Akira Koseki

contact information

Senior Research Scientist
Tokyo Research Laboratory, Yamato, Japan

links

Professional Associations

Professional Associations: IPSJ

2020

Increasing tendency of urine protein is a risk factor for rapid eGFR decline in patients with CKD: A machine learning-based prediction model by using a big database
Daijo Inaguma, Akimitsu Kitagawa, Ryosuke Yanagiya, Akira Koseki, Toshiya Iwamori, Michiharu Kudo, Yukio Yuzawa
PLoS One, Public Library of Science, 2020
Abstract

Artificial intelligence is increasingly being adopted in medical fields to predict various outcomes. In particular, chronic kidney disease (CKD) is problematic because it often progresses to end- stage kidney disease. However, the trajectories of kidney function depend on individual patients. In this study, we propose a machine learning-based model to predict the rapid decline in kidney function among CKD patients by using a big hospital database constructed from the information of 118,584 patients derived from the electronic medical records system. The database included the estimated glomerular filtration rate (eGFR) of each patient, recorded at least twice over a period of 90 days. The data of 19,894 patients (16.8%) were observed to satisfy the CKD criteria. We characterized the rapid decline of kidney function by a decline of 30% or more in the eGFR within a period of two years and classified the available patients into two groups those exhibiting rapid eGFR decline and those exhibiting non-rapid eGFR decline. Following this, we constructed predictive models based on two machine learning algorithms. Longitudinal labora- tory data including urine protein, blood pressure, and hemoglobin were used as covariates. We used longitudinal statistics with a baseline corresponding to 90-, 180-, and 360-day windows prior to the baseline point. The longitudinal statistics included the exponentially smoothed aver- age (ESA), where the weight was defined to be 0.9*(t/b), where t denotes the number of days prior to the baseline point and b denotes the decay parameter. In this study, b was taken to be 7 (7-day ESA). We used logistic regression (LR) and random forest (RF) algorithms based on Python code with scikit-learn library (https://scikit-learn.org/) for model creation. The areas under the curve for LR and RF were 0.71 and 0.73, respectively. The 7-day ESA of urine protein ranked within the first two places in terms of importance according to both models. Further, other features related to urine protein were likely to rank higher than the rest. The LR and RF models revealed that the degree of urine protein, especially if it exhibited an increasing ten- dency, served as a prominent risk factor associated with rapid eGFR decline.

2019

Artificial intelligence predicts the progression of diabetic kidney disease using big data machine learning
Masaki Makino, Ryo Yoshimoto, Masaki Ono, Toshinari Itoko, Takayuki Katsuki, Akira Koseki, Michiharu Kudo, Kyoichi Haida, Jun Kuroda, Ryosuke Yanagiya, Eiichi Saitoh, Kiyotaka Hoshinaga, Yukio Yuzawa, Atsushi Suzuki
Scientific Reports 9(1), 11862, 2019
Abstract

2018

Risk Prediction of Diabetic Nephropathy via Interpretable Feature Extraction from EHR Using Convolutional Autoencoder
Takayuki Katsuki, Masaki Ono, Akira Koseki, Michiharu Kudo, Kyoichi Haida, Jun Kuroda, Masaki Makino, Ryosuke Yanagiya, Atsushi Suzuki
medical informatics europe, pp. 106-110, 2018
Abstract temporal information, medicine, machine learning, feature extraction, event sequence, electronic health record, diabetic nephropathy, data mining, autoencoder, artificial intelligence

Feature Extraction from Electronic Health Records of Diabetic Nephropathy Patients with Convolutioinal Autoencoder
Takayuki Katsuki, Masaki Ono, Akira Koseki, Michiharu Kudo, Kyoichi Haida, Jun Kuroda, Masaki Makino, Ryosuke Yanagiya and Atsushi Suzuki
AAAI Workshop on Health Intelligence, AAAI, 2018

