Abstract
The development of the paper addresses the interdisciplinary challenge of accelerating data processing for flow cytometry. Traditional use of the pandas library in Python is limited by sequential calculations: processing 500 000 leukocyte cells takes 15 seconds. This long waiting time causes user discomfort and may lead to medical errors. The aim of the study is to develop an original method for increasing processing speed using parallel calculations, building on the authors’ previous work on leukocyte population recognition. The approach employs algorithms from the Ray distributed computing framework and the modin. pandas library. Experiments on clinical laboratory equipment confirm the solution’s effectiveness compared to traditional pandas algorithms. The use of the developed method reduced the processing time for 500 000 cells by 11 times, using data from an open Internet source. The method is scalable for medical diagnostics, treatment, forecasting, and artificial intelligence systems.