Shoko Wakamiya

Dr. Shoko Wakamiya


Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

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

Representative Publications

View All Publications
25 September 2018 in JMIR Public Health and Surveillance
Verified
Journal Article
Open access
SWShoko Wakamiya
Shoko Wakamiya
YKYukiko Kawai
Yukiko Kawai
EAEiji Aramaki
Eiji Aramaki
Background: The recent rise in popularity and scale of social networking services (SNSs) has resulted in an increasing need for SNS-based information extraction systems. A popular application of SNS data is health surveillance for predicting an outbreak of epidemics by detecting diseases from text messages posted on SNS platforms. Such applications share the following logic: they incorporate SNS users as social sensors. These social sensor–based approaches also share a common problem: SNS-based surveillance are much more reliable if sufficient numbers of users are active, and small or inactive populations produce inconsistent results. Objective: This study proposes a novel approach to estimate the trend of patient numbers using indirect information covering both urban areas and rural areas within the posts. Methods: We presented a TRAP model by embedding both direct information and indirect information. A collection of tweets spanning 3 years (7 million influenza-related tweets in Japanese) was used to evaluate the model. Both direct information and indirect information that mention other places were used. As indirect information is less reliable (too noisy or too old) than direct information, the indirect information data were not used directly and were considered as inhibiting direct information. For example, when indirect information appeared often, it was considered as signifying that everyone already had a known disease, leading to a small amount of direct information. Results: The estimation performance of our approach was evaluated using the correlation coefficient between the number of influenza cases as the gold standard values and the estimated values by the proposed models. The results revealed that the baseline model (BASELINE+NLP) shows .36 and that the proposed model (TRAP+NLP) improved the accuracy (.70, +.34 points). Conclusions: The proposed approach by which the indirect information inhibits direct information exhibited improved estimation performance not only in rural cities but also in urban cities, which demonstrated the effectiveness of the proposed method consisting of a TRAP model and natural language processing (NLP) classification.

18 June 2022 in BMC Medical Informatics and Decision Making
Verified
Journal Article
Open access
PDF file
FMFaith Wavinya Mutinda
Faith Wavinya Mutinda
KLKongmeng Liew
Kongmeng Liew
SYShuntaro Yada
Shuntaro Yada
SWShoko Wakamiya
Shoko Wakamiya
EAEiji Aramaki
Eiji Aramaki
Background: Meta-analyses aggregate results of different clinical studies to assess the effectiveness of a treatment. Despite their importance, meta-analyses are time-consuming and labor-intensive as they involve reading hundreds of research articles and extracting data. The number of research articles is increasing rapidly and most meta-analyses are outdated shortly after publication as new evidence has not been included. Automatic extraction of data from research articles can expedite the meta-analysis process and allow for automatic updates when new results become available. In this study, we propose a system for automatically extracting data from research abstracts and performing statistical analysis.Materials and methods: Our corpus consists of 1011 PubMed abstracts of breast cancer randomized controlled trials annotated with the core elements of clinical trials: Participants, Intervention, Control, and Outcomes (PICO). We proposed a BERT-based named entity recognition (NER) model to identify PICO information from research abstracts. After extracting the PICO information, we parse numeric outcomes to identify the number of patients having certain outcomes for statistical analysis.Results: The NER model extracted PICO elements with relatively high accuracy, achieving F1-scores greater than 0.80 in most entities. We assessed the performance of the proposed system by reproducing the results of an existing meta-analysis. The data extraction step achieved high accuracy, however the statistical analysis step achieved low performance because abstracts sometimes lack all the required information.Conclusion: We proposed a system for automatically extracting data from research abstracts and performing statistical analysis. We evaluated the performance of the system by reproducing an existing meta-analysis and the system achieved a relatively good performance, though more substantiation is required.

20 February 2019 in Journal of Medical Internet Research
Verified
Journal Article
Open access
SWShoko Wakamiya
Shoko Wakamiya
MMMizuki MORITA
Mizuki MORITA
YKYoshinobu Kano
Yoshinobu Kano
TOTomoko Ohkuma
Tomoko Ohkuma
EAEiji Aramaki
Eiji Aramaki
Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: The amount of medical and clinical-related information on the Web is increasing. Among the various types of information on the Web, social media-based data obtained directly from people are particularly valuable and garnering much attention. To encourage medical natural language processing research exploiting social media data, the NTCIR-13 MedWeb (Medical Natural Language Processing for Web Document) provides pseudo-Twitter messages in a cross-language and multi-label corpus, covering three languages (Japanese, English, and Chinese), and annotated with eight symptom labels (e.g., cold, fever, flu, and so on). Then, participants classify each tweet into one of two categories: those containing a patient’s symptom, and those that do not. Objective: We aim to present the results of groups participated in the Japanese subtask, the English subtask, and the Chinese subtask along with discussions, in order to clarify the issues that need to be resolved in the field of medical natural language processing. Methods: The performance of participant systems is assessed using the exact match accuracy, F-measure based on precision and recall, and Hamming loss. Results: In all, eight groups (19 systems) participated in the Japanese subtask, four groups (12 systems) participated in the English subtask, and two groups (six systems) participated in the Chinese subtask. The best system achieved .880 in exact match accuracy, .920 in F-measure, and .019 in Hamming loss. Conclusions: This paper presented and discussed the performance of systems participated in the NTCIR-13 MedWeb task. Because the MedWeb task settings can be formalized as the factualization of text, the achievement of this task could be applied directly to practical clinical applications.

