A5030896432- Radiomics and Machine Learning in Medical Imaging
- COVID-19 diagnosis using AI
- Medical Imaging Techniques and Applications
- Lung Cancer Diagnosis and Treatment
- Advanced X-ray and CT Imaging
- Medical Imaging and Analysis
- Natural Language Processing Techniques
- Neuropeptides and Animal Physiology
- Radiation Dose and Imaging
- Cardiac Imaging and Diagnostics
- Neuroendocrine Tumor Research Advances
- Artificial Intelligence in Healthcare and Education
- Advanced MRI Techniques and Applications
- Plant Molecular Biology Research
- Medical Image Segmentation Techniques
- Radiology practices and education
- Nitric Oxide and Endothelin Effects
- Growth Hormone and Insulin-like Growth Factors
- Topic Modeling
- Multimodal Machine Learning Applications
- Plant and animal studies
- Anatomy and Medical Technology
- Plant Parasitism and Resistance
- Pancreatic function and diabetes
- Brain Tumor Detection and Classification
Osaka University
2012-2025
Nara Institute of Science and Technology
2016-2021
Kobe University
2018
Kobe Pharmaceutical University
2018
Johns Hopkins University
2017
Johns Hopkins Medicine
2017
Dokkyo University
1988-2009
Institute of Brain and Blood Vessels
2009
University of Yamanashi
2006
Shimizu (Japan)
1997
A considerable quantity of endothelin-like immunoreactivity was demonstrated as present in the CSF patients with subarachnoid hemorrhage. The endothelin levels raised from 0.4 ± 0.2 (Mean SD) pmol/L at Day 0-1 to 2.2 0.6 6 and decreased gradually. result suggest that may contribute generation vasospasm often observed
In computer-aided diagnosis systems for lung cancer, segmentation of nodules is important analyzing image features on computed tomography (CT) images and distinguishing malignant from benign ones. However, it difficult to accurately robustly segment attached the chest wall or with ground-glass opacities using conventional processing methods. Therefore, this study aimed develop a method robust accurate three-dimensional (3D) nodule regions deep learning. study, nested 3D fully connected...
To assess whether transfer learning with a bidirectional encoder representations from transformers (BERT) model, pretrained on clinical corpus, can perform sentence-level anatomic classification of free-text radiology reports, even for classes few positive examples.This retrospective study included reports patients who underwent whole-body PET/CT imaging December 2005 to 2020. Each sentence in these (6272 sentences) was labeled by two annotators according body part ("brain," "head & neck,"...
Abstract Fluorine-18-fluorodeoxyglucose ( 18 F-FDG) positron emission tomography (PET)/computed (CT) is widely used for the detection, diagnosis, and clinical decision-making in oncological diseases. However, daily medical practice, it often difficult to make decisions because of physiological FDG uptake or cancers with poor uptake. False negative diagnoses malignant lesions are critical issues that require attention. In this study, Vision Transformer (ViT) was automatically classify F-FDG...
Parasitic plants infect other by forming haustoria, specialized multicellular organs consisting of several cell types, each which has unique morphological features and physiological roles associated with parasitism. Understanding the spatial organization types is, therefore, great importance in elucidating functions haustoria. Here, we report a three-dimensional (3-D) reconstruction haustoria from two Orobanchaceae species, obligate parasite Striga hermonthica infecting rice (Oryza sativa)...
Abstract Purpose To predict solid and micropapillary components in lung invasive adenocarcinoma using radiomic analyses based on high-spatial-resolution CT (HSR-CT). Materials methods For this retrospective study, 64 patients with were enrolled. All scanned by HSR-CT 1024 matrix. A pathologist evaluated subtypes (lepidic, acinar, solid, micropapillary, or others). Total 61 features the images calculated our modified texture analysis software, then filtered minimized least absolute shrinkage...
