A deep learning model for the detection of various dementia and MCI pathologies based on resting-state electroencephalography data: A retrospective multicentre study

Dementia and mild cognitive impairment (MCI) represent significant health challenges in an aging population. As the search for noninvasive, precise and accessible diagnostic methods continues, the efficacy of electroencephalography (EEG) combined with deep convolutional neural networks (DCNNs) in varied clinical settings remains unverified, particularly for pathologies underlying MCI such as Alzheimer's disease (AD), dementia with Lewy bodies (DLB) and idiopathic normal-pressure hydrocephalus (iNPH). Addressing this gap, our study evaluates the generalizability of a DCNN trained on EEG data from a single hospital (Hospital #1). For data from Hospital #1, the DCNN achieved a balanced accuracy (bACC) of 0.927 in classifying individuals as healthy (n = 69) or as having AD, DLB, or iNPH (n = 188). The model demonstrated robustness across institutions, maintaining bACCs of 0.805 for data from Hospital #2 (n = 73) and 0.920 at Hospital #3 (n = 139). Additionally, the model could differentiate AD, DLB, and iNPH cases with bACCs of 0.572 for data from Hospital #1 (n = 188), 0.619 for Hospital #2 (n = 70), and 0.508 for Hospital #3 (n = 139). Notably, it also identified MCI pathologies with a bACC of 0.715 for Hospital #1 (n = 83), despite being trained on overt dementia cases instead of MCI cases. These outcomes confirm the DCNN's adaptability and scalability, representing a significant stride toward its clinical application. Additionally, our findings suggest a potential for identifying shared EEG signatures between MCI and dementia, contributing to the field's understanding of their common pathophysiological mechanisms.

Sketch-based semantic retrieval of medical images

The volume of medical images stored in hospitals is rapidly increasing; however, the utilization of these accumulated medical images remains limited. Existing content-based medical image retrieval (CBMIR) systems typically require example images, leading to practical limitations, such as the lack of customizable, fine-grained image retrieval, the inability to search without example images, and difficulty in retrieving rare cases. In this paper, we introduce a sketch-based medical image retrieval (SBMIR) system that enables users to find images of interest without the need for example images. The key concept is feature decomposition of medical images, which allows the entire feature of a medical image to be decomposed into and reconstructed from normal and abnormal features. Building on this concept, our SBMIR system provides an easy-to-use two-step graphical user interface: users first select a template image to specify a normal feature and then draw a semantic sketch of the disease on the template image to represent an abnormal feature. The system integrates both types of input to construct a query vector and retrieves reference images. For evaluation, ten healthcare professionals participated in a user test using two datasets. Consequently, our SBMIR system enabled users to overcome previous challenges, including image retrieval based on fine-grained image characteristics, image retrieval without example images, and image retrieval for rare cases. Our SBMIR system provides on-demand, customizable medical image retrieval, thereby expanding the utility of medical image databases.

Development of a Diagnostic Artificial Intelligence Tool for Lateral Lymph Node Metastasis in Advanced Rectal Cancer

BACKGROUND

Metastatic lateral lymph node dissection can improve survival in patients with rectal adenocarcinoma, with or without chemoradiotherapy. However, the optimal imaging diagnostic criteria for lateral lymph node metastases remain undetermined.

OBJECTIVE

To develop a lateral lymph node metastasis diagnostic artificial intelligence tool using deep learning, for patients with rectal adenocarcinoma who underwent radical surgery and lateral lymph node dissection.

DESIGN

Retrospective study.

SETTINGS

Multicenter study.

PATIENTS

A total of 209 patients with rectal adenocarcinoma, who underwent radical surgery and lateral lymph node dissection at 15 participating hospitals, were enrolled in the study and allocated to training (n = 139), test (n = 17), or validation (n = 53) cohorts.

MAIN OUTCOME MEASURES

In the neoadjuvant treatment group, images taken before pretreatment were classified as baseline images and those taken after pretreatment as presurgery images. In the upfront surgery group, presurgery images were classified as both baseline and presurgery images. We constructed 2 types of artificial intelligence, using baseline and presurgery images, by inputting the patches from these images into ResNet-18, and we assessed their diagnostic accuracy.

RESULTS

Overall, 124 patients underwent surgery alone, 52 received neoadjuvant chemotherapy, and 33 received chemoradiotherapy. The number of resected lateral lymph nodes in the training, test, and validation cohorts was 2418, 279, and 850, respectively. The metastatic rates were 2.8%, 0.7%, and 3.7%, respectively. In the validation cohort, the precision-recall area under the curve was 0.870 and 0.963 for the baseline and presurgery images, respectively. Although both baseline and presurgery images provided good accuracy for diagnosing lateral lymph node metastases, the accuracy of presurgery images was better than that of baseline images.

LIMITATIONS

The number of cases is small.

CONCLUSIONS

An artificial intelligence tool is a promising tool for diagnosing lateral lymph node metastasis with high accuracy.

