Oversampling effect in pretraining for bidirectional encoder representations from transformers (BERT) to localize medical BERT and enhance biomedical BERT.

BACKGROUND: Pretraining large-scale neural language models on raw texts has made a significant contribution to improving transfer learning in natural language processing. With the introduction of transformer-based language models, such as bidirectional encoder representations from transformers (BERT), the performance of information extraction from free text has improved significantly in both the general and medical domains. However, it is difficult to train specific BERT models to perform well in domains for which few databases of a high quality and large size are publicly available. OBJECTIVE: We hypothesized that this problem could be addressed by oversampling a domain-specific corpus and using it for pretraining with a larger corpus in a balanced manner. In the present study, we verified our hypothesis by developing pretraining models using our method and evaluating their performance. METHODS: Our proposed method was based on the simultaneous pretraining of models with knowledge from distinct domains after oversampling. We conducted three experiments in which we generated (1) English biomedical BERT from a small biomedical corpus, (2) Japanese medical BERT from a small medical corpus, and (3) enhanced biomedical BERT pretrained with complete PubMed abstracts in a balanced manner. We then compared their performance with those of conventional models. RESULTS: Our English BERT pretrained using both general and small medical domain corpora performed sufficiently well for practical use on the biomedical language understanding evaluation (BLUE) benchmark. Moreover, our proposed method was more effective than the conventional methods for each biomedical corpus of the same corpus size in the general domain. Our Japanese medical BERT outperformed the other BERT models built using a conventional method for almost all the medical tasks. The model demonstrated the same trend as that of the first experiment in English. Further, our enhanced biomedical BERT model, which was not pretrained on clinical notes, achieved superior clinical and biomedical scores on the BLUE benchmark with an increase of 0.3 points in the clinical score and 0.5 points in the biomedical score. These scores were above those of the models trained without our proposed method. CONCLUSIONS: Well-balanced pretraining using oversampling instances derived from a corpus appropriate for the target task allowed us to construct a high-performance BERT model.

Hybrid endovascular treatment for complicated aortic dissection concomitant with true lumen obliteration: a case report.

Background: Thoracic endovascular aortic repair (TEVAR) has been widely introduced. However, unestablished transfemoral approach due to true lumen obliteration disables endovascular option. Case summary: A 74-year-old male with a history of 15-year-ago type B aortic dissection presented with chronic bilateral lower extremity claudication. CT angiography revealed that a large entry tear was located at distal to the left subclavian artery. The thoracic aneurysmal degeneration progressed and eventually required repair. True lumen of infrarenal aorta to bilateral common iliac arteries was totally collapsed by false lumen, and the re-entry tear was open at external iliac artery. Initially, we performed recanalization to the collapsed true lumen. Bidirectional approach was taken from right brachial and bifemoral arteries. The covered endovascular reconstruction of aortic bifurcation (CERAB) technique and double D-shape moulding technique (DDMT) was performed to create covered stent configuration. As secondary treatment, 1-debranching TEVAR with axillary artery bypass was successfully performed by utilizing femoral approach. Discussion: This case demonstrated feasibility of two-stage endovascular therapy for thoracic aneurysmal degeneration concomitant with true lumen obliteration. This combined technique of CERAB and DDMT was absolutely effective to minimize type Ⅲ endoleak in infrarenal segment. Hybrid endovascular treatment offered minimally invasive therapy to the patient.

Development of the Clinical Specimen Information Management System for Multicenter Clinical Studies.

Some multicenter clinical studies require the acquisition of clinical specimens from patients, and the centralized management and analysis of clinical specimens at a research institution. In such cases, it is necessary to manage clinical specimens with anonymized patient information. In addition, clinical specimens need to be managed in connection with clinical information in clinical studies. In this study, we have developed a clinical specimen information management system that works with electronic data capture system for efficient specimen information management and the system workflow has verified at Osaka University Hospital. In addition, by combining this system with medical image collection system that we have developed previously, the integrated management of clinical information, medical image, and clinical specimen information will become possible. This specimen information management system may be expected to provide the platform for integrated analysis utilizing clinical information, medical image, and data from clinical specimens in multicenter clinical studies.

Classification of Diagnostic Certainty in Radiology Reports with Deep Learning.

