery

The aim of the present study was to examine the functional changes that occur when a rabbit carotid artery is cultured in serum-free medium. In endothelium (EC)-intact arteries cultured under serum-free conditions, acetylcholine (ACh)-induced relaxation responses were partially, yet significantly, reduced when compared with freshly isolated arteries. After pretreatment with N(G)-nitro-L-arginine methyl ester (L-NAME), a nitric oxide synthase inhibitor, application of ACh resulted in a significant contraction in organ cultured arteries. The amplitude of the ACh-induced contractions increased with the duration of culture. In EC-denuded arteries cultured under serum-free conditions, ACh induced responses similar to those in EC-intact arteries pretreated with L-NAME. Furthermore, ACh caused a significant increase in intracellular Ca2+ concentration ([Ca2+]i) in EC-denuded arteries cultured under serum-free condition for 7 days. There was little change in either [Ca2+]i or tension in freshly isolated carotid rings. There was no difference in sodium nitroprusside-induced relaxation responses between fresh and cultured arteries. These results suggest that prolonged culture of carotid arteries under serum-free conditions changes the functional properties of vascular reactivity in rabbit carotid arteries.