Hemostasis Today

October, 2026
October 2026
M T W T F S S
 1234
567891011
12131415161718
19202122232425
262728293031  
Victor Daniel: Deep Learning for Automated Platelet Classification
Oct 24, 2026, 16:18

Victor Daniel: Deep Learning for Automated Platelet Classification

Victor Daniel, Data Scientist at Department of Social Health and Care, shared post on LinkedIn about a recent article, he and his colleagues, published in Journal of Imaging, adding:

“I’m pleased to announce that my scientific article titled, ‘Automated Platelet Classification Under Different Solvent Conditions Through Deep Learning’, has been published in the Journal of Imaging, an MDPI journal.

The work stemmed from my MSc project at the University of Hull, and the study set out to build a methodological foundation for the automated classification of platelets using pre-trained convolutional neural networks.

Platelets change shape when they are activated by compounds such as zinc and milrinone, and classifying those changes by eye is subjective and time-consuming.

I evaluated ten architectures on microscopy images from Hull York Medical School, comparing treated and untreated platelets in a binary classification and four treatment groups in a multiclass classification.

Xception and InceptionV3 performed best on both tasks, offering a reproducible alternative to manual classification with potential value for platelet research and, in time, clinical diagnostics.

I would like to thank my supervisor and co-author, Eva Maria Sousa, for her collaboration, guidance and support throughout. My thanks also go to Charlie A. Coupland, Nick Sander, Julius Sechang Mboli, and Simon Calaminus for their contributions, and to Hull York Medical School for the data.”

Title: Automated Platelet Classification Under Different Solvent Conditions Through Deep Learning

Authors: Eva Maria Sousa, Charlie A. Coupland, Nick Sander, Julius Sechang Mboli,  Simon Calaminus

Victor Daniel: Deep Learning for Automated Platelet ClassificationStay updated on all scientific advances with Hemostasis Today.