A radiology report is prepared for communicating clinical information about observed abnormal structures and clinically important findings with referring clinicians. However, such observations and findings are often accompanied by ambiguous expressions, which can prevent clinicians from accurately interpreting the content of reports. To systematically assess the degree of diagnostic certainty for each observation and finding in a report, we defined an ordinal scale comprising five classes: definite, likely, may represent, unlikely, and denial. Furthermore, we applied a deep learning classification model to determine its applicability to in-house radiology reports. We trained and evaluated the model using 540 in-house chest computed tomography reports. The deep learning model achieved a micro F1-score of 97.61%, which indicated that our ordinal scale was suitable for measuring the diagnostic certainty of observations and findings in a report.

Implementation of a Multilingual Electronic Medical Questionnaire: For Use Inside or Outside the Hospital.

We implemented a multilingual medical questionnaire system, which allows patients to answer questionnaires both in and out of the hospital. The response data are sent to and stored as structured data on the server in hospital information system, and could be converted to Japanese and quoted as part of progress notes in the electronic medical record.

Impact of a Cancer History on Cardiovascular Events Among Patients With Myocardial Infarction Who Received Revascularization.

BACKGROUND: It remains controversial whether a cancer history increases the risk of cardiovascular (CV) events among patients with myocardial infarction (MI) who undergo revascularization.Methods and Results: Patients who were confirmed as type 1 acute MI (AMI) by coronary angiography were retrospectively analyzed. Patients who died in hospital or those not undergoing revascularization were excluded. Patients with a cancer history were compared with those without it. A cancer history was examined in the in-hospital cancer registry. The primary outcome was a composite of cardiac death, recurrent type 1 MI, post-discharge coronary revascularization, heart failure hospitalization, and stroke. Among 551 AMI patients, 55 had a cancer history (cancer group) and 496 did not (non-cancer group). Cox proportional hazards model revealed that the risk of composite endpoint was significantly higher in the cancer group than in the non-cancer group (adjusted hazard ratio [HR]: 1.78; 95% confidence interval [CI]: 1.13-2.82). Among the cancer group, patients who were diagnosed as AMI within 6 months after the cancer diagnosis had a higher risk of the composite endpoint than those who were diagnosed as AMI 6 months or later after the cancer diagnosis (adjusted HR: 5.43; 95% CI: 1.55-19.07). CONCLUSIONS: A cancer history increased the risk of CV events after discharge among AMI patients after revascularization.

Enhancing evidence-informed policymaking in medicine and healthcare: stakeholder involvement in the Commons Project for rare diseases in Japan.

BACKGROUND: Although stakeholder involvement in policymaking is attracting attention in the fields of medicine and healthcare, a practical methodology has not yet been established. Rare-disease policy, specifically research priority setting for the allocation of limited research resources, is an area where evidence generation through stakeholder involvement is expected to be effective. We generated evidence for rare-disease policymaking through stakeholder involvement and explored effective collaboration among stakeholders. METHODS: We constructed a space called 'Evidence-generating Commons', where patients, family members, researchers, and former policymakers can share their knowledge and experiences and engage in continual deliberations on evidence generation. Ten rare diseases were consequently represented. In the 'Commons', 25 consecutive workshops were held predominantly online, from 2019 to 2021. These workshops focused on (1) clarification of difficulties faced by rare-disease patients, (2) development and selection of criteria for priority setting, and (3) priority setting through the application of the criteria. For the first step, an on-site workshop using sticky notes was held. The data were analysed based on KJ method. For the second and third steps, workshops on specific themes were held to build consensus. The workshop agendas and methods were modified based on participants' feedback. RESULTS: The 'Commons' was established with 43 participants, resulting in positive effects such as capacity building, opportunities for interactions, mutual understanding, and empathy among the participants. The difficulties faced by patients with rare diseases were classified into 10 categories. Seven research topics were identified as priority issues to be addressed including 'impediments to daily life', 'financial burden', 'anxiety', and 'burden of hospital visits'. This was performed by synthesising the results of the application of the two criteria that were particularly important to strengthen future research on rare diseases. We also clarified high-priority research topics by using criteria valued more by patients and family members than by researchers and former policymakers, and criteria with specific perspectives. CONCLUSION: We generated evidence for policymaking in the field of rare diseases. This study's insights into stakeholder involvement can enhance evidence-informed policymaking. We engaged in comprehensive discussions with policymakers regarding policy implementation and planned analysis of the participants' experiences in this project.

