Performance evaluation of the Cellbytes software in technically diverse conditions

Abstract

As society advances, so too must our tools and methodologies. Cellbytes software represents a significant leap forward in addressing the evolving needs of healthcare, specifically in the field of haematological diagnostics. This software has been developed to enhance the speed and efficiency of diagnosing haematological diseases, bridging the critical gap between increasing patient volumes and the existing shortage of medical specialists. The software is designed to analyse bone marrow aspirate crush samples, categorize cells and structures, and generate comprehensive information sheets for diagnosticians. The primary objective of this study is to validate the robustness of the Cellbytes software's algorithm with diverse scanning settings. While this is not a required step toward its classification as a medical device for diagnostic use, the results help the devices end-users to operate the software reliably. As a diagnostic support tool influencing clinical decisions, the algorithm's evaluation is a pivotal component of the compliance approval process. This study will therefore assess the software's accuracy, repeatability, and reliability.