DESARROLLO DE UNA HERRAMIENTA DE INTELIGENCIA ARTIFICIAL PARA EL DIAGNSTICO DE METSTASIS EN GANGLIOS LINFTICOS LATERALES EN CNCER DE RECTO AVANZADO

ANTECEDENTES:Disección de nódulos linfáticos laterales metastásicos puede mejorar la supervivencia en pacientes con adenocarcinoma del recto, con o sin quimiorradioterapia. Sin embargo, aún no se han determinado los criterios óptimos de diagnóstico por imágenes de los nódulos linfáticos laterales metastásicos.OBJETIVO:Nuestro objetivo fue desarrollar una herramienta de inteligencia artificial para el diagnóstico de metástasis en nódulos linfáticos laterales mediante el aprendizaje profundo, para pacientes con adenocarcinoma del recto que se sometieron a cirugía radical y disección de nódulos linfáticos laterales.DISEÑO:Estudio retrospectivo.AJUSTES:Estudio multicéntrico.PACIENTES:Un total de 209 pacientes con adenocarcinoma del recto, que se sometieron a cirugía radical y disección de nódulos linfáticos laterales en 15 hospitales participantes, se inscribieron en el estudio y se asignaron a cohortes de entrenamiento (n = 139), prueba (n = 17) o validación (n = 53).PRINCIPALES MEDIDAS DE RESULTADO:En el grupo de tratamiento neoadyuvante, las imágenes tomadas antes del tratamiento se clasificaron como imágenes de referencia y las posteriores al tratamiento, como imágenes previas a la cirugía. En el grupo de cirugía inicial, las imágenes previas a la cirugía se clasificaron como imágenes de referencia y previas a la cirugía. Construimos dos tipos de inteligencia artificial, utilizando imágenes de referencia y previas a la cirugía, ingresando los parches de estas imágenes en ResNet-18. Evaluamos la precisión diagnóstica de los dos tipos de inteligencia artificial.RESULTADOS:En general, 124 pacientes se sometieron a cirugía solamente, 52 recibieron quimioterapia neoadyuvante y 33 recibieron quimiorradioterapia. El número de nódulos linfáticos laterales removidos en los cohortes de entrenamiento, prueba y validación fue de 2,418; 279 y 850, respectivamente. Las tasas metastásicas fueron 2.8%, 0.7%, y 3.7%, respectivamente. En el cohorte de validación, el área de recuperación de precisión bajo la curva fue de 0.870 y 0.963 para las imágenes de referencia y antes de la cirugía, respectivamente. Aunque tanto las imágenes previas a la cirugía como las iniciales proporcionaron una buena precisión para diagnosticar metástasis en los nódulos linfáticos laterales, la precisión de las imágenes previas a la cirugía fue mejor que la de las imágenes iniciales.LIMITACIONES:El número de casos es pequeño.CONCLUSIÓN:La inteligencia artificial es una herramienta prometedora para diagnosticar metástasis en los nódulos linfáticos laterales con alta precisión. (Traducción-Dr. Aurian Garcia Gonzalez ).

Spherical Image Generation From a Few Normal-Field-of-View Images by Considering Scene Symmetry

Spherical images taken in all directions (360 degrees by 180 degrees) can represent an entire space including the subject, providing free direction viewing and an immersive experience to viewers. It is convenient and expands the usage scenarios to generate a spherical image from a few normal-field-of-view (NFOV) images, which are partial observations. The primary challenge is generating a plausible image and controlling the high degree of freedom involved in generating a wide area that includes all directions. We focus on scene symmetry, which is a basic property of the global structure of spherical images, such as the rotational and plane symmetries. We propose a method for generating a spherical image from a few NFOV images and controlling the generated regions using scene symmetry. We incorporate the intensity of the symmetry as a latent variable into conditional variational autoencoders to estimate the possible range of symmetry and decode a spherical image whose features are represented through a combination of symmetric transformations of the NFOV image features. Our experiments show that the proposed method can generate various plausible spherical images controlled from asymmetrically to symmetrically, and can reduce the reconstruction errors of the generated images based on the estimated symmetry.

Improved Object Detection Artificial Intelligence Using the Revised RetinaNet Model for the Automatic Detection of Ulcerations, Vascular Lesions, and Tumors in Wireless Capsule Endoscopy

The use of computer-aided detection models to diagnose lesions in images from wireless capsule endoscopy (WCE) is a topical endoscopic diagnostic solution. We revised our artificial intelligence (AI) model, RetinaNet, to better diagnose multiple types of lesions, including erosions and ulcers, vascular lesions, and tumors. RetinaNet was trained using the data of 1234 patients, consisting of images of 6476 erosions and ulcers, 1916 vascular lesions, 7127 tumors, and 14,014,149 normal tissues. The mean area under the receiver operating characteristic curve (AUC), sensitivity, and specificity for each lesion were evaluated using five-fold stratified cross-validation. Each cross-validation set consisted of between 6,647,148 and 7,267,813 images from 217 patients. The mean AUC values were 0.997 for erosions and ulcers, 0.998 for vascular lesions, and 0.998 for tumors. The mean sensitivities were 0.919, 0.878, and 0.876, respectively. The mean specificities were 0.936, 0.969, and 0.937, and the mean accuracies were 0.930, 0.962, and 0.924, respectively. We developed a new version of an AI-based diagnostic model for the multiclass identification of small bowel lesions in WCE images to help endoscopists appropriately diagnose small intestine diseases in daily clinical practice.

Precise Discrimination for Multiple Etiologies of Dementia Cases Based on Deep Learning with Electroencephalography

INTRODUCTION

It is critical to develop accurate and universally available biomarkers for dementia diseases to appropriately deal with the dementia problems under world-wide rapid increasing of patients with dementia. In this sense, electroencephalography (EEG) has been utilized as a promising examination to screen and assist in diagnosing dementia, with advantages of sensitiveness to neural functions, inexpensiveness, and high availability. Moreover, the algorithm-based deep learning can expand EEG applicability, yielding accurate and automatic classification easily applied even in general hospitals without any research specialist.