Stakeholder involvement is significant for effective policymaking in the field of rare diseases. However, practical methods for this involvement have not yet been established. Therefore, we developed the 'Commons project' to generate valuable policymaking information and explore effective ways for stakeholders' collaboration. This article explains the process and results of 25 continuous workshops, held from 2019 to 2021 with 43 participants, including patients, family members, researchers, and former policymakers. The main achievements of the discussion that took place in the 'Commons' included a presentation of the overview of the difficulties faced by patients with rare diseases and formulation of high priority research topics.First, the difficulties faced by patients with rare diseases were grouped into 10 categories. Second, seven research topics were identified as priority issues including 'impediments to daily life', 'financial burden', 'anxiety', and 'burden of hospital visits'. During the project process, positive effects such as capacity building, opportunities for interactions, mutual understanding, and empathy among the participants, were identified. Beyond the context of the field of rare diseases and science of policy, these findings are useful for the future of society, including co-creation among stakeholders and patient and public involvement. Based on this study's results, we have initiated communications with policy stakeholders in the field of rare diseases, with the aim of policy implementation.

The observation of below-the-knee artery by optical frequency domain image and angioscopy: a case series.

Background: Endovascular treatment (EVT) is a well-established treatment for patients with chronic limb-threatening ischaemia, and below-the-knee (BTK) artery is its main target, although the re-intervention rate is still high. Understanding of the characteristics of BTK artery atherosclerosis would be required to overcome this issue. In this case series, we elucidated the characteristics of non-stenotic BTK artery atherosclerosis in the patients who received EVT of the superficial femoral artery (SFA) using optical frequency domain imaging (OFDI) and angioscopy. Case summary: We presented five patients who underwent EVT of SFA and subsequent observation of ipsilateral BTK artery using OFDI and angioscopy. Patients one and two had advanced atherosclerosis; however, patients three, four, and five had only mild atherosclerosis. Discussion: All patients had multiple risk factors for atherosclerosis and stenosis/occlusion of the SFA and ipsilateral BTK arteries. Furthermore, some patients had several other atherosclerotic vascular diseases suggesting the presence of advanced systemic atherosclerosis. On the other hand, some patients with multiple BTK artery stenosis/occlusion did not have advanced atherosclerosis in the examined BTK artery. The absence of significant atherosclerosis in a BTK artery in patients with multiple stenoses or occlusion in other ipsilateral BTK arteries may suggest some mechanism of vessel occlusion other than atherosclerosis. Further investigations are needed to clarify the mechanism.

Extracting Clinical Information From Japanese Radiology Reports Using a 2-Stage Deep Learning Approach: Algorithm Development and Validation.

Background: Radiology reports are usually written in a free-text format, which makes it challenging to reuse the reports. Objective: For secondary use, we developed a 2-stage deep learning system for extracting clinical information and converting it into a structured format. Methods: Our system mainly consists of 2 deep learning modules: entity extraction and relation extraction. For each module, state-of-the-art deep learning models were applied. We trained and evaluated the models using 1040 in-house Japanese computed tomography (CT) reports annotated by medical experts. We also evaluated the performance of the entire pipeline of our system. In addition, the ratio of annotated entities in the reports was measured to validate the coverage of the clinical information with our information model. Results: The microaveraged F1-scores of our best-performing model for entity extraction and relation extraction were 96.1% and 97.4%, respectively. The microaveraged F1-score of the 2-stage system, which is a measure of the performance of the entire pipeline of our system, was 91.9%. Our system showed encouraging results for the conversion of free-text radiology reports into a structured format. The coverage of clinical information in the reports was 96.2% (6595/6853). Conclusions: Our 2-stage deep system can extract clinical information from chest and abdomen CT reports accurately and comprehensively.

Impact of an Electronic Medical Record-Connected Questionnaire on Efficient Nursing Documentation: Usability and Efficacy Study.

