Emiko Shinohara

Emiko Shinohara

Auto-generated profile, based on publicly available information

Emiko Shinohara

Artificial Intelligence and Digital Twin in Healthcare, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 July 2022 in JMIR Medical Informatics
Verified
Journal Article
Open access
YKYoshimasa Kawazoe
Yoshimasa Kawazoe
KSKiminori Shimamoto
Kiminori Shimamoto
DSDaisaku Shibata
Daisaku Shibata
ESEmiko Shinohara
Emiko Shinohara
HKHideaki Kawaguchi
Hideaki Kawaguchi
TYTomotaka Yamamoto
Tomotaka Yamamoto
Background: Falls may cause elderly people to be bedridden, requiring professional intervention; thus, fall prevention is crucial. The use of electronic health records (EHRs) is expected to provide highly accurate risk assessment and length-of-stay data related to falls, which may be used to estimate the costs and benefits of prevention. However, no studies to date have investigated the extent to which hospital stays could be shortened through fall avoidance resulting from the use of prediction tools. Objective: We first estimated the extended length of hospital stay caused by falls among elderly inpatients. Next, we developed a model that predicts falls using clinical text as input and evaluated its accuracy. Finally, we estimated the potentially shortened hospital stay that would be made possible by appropriate interventions based on the prediction model. Methods: Patients aged 65 years or older were selected as subjects, and the EHRs of 1728 falls and 70,586 nonfalls were subjected to analysis. The extended-stay lengths were estimated using propensity score matching of 49 associated variables. Bidirectional encoder representations from transformers and bidirectional long short-term memory methods were used to predict falls from clinical text. The estimated length of stay and the outputs of the prediction model were used to determine stay reductions. Results: The extended length of hospital stay due to falls was estimated to be 17.8 days (95% CI 16.6-19.0), which dropped to 8.6 days when there were unobserved covariates at an odds ratio of 2.0. The accuracy of the prediction model was as follows: area under the receiver operating characteristic curve, 0.851; F-value, 0.165; recall, 0.737; precision, 0.093; and specificity, 0.839. When assuming interventions with 25% or 100% effectiveness against cases where the model predicted a fall, the stay reduction was estimated at 0.022 and 0.099 days/day, respectively. Conclusions: The accuracy of the prediction model using clinical text is considered to be higher than the prediction accuracy of conventional assessments. However, our model’s precision remained low at 9.3%. This may be due, in part, to the inclusion of cases in which falls did not occur because of preventative interventions during hospitalization. Nonetheless, it is estimated that interventions for cases when falls were predicted will reduce medical costs by 886 Yen/day (~US $6.50/day) of intervention, even if the preventative effect is 25%. Limitations include the fact that these results cannot be extrapolated to short- or long-term hospitalization cases, and that this was a single-center study.

19 December 2025 in International Journal of Medical Informatics
Verified
Journal Article
Open access
ESEmiko Shinohara
Emiko Shinohara
YKYoshimasa Kawazoe
Yoshimasa Kawazoe
Named entity recognition (NER) is critical in natural language processing (NLP), particularly in the medical field, where accurate identification of entities, such as patient information and clinical events, is essential. Traditional NER approaches rely heavily on large, annotated corpora, which are resource intensive. Large language models (LLMs) offer new NER approaches, particularly through in-context and few-shot learning. This study investigates the effects of incorporating annotation guidelines into prompts for NER via LLMs, with a specific focus on their impact on few-shot learning performance across various medical corpora. We designed eight different prompt patterns, combining few-shot examples with annotation guidelines of varying complexity, and evaluated their performance via three prominent LLMs: GPT-4o, Claude 3.5 Sonnet, and gpt-oss-120b. Additionally, we employed three diverse medical corpora: i2b2-2014, i2b2-2012, and MedTxt-CR. Accuracy was assessed via precision, recall, and the F1 score, with evaluation methods aligned with those used in relevant shared tasks to ensure the comparability of the results. Our findings indicate that adding detailed annotation guidelines to few-shot prompts improves the recall and F1 score in most cases. Including annotation guidelines in prompts enhances the performance of LLMs in NER tasks, making this a practical approach for developing accurate NLP systems in resource-constrained environments. Although annotation guidelines are essential for evaluation and example creation, their integration into LLM prompts can further optimize few-shot learning, especially within specialized domains such as medical NLP.

25 January 2024
Verified
Book Chapter
Open access
PDF file
DSDaisaku Shibata
Daisaku Shibata
ESEmiko Shinohara
Emiko Shinohara
KSKiminori Shimamoto
Kiminori Shimamoto
YKYoshimasa Kawazoe
Yoshimasa Kawazoe
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.