METHODS

We utilized a novel deep neural network, with which high accuracy of discrimination was archived in neurological disorders in the previous study. Based on this network, we analyzed EEG data of healthy volunteers (HVs, N = 55), patients with Alzheimer's disease (AD, N = 101), dementia with Lewy bodies (DLB, N = 75), and idiopathic normal pressure hydrocephalus (iNPH, N = 60) to evaluate the discriminative accuracy of these diseases.

RESULTS

High discriminative accuracies were archived between HV and patients with dementia, yielding 81.7% (vs. AD), 93.9% (vs. DLB), 93.1% (vs. iNPH), and 87.7% (vs. AD, DLB, and iNPH).

CONCLUSION

This study revealed that the EEG data of patients with dementia were successfully discriminated from HVs based on a novel deep learning algorithm, which could be useful for automatic screening and assisting diagnosis of dementia diseases.

Effectiveness of patient and staff cohorting to reduce the risk of vancomycin-resistant enterococcus (VRE) acquisition: a retrospective cohort study during a VRE outbreak in Japan

BACKGROUND

Patient and staff cohorting is part of a bundle approach in the response to multi-drug-resistant organisms, but its effectiveness is not fully clarified. This study compared the risks of acquiring vancomycin-resistant Enterococcus faecium (VREfm) at a hospital during a VREfm outbreak based on contact characteristics in order to better understand the effectiveness of cohorting.

METHODS

Exposure came from contact with patients with VREfm (infectors), including existing patients with VREfm and patients who acquired VREfm during the study period. Contact was defined as length of contact time, degree of sharing space, and care by the same nurses as those caring for infectors between January and March 2018. The outcome was VREfm acquisition as determined through monthly stool or rectal screening cultures. Incidence rates were calculated based on contact patterns, and incidence rate ratios (IRRs) were compared.

FINDINGS

Among 272 inpatients (4038 patient-days), 43 patients acquired VREfm with the same or similar pulsotype. Incidence rates were 8.45 per 1000 patient-days when susceptible inpatients were on the same ward as an infector but cared for by different nurses (reference), 16.96 when susceptible inpatients were on the same ward as an infector and cared for by the same nurses [IRR 2.01, 95% confidence interval (CI) 0.62-10.28], and 52.91 when susceptible inpatients shared a room with an infector (IRR 6.26, 95% CI 1.61-35.40).

CONCLUSION

Compared with susceptible inpatients in a different room from infectors and not being cared for by the same nurses, the risk of VREfm acquisition could be six times higher for susceptible inpatients who are in the same room as infectors, and could be double for susceptible inpatients cared for by the same nurses as infectors.

AB0382 COMPARISON OF ADVERSE EVENTS (AEs) RELATED TO MAJOR ANTI-RHEUMATIC DRUGS, REPORTED TO THE OFFICIAL JAPANESE ADVERSE DRUG EVENT REPORT DATABASE (JADER)

Background

Currently, many disease-modifying anti-rheumatic drugs (DMARDs) are available for the treatment of rheumatoid arthritis (RA). Among them, methotrexate (MTX), biologic DMARDs (bDMARDs) and Janus kinase inhibitors (JAKi) are the major options. AEs related to these are major concerns. In Japan, the AEs data spontaneously reported to and summarized by Pharmaceuticals and Medical Devices Agency (PMDA) are freely accessible.

Objectives

The major AEs relating to MTX, bDMARDs, and JAKi observed in the real world were compared.

Methods

The number of AEs listed by JADER from 2014 to 2020 was collected. The AEs were classified by System Organ Class (SOC) of Medical Dictionary for Regulatory Activities (MedDRA) and compared using the chi-square test. The bDMARDs included were etanercept (ETN), adalimumab (ADA), golimumab (GOL), tocilizumab (TCZ), and abatacept (ABT), and JAKi was tofacitinib (TOF) and baricitinib (BAR).

Results

The total number of AEs was 27,604. The number was significantly increasing in total and MTX, GOL, TOF during these years, although the number of cases that have each DMARD is not known in this study. The proportion of SOCs among each DMARD was similar throughout the period.

The most frequent was infections/infestations in all DMARDs except for MTX, in which neoplasms were the most. The primary cause of infection was bacterial, including pneumonia. Varicella-zoster virus infection in JAKi, and tuberculosis in ADA and GOL were conspicuous.

Neoplasms were the second major in many DMARDs. Lymphoproliferative disorders were most common in MTX-related neoplasms, whereas solid tumors were more in other DMARDs like ABT or BAR.

Other SOCs include all other categories such as the musculoskeletal, nervous system, cardiac, and vascular disorders. Among them, major adverse cardiovascular events (MACE, including cardiovascular death, myocardial infarction and stroke) and venous thromboembolism (VTE) were both reported in small numbers. However, more MACE was noted in BAR and GOL, and more VTE was in BAR and TOF compared to other DMARDs.

Conclusion

The number of AEs cases related to DMARDs was increasing. Significant difference among AEs related to DMARDs was noted in the JADER database, especially regarding MTX and JAKi.

References

[1]S. Inokuma. Expert Open Drug Saf. 2021 Nov 11. Online ahead of print.

Table 1.

Total case numbers and proportion of adverse events related to each DMARD.