BACKGROUND: Documentation tasks comprise a large percentage of nurses' workloads. Nursing records were partially based on a report from the patient. However, it is not a verbatim transcription of the patient's complaints but a type of medical record. Therefore, to reduce the time spent on nursing documentation, it is necessary to assist in the appropriate conversion or citation of patient reports to professional records. However, few studies have been conducted on systems for capturing patient reports in electronic medical records. In addition, there have been no reports on whether such a system reduces the time spent on nursing documentation. OBJECTIVE: This study aims to develop a patient self-reporting system that appropriately converts data to nursing records and evaluate its effect on reducing the documenting burden for nurses. METHODS: An electronic medical record-connected questionnaire and a preadmission nursing questionnaire were administered. The questionnaire responses entered by the patients were quoted in the patient profile for inpatient assessment in the nursing system. To clarify its efficacy, this study examined whether the use of the electronic questionnaire system saved the nurses' time entering the patient profile admitted between August and December 2022. It also surveyed the usability of the electronic questionnaire between April and December 2022. RESULTS: A total of 3111 (78%) patients reported that they answered the electronic medical questionnaire by themselves. Of them, 2715 (88%) felt it was easy to use and 2604 (85%) were willing to use it again. The electronic questionnaire was used in 1326 of 2425 admission cases (use group). The input time for the patient profile was significantly shorter in the use group than in the no-use group (P<.001). Stratified analyses showed that in the internal medicine wards and in patients with dependent activities of daily living, nurses took 13%-18% (1.3 to 2 minutes) less time to enter patient profiles within the use group (both P<.001), even though there was no difference in the amount of information. By contrast, in the surgical wards and in the patients with independent activities of daily living, there was no difference in the time to entry (P=.50 and P=.20, respectively), but there was a greater amount of information in the use group. CONCLUSIONS: The study developed and implemented a system in which self-reported patient data were captured in the hospital information network and quoted in the nursing system. This system contributes to improving the efficiency of nurses' task recordings.

Novel Algorithm for the Estimation of Cancer Incidence Using Claims Data in Japan: A Feasibility Study.

PURPOSE: We developed algorithms to identify patients with newly diagnosed cancer from a Japanese claims database to identify the patients with newly diagnosed cancer of the sample population, which were compared with the nationwide cancer incidence in Japan to assess the validity of the novel algorithms. METHODS: We developed two algorithms to identify patients with stomach, lung, colorectal, breast, and cervical cancers: diagnosis only (algorithm 1), and combining diagnosis, treatments, and medicines (algorithm 2). Patients with newly diagnosed cancer were identified from an anonymized commercial claims database (JMDC Claims Database) in 2017 with two inclusions/exclusion criteria: selecting all patients with cancer (extract 1) and excluding patients who had received cancer treatments in 2015 or 2016 (extract 2). We estimated the cancer incidence of the five cancer sites and compared it with the Japan National Cancer Registry incidence (calculated standardized incidence ratio with 95% CIs). RESULTS: The number of patients with newly diagnosed cancer ranged from 219 to 17,840 by the sites, algorithms, and exclusion criteria. Standardized incidence ratios were significantly higher in the JMDC Claims Database than in the national registry data for extract 1 and algorithm 1, extract 1 and algorithm 2, and extract 2 and algorithm 1. In extract 2 and algorithm 2, colorectal cancer in male and stomach, lung, and cervical cancers in females showed similar cancer incidence in the JMDC and national registry data. CONCLUSION: The novel algorithms are effective for extracting information about patients with cancer from claims data by using the combined information on diagnosis, procedures, and medicines (algorithm 2), with 2-year cancer-treatment history as an exclusion criterion (extract 2).

Identification of mild cognitive impairment subtypes predicting conversion to Alzheimer's disease using multimodal data.

Mild cognitive impairment (MCI) is a high-risk condition for conversion to Alzheimer's disease (AD) dementia. However, individuals with MCI show heterogeneous patterns of pathology and conversion to AD dementia. Thus, detailed subtyping of MCI subjects and accurate prediction of the patients in whom MCI will convert to AD dementia is critical for identifying at-risk populations and the underlying biological features. To this end, we developed a model that simultaneously subtypes MCI subjects and predicts conversion to AD and performed an analysis of the underlying biological characteristics of each subtype. In particular, a heterogeneous mixture learning (HML) method was used to build a decision tree-based model based on multimodal data, including cerebrospinal fluid (CSF) biomarker data, structural magnetic resonance imaging (MRI) data, APOE genotype data, and age at examination. The HML model showed an average F1 score of 0.721, which was comparable to the random forest method and had significantly more predictive accuracy than the CART method. The HML-generated decision tree was also used to classify-five subtypes of MCI. Each MCI subtype was characterized in terms of the degree of abnormality in CSF biomarkers, brain atrophy, and cognitive decline. The five subtypes of MCI were further categorized into three groups: one subtype with low conversion rates (similar to cognitively normal subjects); three subtypes with moderate conversion rates; and one subtype with high conversion rates (similar to AD dementia patients). The subtypes with moderate conversion rates were subsequently separated into a group with CSF biomarker abnormalities and a group with brain atrophy. The subtypes identified in this study exhibited varying MCI-to-AD conversion rates and differing biological profiles.

