Search

Medical events are often infrequent, thus becomes hard to predict. In this paper, we focus on predictor that forecasts whether a medical event would occur in the next year, and analyzes the impact of event's frequency and data size via predictor's performance. In the experiment, we made 1572 predictors for medical events using Medical Insurance Claims (MICs) data from 800,000 participants and 205.8 million claims over 8 years. The result revealed that (a) forecasting error will be increased when predicting low-frequency events, and (b) increasing the number of training dataset reduces errors. This result suggests that increasing data size is a key to solve low frequency problems. However, we still need additional methods to cope with sparse and imbalanced data. [ABSTRACT FROM AUTHOR]

We analyzed the behavior of patient with a focus on patient-sharing based on the methodology of network analysis. We used an administrative healthcare claims database from September of the years 2008-2020 to identify shared patients with hypertension. The patients' behavior of visiting multiple medical facilities was extracted as graphical data, and we calculated density and centrality as indicators to evaluate the structure of the patient sharing network. Our findings indicate that density, reciprocity, and transitivity increased over time, and that centrality and PageRank were correlated. [ABSTRACT FROM AUTHOR]

Important pieces of information related to patient symptoms and diagnosis are often written in free-text form in clinical texts. To utilize these texts, information extraction using natural language processing is required. This study evaluated the performance of named entity recognition (NER) and relation extraction (RE) using machine-learning methods. The Japanese case report corpus was used for this study, which had 113 types of entities and 36 types of relations that were manually annotated. There were 183 cases comprising 2,194 sentences after preprocessing. In addition, a machine learning model based on bidirectional encoder representations from transformers was used. The results revealed that the maximum micro-averaged F1 scores of NER and RE were 0.912 and 0.759, respectively. The results of this study are comparable to those of previous studies. Hence, these results could be of substantial baseline accuracy. [ABSTRACT FROM AUTHOR]

We document the procedure and performance of a rule-based NLP system that, using transfer learning, automatically extracts essential named entities related to drug errors from Japanese free-text incident reports. Subsequently, we used the rule-based annotated data to fine-tune a pre-trained BERT model and examined the performance of medication-related incident report prediction. The rule-based pipeline achieved a macro-F1-score of 0.81 in an internal dataset and the BERT model fine-tuned with rule-annotated data achieved a macro-F1-score of 0.97 and 0.75 for named entity recognition and relation extraction tasks, respectively. The model can be deployed to other, similar problems in medication-related clinical texts. [ABSTRACT FROM AUTHOR]

This study examined the effectiveness of a systematic approach to the clinical management of COVID-19, focusing on nursing turnover. METHODS: Between 2017 and 2019, a clinical process support system based on structured clinical knowledge (Team Compass with the Patient Condition Adaptive Path System; TC-PCAPS) was developed, and implemented in hospitals. In 2020, the COVID-19 clinical management system (COVID-19-CMS) was developed. In this study, the effectiveness of implementing both systems was analyzed. The analysis covered hospitals N, T, and B, where TC-PCAPS implementation started in 2019, 2020, and 2022, respectively. Data for the period from 2018 to 2022 were collected and compared. RESULTS: Hospitals N and T implemented TC-PCAPS in the first year and the COVID-19-CMS in the following year. The nurse turnover rates of these hospitals were lower than those of the prefectures in which they were located. There was a trend towards a gradual reduction in nurse turnover. In contrast, hospital B, which had only just started to introduce these systems, saw a gradual increase in nurse turnover. CONCLUSION: The data collected from these three hospitals suggested that this systematic approach has the potential to reduce nurse turnover, in addition to the previously reported ability of TC-PCAPS to reduce nurse overtime. In Japan, there is a need to respond to future pandemics and reform the work styles of physicians and nurses. The abovementioned systematic approach has great potential for contributing to both of these aims. [ABSTRACT FROM AUTHOR]

Medical artificial intelligence (AI) systems need to learn to recognize synonyms or paraphrases describing the same anatomy, disease, treatment, etc. to better understand real-world clinical documents. Existing linguistic resources focus on variants at the word or sentence level. To handle linguistic variations on a broader scale, we proposed the Medical Text Radiology Report section Japanese version (MedTxt-RR-JA), the first clinical comparable corpus. MedTxt-RR-JA was built by recruiting nine radiologists to diagnose the same 15 lung cancer cases in Radiopaedia, an open-access radiological repository. The 135 radiology reports in MedTxt-RR-JA were shown to contain word-, sentence- and document-level variations maintaining similarity of contents. MedTxt-RR-JA is also the first publicly available Japanese radiology report corpus that would help to overcome poor data availability for Japanese medical AI systems. Moreover, our methodology can be applied widely to building clinical corpora without privacy concerns. [ABSTRACT FROM AUTHOR]