Time-Discounting Convolution for Event Sequences with Ambiguous Timestamps [pdf]
Takayuki Katsuki, Takayuki Osogami, Akira Koseki, Masaki Ono, Michiharu Kudo, Masaki Makino, and Atsushi Suzuki
Proceedings of the 18th IEEE International Conference on Data Mining (ICDM 2018), pp. 1085-1090

Risk Prediction of Diabetic Nephropathy via Interpretable Feature Extraction from EHR Using Convolutional Autoencoder [pdf]
Katsuki, T and Ono, M and Koseki, A and Kudo, M and Haida, K and Kuroda, J and Makino, M and Yanagiya, R and Suzuki, A
Studies in health technology and informatics247, 106-110, 2018
Abstract

2014

A financial risk evaluation service for integrating private portfolios securely
Yuji Watanabe, Akira Koseki
Service Operations and Logistics, and Informatics (SOLI), 2014 IEEE International Conference on, pp. 59-64
risk management, risk analysis, marketing, it risk management, financial risk management, financial risk, engineering, actuarial science

2013

An event-processing system alerting analytically to networked vehicles
Imai, Haruki and Maeda, Kumiko and Chiba, Tatsuhiro and Negishi, Yasushi and Koseki, Akira and Aihara, Toru and Komatsu, Hideaki
Intelligent Transportation Systems-(ITSC), 2013 16th International IEEE Conference on, pp. 485--492
Abstract

2011

Distributed and fault-tolerant execution framework for transaction processing
T. Suganuma, A. Koseki, K. Ishizaki, Y. Ueda, K. Mizuno, D. Silva, H. Komatsu, T. Nakatani
Proceedings of the 4th Annual International Conference on Systems and Storage, pp. 2, 2011

2010

Parallel Programming Framework for Large Batch Transaction Processing on Scale-out Systems
Kazuaki Ishizaki, Ken Mizuno, Toshio Suganuma, Daniel Silva, Akira Koseki, Hideaki Komatsu, Yohei Ueda, Toshio Nakatani
Proceedings of the 3rd Annual Haifa Experimental Systems Conference, pp. 15:1--15:14, ACM, 2010

2004

Lock reservation for Java reconsidered
Tamiya Onodera, Kikyokuni Kawachiya, Akira Koseki
ECOOP 2004--Object-Oriented Programming, pp. 559--583, Springer

2003

Spill code minimization by spill code motion
Akira Koseki, Hideaki Komatsu, Toshio Nakatani
Proceedings of the 12th International Conference on Parallel Architectures and Compilation Techniques, pp. 125-134, 2003
Abstract theoretical computer science, small number, register allocation, redundancy, real time computing, parallel computing, minification, instructions per cycle, instruction set, heuristics, graph coloring, computer science

Effectiveness of cross-platform optimizations for a Java just-in-time compiler

SIGPLAN Not. 38(11), ACM, 2003
Abstract

2002

Preference-directed graph coloring
Akira Koseki, Hideaki Komatsu, Toshio Nakatani
Proceedings of the ACM SIGPLAN 2002 conference on Programming language design and implementation, pp. 33-44
Abstract theoretical computer science, strength of a graph, list coloring, greedy coloring, graph coloring, fractional coloring, distance hereditary graph, discrete mathematics, computer science, complete coloring, comparability graph, algorithm

Lock reservation: Java locks can mostly do without atomic operations
Kiyokuni Kawachiya, Akira Koseki, Tamiya Onodera
OOPSLA '02 Proceedings of the 17th ACM SIGPLAN conference on Object-oriented programming, systems, languages, and applications , pp. 130--141, 2002

Year Unknown

Evolution of a Java just-in-time compiler for IA-32 platforms

IBM Journal of ..., 2010 - ieeexplore.ieee.org

Akira Koseki

contact information

links

Professional Associations

2020

2019

2018

2014

2013

2011

2010

2004

2003

2002

Year Unknown

Projects and Groups

Technical Areas

Akira Koseki gen_feedback_link(leftr, rightr);

contact information

links

Professional Associations

2020

2019

2018

2014

2013

2011

2010

2004

2003

2002

Year Unknown

Projects and Groups

Technical Areas

Akira Koseki