DMARDs (year of launch)TotalMTX (1999)ETN (2005)ADA (2008)GOL (2011)TCZ (2008)ABT (2010)TOF (2013)BAR (2017)Number of AEs2760411636297419591302424517093065714Blood/lymphatic system disorders5.89.92.03.22.64.00.92.83.5Gastrointestinal disorders5.04.13.47.55.97.73.95.24.2General disorders/administration site reactions4.63.211.34.12.23.04.07.62.5Infections/infestations28.020.321.332.234.236.537.137.049.2 Pneumonia (bacterial)6.64.06.15.210.97.712.59.415.1 Other bacterial infection9.15.95.612.39.517.69.19.810.6 Herpes zoster2.11.10.41.41.21.01.28.211.2 Tuberculosis1.40.91.96.03.50.60.60.31.0Investigations5.04.95.52.32.65.72.28.32.8Neoplasms benign, malignant, unspecified21.134.811.013.512.77.415.610.815.4 Lymphoproliferative diseases14.329.82.84.14.02.24.52.12.2 Solid tumors5.63.87.37.87.54.09.97.310.5Respiratory, thoracic, mediastinal disorders7.16.88.38.49.56.57.46.26.7Other System Organ Class23.315.937.328.830.229.228.922.215.7 Major adverse cardiovascular events1.10.41.11.62.51.51.61.32.9 Venous thromboembolism0.40.10.40.50.60.30.21.11.4

Background colors indicate: comparing to the total cases, higher with p<0.05, light pink; higher with p<0.00001, dark pink; higher with p<1E-10, red. Lower with p<0.05, light blue; lower with p<0.00001, blue; lower with p<1E-10, dark blue, using chi-square test. Yellow indicates System Organ Classes.

Figure 1.

Total number of adverse events reported from 2014 to 2020.

Linear regression is shown only for DMARDs with increasing numbers of AEs.

Disclosure of Interests

None declared

Precise Measurement of Differential Cross Sections of the Σ^{-}p→Λn Reaction in Momentum Range 470-650 MeV/c

The differential cross sections of the Σ^{-}p→Λn reaction were measured accurately for the Σ^{-} momentum (p_{Σ}) ranging from 470 to 650  MeV/c at the J-PARC Hadron Experimental Facility. Precise angular information about the Σ^{-}p→Λn reaction was obtained for the first time by detecting approximately 100 reaction events at each angular step of Δcosθ=0.1. The obtained differential cross sections show a slightly forward-peaking structure in the measured momentum regions. The cross sections integrated for -0.7≤cosθ≤1.0 were obtained as 22.5±0.68 [statistical error(stat.)] ±0.65 [systematic error(syst.)] mb and 15.8±0.83(stat)±0.52(syst)  mb for 470<p_{Σ}(MeV/c)<550 and 550<p_{Σ}(MeV/c)<650, respectively. These results show a drastic improvement compared with past measurements of the hyperon-proton scattering experiments. They will play essential roles in updating the theoretical models of the baryon-baryon interactions.

Two-dimensional echocardiographic scoring system of the left ventricular filling pressure and clinical outcomes in heart failure with preserved ejection fraction

Funding Acknowledgements

Type of funding sources: Public Institution(s). Main funding source(s): Japan Society for the Promotion of Science (JSPS)

Introduction

Elevated left ventricular (LV) filling pressure in non-decompensated state is a powerful indicator of worse clinical outcomes in heart failure regardless of LV ejection fraction. However, its detection is often challenging in heart failure with preserved ejection fraction (HFpEF).

Purpose

This study aimed to elucidate the predictive value of recently proposed echocardiographic parameter of LV filling pressure, Visually assessed time difference between the Mitral valve and Tricuspid valve opening (VMT) score in HFpEF.

Methods

We retrospectively analyzed 310 well-differentiated HFpEF patients in stable conditions. Using two-dimensional echocardiographic images, time sequence of opening of mitral valve and tricuspid valve was visually assessed in the apical four-chamber view and scored to 0 to 2 (0: tricuspid valve first, 1: simultaneous, 2: mitral valve first). When the inferior vena cava diameter was dilated, 1 point was added and VMT score was calculated as four grades from 0 to 3. Based on the previous study, VMT≥2 was regarded as a sign of elevated LV filling pressure (Figure 1). LV diastolic function was graded according to the guidelines. The primary endpoint was defined as a composite of cardiac death and heart failure hospitalisation during the two years after echocardiographic examination.

Results

During the follow-up period, 55 events (18%) occurred, including four cardiac deaths and 51 heart failure hospitalisations. Kaplan-Meier curves demonstrated that VMT≥2 (n = 54) was associated with worse outcomes compared to patients showing VMT ≤ 1 (n = 256) (log-rank test P &lt;0.001). Furthermore, VMT≥2 was associated with worse outcomes when tested in 100 HFpEF patients with atrial fibrillation (log-rank test P = 0.026) (Figure 2). In the adjusted model including age, systolic blood pressure, serum albumin level, and the LV diastolic function grading, VMT≥2 was independently associated with the primary outcome (hazard ratio: 2.23; 95% confidence interval: 1.17 to 4.24, P = 0.014). Additionally, the nested regression model showed that VMT scoring provided an incremental prognostic value over clinically relevant variables (age, sex, the plasma brain natriuretic peptide level, atrial fibrillation) and LV diastolic function grading (chi-square 10.8 vs 16.3, P = 0.035).