1 August 2025 in JCO Clinical Cancer Informatics
Verified
Journal Article
Open access
MTMasami Tsuchiya
Masami Tsuchiya
YKYoshimasa Kawazoe
Yoshimasa Kawazoe
KSKiminori Shimamoto
Kiminori Shimamoto
TSTomohisa Seki
Tomohisa Seki
SIShungo Imai 
Shungo Imai 
SHSatoko Hori
Satoko Hori
PURPOSE: Capecitabine, an oral anticancer agent, frequently causes hand-foot syndrome (HFS), affecting patients' quality of life and treatment adherence. However, such symptomatic toxicities are often difficult to detect in structured electronic health record (EHR) data. This study primarily aimed to validate a natural language processing (NLP) approach to identifying capecitabine-induced HFS from unstructured clinical text and demonstrate its application in evaluating medication-associated adverse event trends in real-world settings. METHODS: We conducted a retrospective cohort study using EHRs from the University of Tokyo Hospital (2004-2021). HFS cases were identified using the MedNERN-CR-JA NLP model. After propensity score matching, we compared capecitabine users with and without celecoxib and assessed time to HFS onset using Cox proportional hazards models. NLP-based HFS detection was validated through manual annotation of aggregated clinical notes. Negative control and sensitivity analyses ensured robustness. RESULTS: Among 44,502 patients with cancer, 669 capecitabine users were analyzed. HFS incidence was significantly higher among capecitabine users (hazard ratio [HR], 1.93 [95% CI, 1.48 to 2.52]; P < .001) compared with nonusers. Celecoxib use showed a suggestive association with a reduced HFS risk (HR, 0.51 [95% CI, 0.24 to 1.07]; P = .073). The NLP model demonstrated high accuracy in identifying HFS, achieving a precision of 0.875, recall of 1.000, and F 1 score of 0.933, based on manual annotation of patient-level clinical notes. Outcome trends remained consistent when using manually annotated HFS case labels instead of NLP-detected events, supporting the method's robustness. CONCLUSION: These findings demonstrate the effectiveness of NLP in detecting HFS from real-world clinical records. The application to celecoxib-HFS detection illustrates the potential utility of this approach for retrospective safety analysis. Further work is needed to evaluate generalizability across diverse clinical settings.

5 September 2023 in JMIR Formative Research
Verified
Journal Article
Open access
MHMai Hirabayashi
Mai Hirabayashi
DSDaisaku Shibata
Daisaku Shibata
ESEmiko Shinohara
Emiko Shinohara
YKYoshimasa Kawazoe
Yoshimasa Kawazoe
Background: As of December 2022, the outbreak of COVID-19 showed no sign of abating, continuing to impact people’s lives, livelihoods, economies, and more. Vaccination is an effective way to achieve mass immunity. However, in places such as Japan, where vaccination is voluntary, there are people who choose not to receive the vaccine, even if an effective vaccine is offered. To promote vaccination, it is necessary to clarify what kind of information on social media can influence attitudes toward vaccines. Objective: False rumors and counterrumors are often posted and spread in large numbers on social media, especially during emergencies. In this paper, we regard tweets that contain questions or point out errors in information as counterrumors. We analyze counterrumors tweets related to the COVID-19 vaccine on Twitter. We aimed to answer the following questions: (1) what kinds of COVID-19 vaccine–related counterrumors were posted on Twitter, and (2) are the posted counterrumors related to social conditions such as vaccination status? Methods: We use the following data sets: (1) counterrumors automatically collected by the “rumor cloud” (18,593 tweets); and (2) the number of COVID-19 vaccine inoculators from September 27, 2021, to August 15, 2022, published on the Prime Minister’s Office’s website. First, we classified the contents contained in counterrumors. Second, we counted the number of COVID-19 vaccine–related counterrumors from data set 1. Then, we examined the cross-correlation coefficients between the numbers of data sets 1 and 2. Through this verification, we examined the correlation coefficients for the following three periods: (1) the same period of data; (2) the case where the occurrence of the suggestion of counterrumors precedes the vaccination (negative time lag); and (3) the case where the vaccination precedes the occurrence of counterrumors (positive time lag). The data period used for the validation was from October 4, 2021, to April 18, 2022. Results: Our classification results showed that most counterrumors about the COVID-19 vaccine were negative. Moreover, the correlation coefficients between the number of counterrumors and vaccine inoculators showed significant and strong positive correlations. The correlation coefficient was over 0.7 at −8, −7, and −1 weeks of lag. Results suggest that the number of vaccine inoculators tended to increase with an increase in the number of counterrumors. Significant correlation coefficients of 0.5 to 0.6 were observed for lags of 1 week or more and 2 weeks or more. This implies that an increase in vaccine inoculators increases the number of counterrumors. These results suggest that the increase in the number of counterrumors may have been a factor in inducing vaccination behavior. Conclusions: Using quantitative data, we were able to reveal how counterrumors influence the vaccination status of the COVID-19 vaccine. We think that our findings would be a foundation for considering countermeasures of vaccination.

7 August 2025
Verified
Book Chapter
Open access
MTMasami Tsuchiya
Masami Tsuchiya
KSKiminori Shimamoto
Kiminori Shimamoto
YKYoshimasa Kawazoe
Yoshimasa Kawazoe
ESEmiko Shinohara
Emiko Shinohara
SYShuntaro Yada
Shuntaro Yada
EAEiji Aramaki
Eiji Aramaki
This study assessed the effectiveness of natural language processing (NLP) in detecting adverse events (AEs) from anticancer agents by analyzing data from over 39,000 cancer patients. A specialized machine learning model identified known AEs from anticancer agents like capecitabine, oxaliplatin, and anthracyclines, revealing a significantly higher incidence in the treatment groups compared to non-users. While the NLP approach effectively detected most symptomatic AEs requiring manual review, it struggled with rarely documented conditions and commonly used clinical terms. Overall, the method shows promise for automated AE detection in medical records, particularly for symptoms without laboratory markers or diagnosis codes.