Automated segmentation of multiple organs in CT data the upper abdomen is addressed. In order to explicitly incorporate spatial interrelations among organs, we propose a method for finding and representing based on canonical correlation analysis. Furthermore, methods are developed constructing utilizing statistical atlas which inter-organ constraints incorporated improve accuracy multi-organ segmentation. The proposed were tested perform eight abdominal (liver, spleen, kidneys, pancreas,...
Endothelin-1 (ET-1) induces pulmonary vascular remodeling and hypertension secondary to fibrosis. Given that endothelial cells are the main source of ET-1 from other may encounter difficulty penetrating compartments, we hypothesize endothelial-derived promotes We used knock-out (VEETKO) Wild type mice for this research. They were given intratracheal bleomycin euthanized at day 28. quantified fibrosis, measured lung its receptors' expression, assessed by calculating medial wall index,...
We investigated the possible relation between neuropeptides and cerebral vasoconstriction in samples of ventricular or cisternal cerebrospinal fluid from 14 patients with subarachnoid hemorrhage. Neuropeptide Y, calcitonin gene-related peptide, atrial natriuretic pituitary polypeptide 7B2 were present these patients. Concentrations peptide not significantly different those control subjects, but that was lower. Although mean concentration neuropeptide Y higher than control, consecutive...
To propose a five-point scale for radiology report importance called Report Importance Category (RIC) and to compare the performance of natural language processing (NLP) algorithms in assessing RIC using head computed tomography (CT) reports written Japanese. 3728 Japanese CT performed at Osaka University Hospital 2020 were included. (category 0: no findings, category 1: minor 2: routine follow-up, 3: careful 4: examination or therapy) was established based not only on patient severity but...
We proposed a bimodal artificial intelligence that integrates patient information with images to diagnose spinal cord tumors. Our model combines TabNet, state-of-the-art deep learning for tabular data information, and convolutional neural network images. As training data, we collected 259 tumor patients (158 schwannoma 101 meningioma). compared the performance of image-only unimodal model, table-only using gradient-boosting decision tree, TabNet. TabNet performed best (area under...
In total hip arthroplasty, analysis of postoperative images is important to evaluate surgical outcome. Since CT most prevalent modality in orthopedic surgery, we aimed at the image. The challenge this work metal artifact caused by metallic implant, which reduces accuracy segmentation especially vicinity implant. Our goal was develop an automated method muscles images. paper, propose a that combines Normalized Metal Artifact Reduction (NMAR), one state-of-the-art reduction methods, and CNN-...
Artificial intelligence (AI) systems designed to detect abnormalities in abdominal computed tomography (CT) could reduce radiologists' workload and improve diagnostic processes. However, development of such models has been hampered by the shortage large expert-annotated datasets. Here, we used information from free-text radiology reports, rather than manual annotations, develop a deep-learning-based pipeline for comprehensive detection CT abnormalities.
PET/CT can scan low-dose computed tomography (LDCT) images with morphological information and PET functional information. Because the whole body is targeted for imaging, examinations are important in cancer diagnosis. However, several obtained by place a heavy burden on radiologists during Thus, development of computer-aided diagnosis (CAD) technologies assisting has been requested. because FDG accumulation differs each organ, recognizing organ regions essential developing lesion detection...
Abstract Cervical sagittal alignment is an essential parameter for the evaluation of spine disorders. Manual measurement time-consuming and burdensome to measurers. Artificial intelligence (AI) in form convolutional neural networks has begun be used measure x-rays. This study aimed develop AI automated lordosis on lateral cervical We included 4546 x-rays from 1674 patients. For all x-rays, caudal endplates C2 C7 were labeled based consensus among well-experienced surgeons, data which as...
Radiology report generation systems have the potential to reduce workload of radiologists by automatically describing findings in medical images.To broaden application system, system should generate reports that are not only factually accurate but also chronologically consistent, images presented time order, is, correct order.We employ a planning-based radiology generates overall structure as “plans’” prior generating and consistent order.Additionally, we propose novel reinforcement learning...