Conclusions

In patients with HFpEF, the VMT score was independently and incrementally associated with adverse clinical outcomes. Moreover, it discriminated worse clinical outcome even in HFpEF patients with atrial fibrillation. Abstract Figure. VMT scoring  Abstract Figure. Kaplan-Meier analysis

Recent progress and future prospects of hyperon nucleon scattering experiment

A new hyperon-proton scattering experiment, dubbed J-PARC E40, was performed to measure differential cross sections of the Σ+p, Σ−p elastic scatterings and the Σ−p → Λn scattering by identifying a lot of Σ particles in the momentum ranging from 0.4 to 0.8 GeV/c produced by the π±p → K+Σ± reactions. We successfully measured the differential cross sections of these three channels with a drastically improved accuracy with a fine angular step. These new data will become important experimental constraints to improve the theories of the two-body baryon-baryon interactions. Following this success, we proposed a new experiment to measure the differential cross sections and spin observables by using a highly polarized Λ beam for providing quantitative information on the ΛN interaction. The results of three Σp channels and future prospects of the Λp scattering experiment are described.

Model-Induced Generalization Error Bound for Information-Theoretic Representation Learning in Source-Data-Free Unsupervised Domain Adaptation

Many unsupervised domain adaptation (UDA) methods have been developed and have achieved promising results in various pattern recognition tasks. However, most existing methods assume that raw source data are available in the target domain when transferring knowledge from the source to the target domain. Due to the emerging regulations on data privacy, the availability of source data cannot be guaranteed when applying UDA methods in a new domain. The lack of source data makes UDA more challenging, and most existing methods are no longer applicable. To handle this issue, this paper analyzes the cross-domain representations in source-data-free unsupervised domain adaptation (SF-UDA). A new theorem is derived to bound the target-domain prediction error using the trained source model instead of the source data. On the basis of the proposed theorem, information bottleneck theory is introduced to minimize the generalization upper bound of the target-domain prediction error, thereby achieving domain adaptation. The minimization is implemented in a variational inference framework using a newly developed latent alignment variational autoencoder (LA-VAE). The experimental results show good performance of the proposed method in several cross-dataset classification tasks without using source data. Ablation studies and feature visualization also validate the effectiveness of our method in SF-UDA.

Endoscope-assisted greater neurovascular palatal bundle release in cleft palatoplasty

Performing surgery in the oral cavity is difficult because of the limited view of the surgical field. Intraoral surgery for infantile oral disorders, such as cleft palate, is even more challenging. Endoscopy provides a minimally invasive approach and clear surgical view in surgeries with a constrained field of view. To date, very few reports have described endoscope-assisted palate surgery for children with cleft palate. At the authors' institution, endoscopes have been used in primary palatoplasty using the double-opposing Z-plasty technique. A novel endoscope-assisted procedure is described herein, in which a dissection around the greater palatine neurovascular bundle is used to obtain tension-free closure of the palatal cleft. With this technique, it was possible to minimize the application of additional von Langenbeck-type relaxation incisions, which were previously introduced in most of our cases; the relaxation incision was successfully circumvented in 42.3% of cases. This led to lesser surgical interference, which possibly resulted in favourable palatal development. It was also found that the endoscopic procedure did not increase the operation time or blood loss when compared to those patients who underwent the non-endoscopic procedure. It is concluded that endoscopic guidance is quite useful in primary palatoplasty procedures with a constricted surgical view.

Not Only Hypoxia- but Radiation-Induced Epithelial-Mesenchymal Transition Is Modulated by Hypoxia-Inducible Factor 1 in A549 Lung Cancer Cells

Hypoxia leads to post-treatment metastasis and recurrences of cancer via the epithelial-mesenchymal transition (EMT). Radiotherapy itself may also contribute to the acquisition of EMT phenotypes. Despite extensive studies on the EMT driven by either hypoxia or radiation stimuli, the molecular mechanisms characterizing these EMT events remain unclear. Thus, we aimed to evaluate the differences in the molecular pathways between hypoxia-induced EMT (Hypo-EMT) and radiation-induced EMT (R-EMT). Further, we investigated the therapeutic effects of HIF-1α inhibitor (LW6) on Hypo-EMT and R-EMT cells. A549 cells, lung adenocarcinoma cell line, acquired enhanced wound-healing activity under both hypoxia and irradiation. Localization of E-cadherin was altered from the cell membrane to the cytoplasm in both hypoxia and irradiated conditions. Of note, the expression levels of vimentin, one of the major EMT markers, was enhanced in irradiated cells, while it decreased under hypoxia condition. Importantly, LW6 significantly blocked EMT-related malignant phenotypes in both Hypo-EMT cells and R-EMT cells with concomitant re-location of E-cadherin onto the cell membrane. Moreover, LW6 deflected stress responsive signalling, JNK, activated sustainably under hypoxic condition, and the blockage of JNK impaired EMT phenotypes. Together, this work demonstrated the molecular events underlying Hypo-EMT and R-EMT, and highlighted HIF-1α as a therapeutic target not only in Hypo- EMT, but also in R-EMT.