Relationships of Atrial Fibrillation at Diagnosis and Type of Atrial Fibrillation During Follow-up With Long-Term Outcomes for Heart Failure With Preserved Ejection Fraction.

Background: Few data are available regarding the impact of atrial fibrillation (AF) at diagnosis and type of AF during the follow-up period on long-term outcomes in patients with heart failure with preserved ejection fraction (HFpEF). Methods and Results: In all, 1,697 patients diagnosed as HFpEF between March 2010 and December 2017 were included in this study. At enrollment, 698 (41.1%) patients had AF. Over a median follow-up of 1,017 days, there were no significant differences between patients with and without AF in the adjusted hazard ratio (HR) for all-cause death or admission for heart failure. However, those with AF had a higher risk of stroke (HR 1.831; P=0.003). Of 998 patients with sinus rhythm at enrollment, 139 (13.9%) developed new-onset AF. Predictors of new-onset AF were pulse, hemoglobin, left ventricular end-diastolic dimension, and B-type natriuretic peptide. Compared with sinus rhythm, paroxysmal AF had a similar risk for all-cause death, admission for HF, and stroke; persistent AF had a lower risk of all-cause death (HR 0.701; P=0.015), but a higher risk for admission for HF (HR 1.608; P=0.002); and new-onset AF had a lower risk for all-cause death (HR 0.654; P=0.040), but a higher risk of admission for HF (HR 2.475; P<0.001). Conclusions: In patients with HFpEF, long-term outcome may differ by type of AF. Physicians need to consider individual risk with regard to AF type.

Prospects of Hospital Information Systems and Patient Safety in Japan.

OBJECTIVES: Approximately 20 years have passed since hospital information systems (HISs) featuring full-scale electronic medical records were first implemented in Japan. Patient safety is one of the most important of the several "safety" roles that HISs are expected to fulfill. However, insufficient research has analyzed the contribution of HISs to patient safety. This paper reviews the history of HISs in connection with patient safety in Japan and discusses the future of the patient safety function of HISs in a favorable environment for digitization. METHODS: A review on the history of HISs with functions that contribute to patient safety was conducted, analyzing evidence from reports published by the Japanese government and papers on patient safety and HISs published in various countries. RESULTS: Patient safety has become a concern, and initiatives to promote patient safety have progressed simultaneously with the spread of HISs. To address the problem of patient safety, most large hospitals prioritize patients' welfare when building HISs. However, no HIS-associated reduction in adverse events due to medical treatment could be confirmed. CONCLUSIONS: HISs are expected to help prevent medical accidents, such as patient- and drug-related errors. It is hoped that the patient safety functions of HISs will become generalized and contribute to patient safety in the future. To achieve this, the government and academic societies should provide regulations and guidelines on HISs and patient safety to the medical community and medical-device vendors. Furthermore, departments responsible for HISs and patient safety should collaborate to gather evidence for the effectiveness of HISs.

Temporary Rise in Blood Thrombogenicity in Patients with Acute Myocardial Infarction.

