Performance evaluation of the fully automated urine analyzer US-3500 with improved algorithm for determining urine color tone

DOI Web Site
  • OKA Aki
    Division of Clinical Laboratory, Gifu University Hospital
  • MATSUNO Hiroko
    Division of Clinical Laboratory, Gifu University Hospital
  • KATO Yohei
    Division of Clinical Laboratory, Gifu University Hospital
  • ISHIDA Mariko
    Division of Clinical Laboratory, Gifu University Hospital
  • SHIRAKAMI Yohei
    Division of Clinical Laboratory, Gifu University Hospital Department of Gastroenterology, Gifu University Graduate School of Medicine
  • WATANABE Takatomo
    Division of Clinical Laboratory, Gifu University Hospital Department of Cardiology, Gifu University Graduate School of Medicine
  • KIKUCHI Ryosuke
    Division of Clinical Laboratory, Gifu University Hospital
  • OHKURA Hiroyuki
    Division of Clinical Laboratory, Gifu University Hospital Department of Cardiology, Gifu University Graduate School of Medicine

Bibliographic Information

Other Title
  • 全自動尿分析装置US-3500における尿色調判定アルゴリズム改良後の性能評価
  • ゼン ジドウ ニョウ ブンセキ ソウチ US-3500 ニ オケル ニョウ シキチョウ ハンテイ アルゴリズム カイリョウ ゴ ノ セイノウ ヒョウカ

Search this article

Abstract

<p>There are cases in which the actual color tone differs from the device’s judgment of the urine color tone in the urine qualitative test. In particular, the cases in which “RED” and “OTHERS” are determined to be different from the actual color tone have been a problem. Here, we evaluated the performance of the color tone judgment function of the fully automatic urine analyzer US-3500, which uses a color tone judgment algorithm improved by Eiken Chemical Co. We collected color tone results from 1,190 specimens submitted to our laboratory and photographed the color tone of the specimens after measurement using a digital camera, and we then compared the algorithm judgments before and after the improvement. Of the 1,190 specimens, seven specimens were judged “RED” by the conventional algorithm (hereinafter referred to as “conventional”), three specimens by the improved algorithm (hereinafter referred to as “improved”), and 12 specimens by the conventional algorithm and three specimens by the improved algorithm were judged “OTHERS”. Next, we checked the images of a total of 19 specimens that were judged “RED” or “OTHERS” by the conventional method. Of the seven specimens that were judged “RED” by the conventional method, three specimens that were judged “RED” by the improved method also showed a red color tone. The remaining four specimens were determined to be “DARK BROWN” and “AMBER,” which were consistent with the images. Nine of the 12 specimens judged as “OTHERS” by the conventional algorithm were judged as “AMBER,” “YELLOW,” “STRAW,” or “LIGHT YELLOW” by the improved algorithm. Overall, the improved algorithm improved the accuracy of “RED” and “OTHERS” judgments in particular.</p>

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top