Semi-Supervised Learning in Medical Images Through Graph-Embedded Random Forest

One major challenge in medical imaging analysis is the lack of label and annotation which usually requires medical knowledge and training. This issue is particularly serious in the brain image analysis such as the analysis of retinal vasculature, which directly reflects the vascular condition of Central Nervous System (CNS). In this paper, we present a novel semi-supervised learning algorithm to boost the performance of random forest under limited labeled data by exploiting the local structure of unlabeled data. We identify the key bottleneck of random forest to be the information gain calculation and replace it with a graph-embedded entropy which is more reliable for insufficient labeled data scenario. By properly modifying the training process of standard random forest, our algorithm significantly improves the performance while preserving the virtue of random forest such as low computational burden and robustness over over-fitting. Our method has shown a superior performance on both medical imaging analysis and machine learning benchmarks.

Effect of PM2.5 for hospitalization of cardiovascular diseases and medical expenses by age group: a nationwide study from the Japanese Registry Of All cardiac and vascular Diseases (JROAD)

Introduction

Although particulate matter with an aerodynamic diameter less than or equal to 2.5 μm (PM2.5) exposure influences cardiovascular events, there is limited knowledge how PM2.5 exposure is associated with cardiovascular hospitalization in Japan.

Purpose

We tested the hypothesis that PM2.5 exposure was related to the number of cardiovascular hospitalizations, hospitalization period, and medical expenses.

Methods

We included 835,405 subjects who were admitted to acute care hospitals in Japan. We classified PM2.5 exposure concentration into quintile groups (from low to high in the order of “PM-1” to “PM-5” group). Multilevel mixed-effects Poisson and linear regression analysis were used to estimate the association of PM2.5 exposure concentration with the cardiovascular hospitalization events, duration and medical expenses. We also analyzed those factors classified by age.

Results

PM-2, 3, 4, 5 group were positively related to the number of cardiovascular hospitalization events compared with PM-1 group (Table 1). PM-3, 4, 5 group were positively associated with a 0.018, 0.029, and 0.029 (all p&lt;0.001) of the difference of log cardiovascular hospitalization period compared with PM-1 group (p for trend &lt;0.001). These groups were also positively associated with a 0.016, 0.023, and 0.021 (all p&lt;0.001) of the difference of log medical expenses compared with PM-1 group (p for trend &lt;0.001). Analyzed by age group, hospitalization duration was longer (p&lt;0.05) and medical expenses was lower (p&lt;0.05) significantly in super-aging group (over 90 years old) than the group under 64 years old (Figure 1).

Conclusions

PM2.5 exposure concentration has harmful effect on not only cardiovascular events but cardiovascular hospitalization period and medical expenses due to big-data in Japan. Medical costs were kept low in comparison with their hospitalization period in subjects over 90 years old.

Figure 1

Funding Acknowledgement

Type of funding source: None

Prognostic relevance of exercise pulmonary hypertension for new-onset atrial fibrillation in primary mitral regurgitation

Background/Introduction

New-onset of atrial fibrillation (AF) portends poor prognosis in patients with primary mitral regurgitation (MR). However less is known about prognostic indicator for new-onset AF.

Purpose

The purpose of this study was to identify the prognostic relevance of exercise pulmonary hypertension for the new-onset AF in patients with primary MR.

Methods

Total of 114 consecutive patients with primary MR who underwent symptom-limited exercise echocardiography using supine-cycle ergometer were followed for new-onset AF over mean follow-up time of 3.6±2.6 years. Those with prevalent AF and pulmonary hypertension (estimated systolic pulmonary artery pressure ≥50mmHg) prior to exercise echocardiography were excluded from our analysis. We defined exercise-induced pulmonary hypertension (ExPHT) as those with peak estimated systolic pulmonary artery pressure (SPAP) ≥60mmHg or delta SPAP defined as differences between rest and peak SPAP ≥20mmHg.

Results

The mean age was 61±15 years old and 70 (61%) were male. Of those, 8 (7.8%) had mild MR, 32 (31.1%) had moderate MR, and 63 (61.2%) had severe MR. 60 (52.2%) patients had ExPHT. A total of 27 cases of new-onset AF were found during follow-up where the ExPHT group had higher prevalence of new-onset AF than the non-ExPHT group (35% vs. 11%, p=0.002). Those with ExPHT had significantly stronger association with shorter event-free survival time of new-onset AF (log-rank p&lt;0.001, Figure). When adjusted for age, sex, body surface area, the American Society of Echocardiography MR grade, left atrial dimension, peak systolic blood pressure and heart rate, the multivariable Cox regression analysis showed that those with ExPHT had a hazard risk of 3.1 ([95% CI 1.1–9.1], p=0.039) for new-onset of AF.

Conclusions

Exercise-induced pulmonary hypertension predicted incident of new-onset AF in those with primary MR. Exercise echocardiography is expected to play an important role in decision making with regards to the optimal timing for surgical intervention in primary MR.

Figure 1

Funding Acknowledgement

Type of funding source: None

Automatic detection of different types of small-bowel lesions on capsule endoscopy images using a newly developed deep convolutional neural network

BACKGROUND : Previous computer-aided detection systems for diagnosing lesions in images from wireless capsule endoscopy (WCE) have been limited to a single type of small-bowel lesion. We developed a new artificial intelligence (AI) system able to diagnose multiple types of lesions, including erosions and ulcers, vascular lesions, and tumors. METHODS : We trained the deep neural network system RetinaNet on a data set of 167 patients, which consisted of images of 398 erosions and ulcers, 538 vascular lesions, 4590 tumors, and 34 437 normal tissues. We calculated the mean area under the receiver operating characteristic curve (AUC) for each lesion type using five-fold stratified cross-validation. RESULTS : The mean age of the patients was 63.6 years; 92 were men. The mean AUCs of the AI system were 0.996 (95 %CI 0.992 - 0.999) for erosions and ulcers, 0.950 (95 %CI 0.923 - 0.978) for vascular lesions, and 0.950 (95 %CI 0.913 - 0.988) for tumors. CONCLUSION : We developed and validated a new computer-aided diagnosis system for multiclass diagnosis of small-bowel lesions in WCE images.