Objective Although blood thrombogenicity seems to be one of the determinant factors for the development of acute myocardial infarction (MI), it has not been dealt with in-depth. This study aimed to investigate blood thrombogenicity and its change in acute MI patients. Methods and Results We designed a prospective, observational study that included 51 acute MI patients and 83 stable coronary artery disease (CAD) patients who underwent cardiac catheterization, comparing thrombogenicity of the whole blood between: (1) acute MI patients and stable CAD patients; and (2) acute and chronic phase in MI patients. Blood thrombogenicity was evaluated by the Total Thrombus-Formation Analysis System (T-TAS) using the area under the flow pressure curve (AUC 30 ) for the AR-chip. Acute MI patients had significantly higher AUC 30 than stable CAD patients (median [interquartile range], 1,771 [1,585-1,884] vs. 1,677 [1,527-1,756], p = 0.010). Multivariate regression analysis identified acute MI with initial TIMI flow grade 0/1 as an independent determinant of high AUC 30 ( ß = 0.211, p = 0.013). In acute MI patients, AUC 30 decreased significantly from acute to chronic phase (1,859 [1,550-2,008] to 1,521 [1,328-1,745], p = 0.001). Conclusion Blood thrombogenicity was significantly higher in acute MI patients than in stable CAD patients. Acute MI with initial TIMI flow grade 0/1 was significantly associated with high blood thrombogenicity by multivariate analysis. In acute MI patients, blood thrombogenicity was temporarily higher in acute phase than in chronic phase.

Practical use of a multicenter clinical research support system connected to electronic medical records.

BACKGROUND: Electronic medical records (EMRs) are widely used, but in many cases, they are used within a network physically separated from the Internet. Multicenter clinical studies use Internet-connected electronic data capture (EDC) systems to collect data, where data entered into the EMR are manually transcribed into the EDC system. In addition, medical images for clinical research are also collected manually. Variations in EMRs and differing data structures among vendors hamper the use of data for clinical research. METHODS: We solved this problem by developing a network infrastructure for clinical research between Osaka University Hospital and affiliated hospitals in the Osaka area and introducing a clinical data collection system (CDCS). In each hospital's EMR network, we implemented a CRF reporter that accumulated data for clinical research using a template and then sent the data to a management server in the Osaka University Hospital Data Center. To organize the patient profile data and clinical laboratory data stored in each EMR for use in clinical research, the data are retrieved from the template by an interface module developed by each vendor, according to our common data output interface specification. The data entered into the CRF reporter template for clinical research are also recorded in the EMR progress notes and sent to the data management server. This network infrastructure can also be used as a medical image collection system that automatically collects images for research from PACS at each hospital. These systems are managed under common subject numbers issued by the CDCS. RESULTS: A network infrastructure was established among 19 hospitals, and a CRF reporter was incorporated into the EMR. A medical image transfer system was introduced in 13 hospitals. Since 2013, 28 clinical studies have been conducted using this system, and data for 9,987 cases have been collected as of December 31, 2020. CONCLUSION: Incorporating a CRF reporter with medical image transfer system into the EMR has proven useful for collecting research data.

Automatic creation of annotations for chest radiographs based on the positional information extracted from radiographic image reports.

BACKGROUND AND OBJECTIVE: In this study, we tried to create a machine-learning method that detects disease lesions from chest X-ray (CXR) images using a data set annotated with extracted CXR reports information. We set the nodule as the target disease lesion. Manually annotating nodules is costly in terms of time. Therefore, we used the report information to automatically produce training data for the object detection task. METHODS: First, we use semantic segmentation model PSP-Net to recognize lung fields described in the CXR reports. Next, a classification model ResNeSt-50 is used to discriminate the nodule in segmented right and left field. It also can provide attention map by Grad-Cam. If the attention region corresponds to the location of the nodule in the CXR reports, an attention bounding box is generated. Finally, object detection model Faster-RCNN was performed using generated attention bounding box. The bounding boxes predicted by Faster-RCNN were filtered to satisfy the location extracted from CXR reports. RESULTS: For lung field segmentation, a mean intersection of union of 0.889 was achieved in our best model. 15,156 chest radiographs are used for classification. The area under the receiver operating characteristics curve was 0.843 and 0.852 for the left and right lung, respectively. The detection precision of the generated attention bounding box was 0.341 to 0.531 depending on the binary setting for attention map. Through object detection process, the detection precisions of the bounding boxes were improved to 0.567 to 0.800. CONCLUSION: We successfully generated bounding boxes with nodule on CXR images based on the positional information of the diseases extracted from the CXR reports. Our method has the potential to provide bounding boxes for various lung lesions which can reduce the annotation burden for specialists. SHORT ABSTRACT: Machine learning for computer aided image diagnosis requires annotation of images, but manual annotation is time-consuming for medical doctor. In this study, we tried to create a machine-learning method that creates bounding boxes with disease lesions on chest X-ray (CXR) images using the positional information extracted from CXR reports. We set the nodule as the target lesion. First, we use PSP-Net to segment the lung field according to the CXR reports. Next, a classification model ResNeSt-50 was used to discriminate the nodule in segmented lung field. We also created an attention map using the Grad-Cam algorithm. If the area of attention matched the area annotated by the CXR report, the coordinate of the bounding box was considered as a possible nodule area. Finally, we used the attention information obtained from the nodule classification model and let the object detection model trained by all of the generated bounding boxes. Through object detection model, the precision of the bounding boxes to detect nodule is improved.