1 June 2021 in Methods of Information in Medicine
Verified
Journal Article
Open access
PDF file
FMFaith Wavinya Mutinda
Faith Wavinya Mutinda
SYShuntaro Yada
Shuntaro Yada
SWShoko Wakamiya
Shoko Wakamiya
EAEiji Aramaki
Eiji Aramaki
Background Semantic textual similarity (STS) captures the degree of semantic similarity between texts. It plays an important role in many natural language processing applications such as text summarization, question answering, machine translation, information retrieval, dialog systems, plagiarism detection, and query ranking. STS has been widely studied in the general English domain. However, there exists few resources for STS tasks in the clinical domain and in languages other than English, such as Japanese. Objective The objective of this study is to capture semantic similarity between Japanese clinical texts (Japanese clinical STS) by creating a Japanese dataset that is publicly available. Materials We created two datasets for Japanese clinical STS: (1) Japanese case reports (CR dataset) and (2) Japanese electronic medical records (EMR dataset). The CR dataset was created from publicly available case reports extracted from the CiNii database. The EMR dataset was created from Japanese electronic medical records. Methods We used an approach based on bidirectional encoder representations from transformers (BERT) to capture the semantic similarity between the clinical domain texts. BERT is a popular approach for transfer learning and has been proven to be effective in achieving high accuracy for small datasets. We implemented two Japanese pretrained BERT models: a general Japanese BERT and a clinical Japanese BERT. The general Japanese BERT is pretrained on Japanese Wikipedia texts while the clinical Japanese BERT is pretrained on Japanese clinical texts. Results The BERT models performed well in capturing semantic similarity in our datasets. The general Japanese BERT outperformed the clinical Japanese BERT and achieved a high correlation with human score (0.904 in the CR dataset and 0.875 in the EMR dataset). It was unexpected that the general Japanese BERT outperformed the clinical Japanese BERT on clinical domain dataset. This could be due to the fact that the general Japanese BERT is pretrained on a wide range of texts compared with the clinical Japanese BERT.

27 September 2018 in JMIR Medical Informatics
Verified
Journal Article
Open access
MUMisa Usui
Misa Usui
EAEiji Aramaki
Eiji Aramaki
TITomohide Iwao
Tomohide Iwao
SWShoko Wakamiya
Shoko Wakamiya
TSTohru Sakamoto
Tohru Sakamoto
MMMayumi Mochizuki
Mayumi Mochizuki
Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Although methods of obtaining knowledge from texts written by healthcare professionals such as electronic medical records and discharge summaries have been studied, there are few reports analyzing free-text data on patients’ complaints in Japanese. Objective: This study aimed to establish a new method for extracting keywords from patients’ free descriptions accumulated in Japanese medical institutions. Methods: We developed a system that automatically annotates free-text data with the codes of the Tenth Revision of the International Statistical Classification of Diseases and Related Health Problems (ICD10) using electronic medication history data (target period: September 1, 2015 to August 31, 2016). The performance of the system was evaluated through comparisons with data manually annotated by healthcare workers. Results: The number of ICD10 codes extracted from 5,000 patient statements by healthcare workers was 2,348, while the system extracted 2,236 codes. Of those cases, 1,480 matched. Compared with manual extraction, the performance of the system was 0.66 in terms of precision, 0.63 in recall, and 0.65 for the F-measure. Conclusions: Our results suggested that the system was helpful for extracting and standardizing patient’s words related to symptoms from massive amounts of free-text data instead of manual work. After improving the extraction accuracy, we expect to utilize this system to detect the signals of adverse drug reactions from patients’ statements in the future.

20 February 2019 in Journal of Medical Internet Research
Verified
Journal Article
Open access
SWShoko Wakamiya
Shoko Wakamiya
SMShoji Matsune
Shoji Matsune
KOKimihiro none Okubo
Kimihiro none Okubo
EAEiji Aramaki
Eiji Aramaki
Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Health-related social media data are increasingly used in disease surveillance studies, which have demonstrated moderately high correlations between the number of social media posts and the number of patients. However, there is a need to understand the causal relationship between the behavior of social media users and the actual number of patients to increase the credibility of disease surveillance based on social media data. Objective: To clarify the causal relationships among pollen count, the posting behavior of social media users, and the number of SAR patients in the real world. Methods: This analysis was conducted using datasets of pollen count, tweet numbers, and SAR patient numbers from Kanagawa Prefecture, Japan. We examined daily pollen counts for Japanese cedar (the major cause of SAR in Japan) and hinoki cypress (which commonly complicates SAR) from February 1 to May 31, 2017. The daily numbers of tweets that included the keyword “kafunshō” (or SAR) were calculated between January 1 and May 31, 2017. Daily SAR patient numbers from January 1 to May 31, 2017 were obtained from three healthcare institutes that participated in the study. The Granger causality test was used to examine the causal relationships among pollen count, tweet numbers, and the number of SAR patients during the study period from February to May 2017. To determine if time-variant factors affect these causal relationships, we also analyzed the main SAR phase (February to April) when Japanese cedar trees actively produce and release pollen. Results: Increases in pollen count were found to cause increases in the number of tweets during the overall study period (P = .04), but not the main SAR phase (P = .05). In contrast, increases in pollen count were found to increase patient numbers in both the study period (P = .04) and the main SAR phase (P = .01). In addition, increases in the number of tweets also caused increases in patient numbers during the main SAR phase (P = .02), but not the overall study period (P = .89). Patient numbers did not affect the number of tweets in both the overall study period (P = .24) and the main SAR phase (P = .47). Conclusions: Understanding the causal relationships between these factors is an important step to increase the credibility of surveillance systems that use social media data. Further in-depth studies are needed to identify the determinants of social media posts described in this exploratory analysis.