Successful pregnancy and delivery achieved following intracytoplasmic sperm injection using teratozoospermic sperm exhibiting marked dysplasia of the fibrous sheath: a case report

A husband and his wife, both 34 years old, consulted our clinic because of primary infertility. Sperm analysis revealed that the sperm concentration, motility, and progressive motility were (42.8±22.8)×10⁶/mL, 23.3%±12.2%, and 12.9%±6.1%, respectively. Based on Krugar strict morphology criteria, 100% of the sperm were teratozoospermic, with 7.9% DNA fragmentation index. Observation of the sperm under a transmission electron microscope revealed that most parts of the fibrous sheath (FS) surrounding the tails of the sperm were missing from midway through the principal piece to the end piece, although the sperm’s heads, necks and midpieces were morphologically normal. To collect oocytes, the gonadotropin-releasing hormone antagonist protocol was carried out, and 7 oocytes were retrieved. Intracytoplasmic sperm injection (ICSI) was performed for all the teratozoospermic sperm. Of the 7 oocytes, 3 were fertilized, and one 8-cell embryo and 2 expanded blastocysts were vitrified. Although repeated transfers of expanded blastocysts resulted in no implantation, one 8-cell embryo transfer in a hormone replacement therapy cycle led to pregnancy. The pregnancy using an 8-cell vitrified embryo resulted in the delivery of a healthy female baby at 38 weeks of gestation. No congenital malformations were found until 28 days after birth. Our results demonstrated that healthy birth could be achieved following the transfer of an embryo derived from ICSI using teratozoospermic sperm exhibiting the dysplasia of the fibrous sheath (DFS). Furthermore, while the previous reports on DFS have not investigated male infertility, we evaluated sperms from various aspects such as Kruger sperm function test, chromatin dispersion test, electron microscopy findings, time-lapse images of the obtained embryos, and concluded that ICSI could be desirable as a treatment policy for DFS.

P304Prognostic significance of transvalvular flow rate during exercise in asymptomatic patients with aortic stenosis

Background

The optimal management of asymptomatic aortic stenosis (AS) remains controversial. The aim of this study is to investigate the prognostic value of exercise stress echocardiography (ESE) in asymptomatic patients with AS.

Purpose

To investigate the additive value of ESE in asymptomatic patients with AS.

Methods and results

We retrospectively enrolled 109 consecutive patients (mean age 73±13 years, 55 men) with AS (aortic valve area ≤1.5cm2) who underwent ESE. Of these, 10 patients referred for aortic valve replacement without symptoms were excluded; finally, 99 conservatively managed patients were enrolled. During the mean follow-up period of 14±11 months, 23 patients (23%) suffered from AS related events. Although no differences in mean pressure gradient at rest and during exercise were found between the patients with and without adverse events, transvalvular flow rate during exercise (Ex-FR) was lower in patients with adverse events than those without adverse events (236±55 ml/sec vs 274±64 ml/sec, P=0.01). Using multivariate Cox regression analysis, low Ex-FR (<270 ml/sec) was an independent predictor for adverse events in patients with asymptomatic AS (hazard ratio: 3.53, P<0.01). The Kaplan–Meier analysis showed that Ex-FR was clearly stratified the event-free survival (Figure, log-rank P<0.01).

K-M curve according to Ex-FR

Conclusion

These results suggested that Ex-FR measured by ESE should play crucial roles in risk stratification in asymptomatic patients with AS.

P1493Prognostic significance of right ventricular function during exercise in patients with non-obstructive hypertrophic cardiomyopathy

Background

Hypertrophic cardiomyopathy (HCM) is a heterogeneous condition that may present crucial complication including life-threatening arrhythmia and sudden cardiac death. However, the risk stratification of HCM without left ventricular outflow tract (LVOT) obstruction had not been fully elucidated. Moreover, although recent studies have revealed the right ventricle (RV) involvement of HCM, the prognostic importance of RV function during exercise is unclear.

Purpose

To investigate the prognostic significance of RV function in patients with non-obstructive HCM using exercise stress echocardiography (ESE).

Methods and results

This study conducted on 100 HCM patients (age 62.9±13.6 years, 63% men) with preserved left ventricular ejection fraction who underwent ESE using semi-supine bicycle ergometer. Ten patients with significant LVOT obstruction (≥30mmHg) were excluded and 9 were also excluded because of the inadequate imaging quality or insufficiency of data. Among remaining 81 non-obstructive HCM patients, 9 patients suffered from HCM related cardiac events including cardiac death, unexpected hospitalization, life-threatening arrhythmias, and new-onset of syncope during the mean follow up period of 2.6±1.6 years. A multivariate Cox Hazard analysis revealed that low tricuspid annular plane systolic excursion during exercise (Ex-TAPSE, cut-off: 24mm) was an independent predictor of cardiac events. (hazard ratio: 18.66, 95% confidence interval: 3.66–338.46, P<0.001) The estimated cumulative cardiac event free survival using the Kaplan-Meier method was significantly lower in patients with reduced Ex-TAPSE (<24mm) than those with preserved Ex-TAPSE (Log-rank, P<0.01).