Detecting Adverse Drug Events Through the Chronological Relationship Between the Medication Period and the Presence of Adverse Reactions From Electronic Medical Record Systems: Observational Study.

BACKGROUND: Medicines may cause various adverse reactions. An enormous amount of money and effort is spent investigating adverse drug events (ADEs) in clinical trials and postmarketing surveillance. Real-world data from multiple electronic medical records (EMRs) can make it easy to understand the ADEs that occur in actual patients. OBJECTIVE: In this study, we generated a patient medication history database from physician orders recorded in EMRs, which allowed the period of medication to be clearly identified. METHODS: We developed a method for detecting ADEs based on the chronological relationship between the presence of an adverse event and the medication period. To verify our method, we detected ADEs with alanine aminotransferase elevation in patients receiving aspirin, clopidogrel, and ticlopidine. The accuracy of the detection was evaluated with a chart review and by comparison with the Roussel Uclaf Causality Assessment Method (RUCAM), which is a standard method for detecting drug-induced liver injury. RESULTS: The calculated rates of ADE with ALT elevation in patients receiving aspirin, clopidogrel, and ticlopidine were 3.33% (868/26,059 patients), 3.70% (188/5076 patients), and 5.69% (226/3974 patients), respectively, which were in line with the rates of previous reports. We reviewed the medical records of the patients in whom ADEs were detected. Our method accurately predicted ADEs in 90% (27/30patients) treated with aspirin, 100% (9/9 patients) treated with clopidogrel, and 100% (4/4 patients) treated with ticlopidine. Only 3 ADEs that were detected by the RUCAM were not detected by our method. CONCLUSIONS: These findings demonstrate that the present method is effective for detecting ADEs based on EMR data.

High neutrophil/lymphocyte ratio at cancer diagnosis predicts incidence of stroke in cancer patients.

Although cancer increases the incidence and severity of ischaemic stroke, there is no reliable method for predicting ischaemic stroke in cancer patients. To evaluate the prognostic capacity of the neutrophil-to-lymphocyte ratio at cancer diagnosis for predicting the incidence of ischaemic stroke, we used a hospital-based cancer registry that contained clinical data from all patients treated for cancer at Osaka University Hospital between 2007 and 2015. The neutrophil-to-lymphocyte ratio was calculated after dividing absolute neutrophil counts by absolute lymphocyte counts. These counts were obtained within 1 month after cancer diagnosis. The primary endpoint was new-onset ischaemic stroke within 2 years after cancer diagnosis. Of the 18 217 included cancer patients (median age: 65.2 years), 69 (0.38%) had ischaemic stroke. Unadjusted Cox regression analysis stratified by cancer site demonstrated that each 1-unit increase in the neutrophil-to-lymphocyte ratio was associated with a significant 7.2% increase in the risk of an ischaemic stroke event (95% confidence interval 1.041-1.103, P < 0.001). Survival tree analysis and the Kaplan-Meier method suggested that patients with and without atrial fibrillation who had increased neutrophil-to-lymphocyte ratios had a higher risk of ischaemic stroke. Multivariate Cox proportional hazard models, adjusted for cancer site and stage, revealed that patients with high neutrophil-to-lymphocyte ratios (>15) had higher ischaemic stroke risk than patients with low neutrophil-to-lymphocyte ratios (<5). This was true among cancer patients both with (hazard ratio 11.598; 95% confidence interval 0.953-141.181) and without (hazard ratio 7.877; 95% confidence interval 2.351-26.389) atrial fibrillation. The neutrophil-to-lymphocyte ratio at cancer diagnosis is associated with the incidence of ischaemic stroke among cancer patients and might thus be useful for identifying patients at high risk of ischaemic stroke, allowing us to guide future preventive interventions.