Automatic Identification and Counting System of Thrombocytes in Chickens (Gallus Domesticus) via Deep Learning

Accurate identification and counting of thrombocytes (platelets) in chicken blood are critical for assessing health, investigating diseases, and ensuring optimal management practices. Traditional manual counting methods were time-consuming, labor-intensive, and prone to human error. This study introduced an automatic identification and counting system for thrombocytes in chickens, utilizing digital imaging and machine learning techniques. High-resolution microscopic images of blood samples were processed and analyzed using convolutional neural networks (CNNs) to accurately identify and count thrombocytes. The automated system enhanced diagnostic efficiency, consistency, and reliability, which significantly reduced the time required for analysis. The result of the system's potential confirmed the system can provide a promising innovative tool for supporting timely health interventions and sustainable poultry farming by providing a robust tool for veterinary diagnosis.