Haematology International Journal (HIJ)

ISSN: 2578-501X

Upcoming Article

A Comparison of Artificial Intelligence (AI) Based System, Automated Hematology Analyzer, & Manual Method in Platelet Estimation in Peripheral Blood Smear

Abstract

Introduction: AI based automated peripheral blood smear slide analyzers are tested for in vitro diagnosis that provides morphological analysis as well as quantification of haematology parameters.
Aims & objectives: The present study aims at determining the correlation between Sysmex XN 1000 & AI based system results of platelet count with the manual method and to assess accuracy, utility & reliability of AI platform in diagnostic hematology.
Material & method: The study conducted in the department of Pathology, at tertiary care hospital in western India, included 171 patients for platelet count estimation were included in the study after taking written consent from the participants.
Results: There was an acceptable correlation between platelet counts as estimated by AI based system with manual platelet counting with coefficient of correlation .7399. The correlation coefficient for Sysmex XN 1000 platelet count with manual platelet count was .8878.( P value- less than .001 , statistically significant) & between AI based system and Sysmex XN 1000 was .7206 . The platelet count measurement by Sysmex had higher precision 0.8878 and accuracy 0.9802 in comparison to AI based system where precision was 0.7399 and accuracy stood at 0.8878.
Conclusion: The AI Based system could identify platelet clumps and macro platelet better. Sysmex XN 1000 analyzer has better correlation with manual method of estimation than AI based system with manual count. The advent of new technology needs more studies before they could be used in clinical practice.

Note: This article has been accepted for publication in the next issue.  A peer‑reviewed version will be posted soon.
Submit Manuscript

Chat with us on WhatsApp

Welcome to Medwin Publishers. How can we help you today?