Japan Journal of Medical Informatics
Online ISSN : 2188-8469
Print ISSN : 0289-8055
ISSN-L : 0289-8055
Current issue
Showing 1-3 articles out of 3 articles from the selected issue
Original Article-Notes
  • Kiyotaka Fujii, Yuko Ohno, Michiko Kido, Kai Ishida, Hieyong Jeong
    2018 Volume 38 Issue 6 Pages 321-336
    Published: 2018
    Released: February 28, 2020
    JOURNALS FREE ACCESS

     In medical institutions, wandering elderly patients have become a problem, and the introduction of wandering sensing systems is underway. However, wandering sensing systems are known to affect wireless medical telemetry systems. In this research, we aimed to clarify the following two points. The first point is to visualize the signal of the wandering sensing equipment and the medical telemeter which influence by using the real time USB spectrum analyzer and to clarify the cause of the electromagnetic interference. The second point is to focus on the medical telemeter channel and to investigate how the wandering sensing system affects to which channel of the medical telemeter. With DPX real time display, we could observe three signals of wandering sensing system, main signal, out of band emission, instantaneous signal. Furthermore, the center frequency in the main signal of the wandering sensing system corresponds to 3000 series of wireless medical telemetry system. Those experiments suggested the possibility of electromagnetic interference when using the wandering sensing system and the wireless medical telemetry system (the channel numbers of 3000 series) simultaneously at relatively short distances (10 cm, 1 m, and 3 m). In addition to being the frequency band allocated for wireless medical telemetry systems, these channels are used in the wandering sensing system. Hence, it is necessary to consider possible electromagnetic interference problems.

    Download PDF (3338K)
Original Article-Short Notes
  • E Aramaki, S Wakamiya, T Iwao, C Kawakami, M Nakae, T Matsumoto, K Tom ...
    2018 Volume 38 Issue 6 Pages 337-348
    Published: 2018
    Released: February 28, 2020
    JOURNALS FREE ACCESS

     Due to the rapid development of IT in medical field, various kinds of medical and drug information have been digitized. However, with regard to the quality of the drug package insert indispensable for proper use of drugs, systematic surveys on whether the notation is unified have not been conducted so far. This time, we paid attention to pediatric areas and investigated how the information on pediatrics is described in the package insert. As a result, it was found that the proportion of information on dosage regimen for children was only 13.5% in the package insert and 49.2% in frequent medicine for children. In particular, it was found that there were few descriptions for low birth weight infants and newborn babies, which are more detailed childhood classifications. Furthermore, with regard to the method of describing the package insert, it is highly likely that there is ambiguity in the age classification, safety, and the place of entry, which is a barrier to proper use of medicines in pediatric medical field. It is found that it is also necessary to improve the quality of the package insert itself, such as eliminating the fluctuation and ambiguity in the expression of the package insert.

    Download PDF (890K)
Proceeding of the Spring Meeting on Medical Informatics
  • M Suzuki, T Shibahara, Y Muragaki
    2018 Volume 38 Issue 6 Pages 351-357
    Published: 2018
    Released: February 28, 2020
    JOURNALS FREE ACCESS

     Over the last half century, a number of new learning methods have been developed, including SVMs and deep neural networks. These are very accurate, but unfortunately they also lack explainability. In particular, deep neural networks provide no information about the importance of feature variables. High explainability is expected to guarantee the reliability of prediction models made by learning methods other than the evaluation of prediction accuracy. To address this problem, we have developed a factor analysis technique for nonlinear machine learning methods. The technique has two statistical steps as follows. The first step, called backward analysis, generates probability distributions of the positive and negative classes estimated by the prediction model. The second step uses backward elimination based on Hilbert-Schmidt independence criteria to extract feature variables for which there is a nonlinear correlation between the feature variables and outcome. This factor analysis technique was verified by simulation. In the experiment, we extracted new factors that are relevant to prostate cancer from the feature variables of gene expression data. Experimental results show that this technique has the potential to play a vital role in clinical research.

    Download PDF (839K)
feedback
Top