K-M curve according to Ex-TAPSE

Conclusion

Ex-TAPSE had a strong predictive value of clinical outcomes in non-obstructive HCM patients. Right ventricular function during exercise may have crucial role in the risk stratification of non-obstructive HCM.

Electric dipole effect in PdCoO2/β-Ga2O3 Schottky diodes for high-temperature operation

High-temperature operation of semiconductor devices is widely demanded for switching/sensing purposes in automobiles, plants, and aerospace applications. As alternatives to conventional Si-based Schottky diodes usable only at 200°C or less, Schottky interfaces based on wide-bandgap semiconductors have been extensively studied to realize a large Schottky barrier height that makes high-temperature operation possible. Here, we report a unique crystalline Schottky interface composed of a wide-gap semiconductor β-Ga₂O₃ and a layered metal PdCoO₂. At the thermally stable all-oxide interface, the polar layered structure of PdCoO₂ generates electric dipoles, realizing a large Schottky barrier height of ~1.8 eV, well beyond the 0.7 eV expected from the basal Schottky-Mott relation. Because of the naturally formed homogeneous electric dipoles, this junction achieved current rectification with a large on/off ratio approaching 10⁸ even at a high temperature of 350°C. The exceptional performance of the PdCoO₂/β-Ga₂O₃ Schottky diodes makes power/sensing devices possible for extreme environments.

Random-start ovarian stimulation with aromatase inhibitor for fertility preservation in women with Japanese breast cancer

PURPOSE

Fertility preservation is an important issue for young cancer patients. Random-start controlled ovarian stimulation and double ovarian stimulation have been proposed for efficient oocyte retrieval within the limited time before cancer therapy. We aimed to clarify the efficacy of these new protocols within the Japanese population.

METHODS

We performed a retrospective observational study at a multicenter from February 2012 to August 2017. The study entailed 50 cycles with 34 patients who underwent fertility preservation due to breast cancer. Follicular phase or luteal phase ovarian stimulation with aromatase inhibitor was performed. A second ovarian stimulation was started with or without waiting until the next menstruation. We measured the number of retrieved oocytes and cryopreserved oocytes/embryos, the ratio of mature oocytes, and the fertilization rate.

RESULTS

The numbers of retrieved oocytes and frozen oocytes/embryos were not significantly different between follicular phase and luteal phase ovarian stimulation. The number of retrieved oocytes was not reduced at the second ovum pick up compared to the first ovum pick up in the double ovarian stimulation.

CONCLUSIONS

Random-start controlled ovarian stimulation and double ovarian stimulation with aromatase inhibitor for breast cancer patients were effective protocols for retrieving a greater number of oocytes within the limited time.

Automatic diagnosis of neurological diseases using MEG signals with a deep neural network

The application of deep learning to neuroimaging big data will help develop computer-aided diagnosis of neurological diseases. Pattern recognition using deep learning can extract features of neuroimaging signals unique to various neurological diseases, leading to better diagnoses. In this study, we developed MNet, a novel deep neural network to classify multiple neurological diseases using resting-state magnetoencephalography (MEG) signals. We used the MEG signals of 67 healthy subjects, 26 patients with spinal cord injury, and 140 patients with epilepsy to train and test the network using 10-fold cross-validation. The trained MNet succeeded in classifying the healthy subjects and those with the two neurological diseases with an accuracy of 70.7 ± 10.6%, which significantly exceeded the accuracy of 63.4 ± 12.7% calculated from relative powers of six frequency bands (δ: 1-4 Hz; θ: 4-8 Hz; low-α: 8-10 Hz; high-α: 10-13 Hz; β: 13-30 Hz; low-γ: 30-50 Hz) for each channel using a support vector machine as a classifier (p = 4.2 × 10-2). The specificity of classification for each disease ranged from 86-94%. Our results suggest that this technique would be useful for developing a classifier that will improve neurological diagnoses and allow high specificity in identifying diseases.

Characterizing Adversity of Lysosomal Accumulation in Nonclinical Toxicity Studies: Results from the 5th ESTP International Expert Workshop

Lysosomes have a central role in cellular catabolism, trafficking, and processing of foreign particles. Accumulation of endogenous and exogenous materials in lysosomes represents a common finding in nonclinical toxicity studies. Histologically, these accumulations often lack distinctive features indicative of lysosomal or cellular dysfunction, making it difficult to consistently interpret and assign adverse dose levels. To help address this issue, the European Society of Toxicologic Pathology organized a workshop where representative types of lysosomal accumulation induced by pharmaceuticals and environmental chemicals were presented and discussed. The expert working group agreed that the diversity of lysosomal accumulations requires a case-by-case weight-of-evidence approach and outlined several factors to consider in the adversity assessment, including location and type of cell affected, lysosomal contents, severity of the accumulation, and related pathological effects as evidence of cellular or organ dysfunction. Lysosomal accumulations associated with cytotoxicity, inflammation, or fibrosis were generally considered to be adverse, while those found in isolation (without morphologic or functional consequences) were not. Workshop examples highlighted the importance of thoroughly characterizing the biological context of lysosomal effects, including mechanistic data and functional in vitro readouts if available. The information provided here should facilitate greater consistency and transparency in the interpretation of lysosomal effects.