A5009899081- Artificial Intelligence in Healthcare
- Computational Drug Discovery Methods
- Pressure Ulcer Prevention and Management
- Imbalanced Data Classification Techniques
- Diabetic Foot Ulcer Assessment and Management
- Face and Expression Recognition
- Liver Disease Diagnosis and Treatment
- Pharmacogenetics and Drug Metabolism
- Intracerebral and Subarachnoid Hemorrhage Research
- Adipokines, Inflammation, and Metabolic Diseases
- Diagnosis and Treatment of Venous Diseases
- Fault Detection and Control Systems
- Blind Source Separation Techniques
- Vascular Malformations Diagnosis and Treatment
- Digital Imaging for Blood Diseases
- Neural Networks and Applications
- Biomedical Text Mining and Ontologies
- Cancer, Lipids, and Metabolism
- Intracranial Aneurysms: Treatment and Complications
- Machine Learning and Data Classification
- Wound Healing and Treatments
- Spectroscopy and Chemometric Analyses
Tokushima University
2015-2020
Institute of Biomedical Science
2016
Nagoya Institute of Technology
2004
To develop a prediction model for pressure ulcer cases that continue to occur at an acute care hospital with low occurrence rate of ulcers.Analyzing data were collected from patients hospitalized Tokushima University Hospital during 2012 using alternating decision tree (ADT) mining method.The ADT-based analysis revealed transfer activity, operation time, and body mass index (BMI) as important factors predicting development.Among the identified, only "transfer activity" can be modified by...
The severity of clinical signs and symptoms cranial dural arteriovenous fistulas (DAVFs) are well correlated with their pattern venous drainage. Although the presence cortical drainage can be considered a potential predictor aggressive DAVF behaviors, such as intracranial hemorrhage or progressive neurological deficits due to congestion, accurate statistical analyses currently not available. Using decision tree data mining method, authors aimed at clarifying predictability future development...
Background: Pressure ulcers (PUs) are considered a serious problem in nursing care and require preventive measures. Many risk assessment methods currently being used, but most the collection of data not available on admission. Although nurses assess Nursing Needs Score (NNS) daily basis Japanese acute hospitals, these primarily used to standardize cost public insurance system for appropriate nurse staffing, have never been PU assessment. Objective: The objective this study was predict...
Abstract To investigate unknown patterns associated with type 2 diabetes in the Japanese population, we first used an alternating decision tree (ADTree) algorithm, a powerful classification algorithm from data mining, for 1,102 subjects aged 35–69 years. On basis of investigated patterns, then evaluated associations serum high-sensitivity C-reactive protein (hs-CRP) as biomarker systemic inflammation and family history (negative, positive or unknown) prevalence because their detailed have...
Hepatocellular carcinoma (HCC) is a highly lethal tumor and the majority of postoperative patients experience recurrence. In present study, we focus on predictability recurrence HCC through data mining method. total, 323 patients with who underwent hepatic resection were included in 156 whom suffered from cancer Clinicopathological including prognosis analyzed using method for HCC. The resulting alternating decision tree (ADT) was described This tree validated 10‑fold cross...
The analysis of Electronic Health Records (EHRs) is attracting a lot research attention in the medical informatics domain. Hospitals and institutes started to use data mining techniques gain new insights from massive amounts that can be made available through EHRs. Researchers field have often used descriptive statistics classical statistical methods prove assumed hypotheses. However, discovering large solely based on experts' observations difficult. Using visualizations, practitioners find...
In the field of artificial neural networks, large-scale classification problems are still challenging due to many obstacles such as local minima state, long time computation, and requirement large amount memory. The network CombNET-II overcomes state proves give good recognition rate in applications. However requires a memory used for training database feature space. We propose revised version with considerably lower requirement, which makes problem more tractable. reduction is achieved by...
In this paper we propose a new method for generating an informative QSAR model (called VSVR-QSAR) using Voronoi grid and support vector machines regression. The procedure enables researchers to understand the physicochemical meaning of steric electrostatics measurements inclusion those as latent variables in generated model. proved be comparable or better than classical QSAR, well conventional 3D-QSAR procedures
The possibility of extracting useful medical information from data collected by nurses for management purposes is investigated. An alternating decision tree predicting pressure ulcer development generated nursing needs score (NNS) usually recorded in Japanese hospitals.