BlasT: Blood Clot Classification Model Using Transfer Learning Based Convolutional Neural Network

Blood cells contain a plethora of clotting factors vital for sustaining human life. Blood coagulation manages excessive blood loss from the blood vessels of an injured individual. Blood clots can appear due to various medical reasons in humans, such as platelet aggregation and converting fibrinogen into fibrin. Consequently, blood clots generate multiple disorders and even elevate the risk of death when not promptly diagnosed. Therefore, the prompt identification and treatment of blood clotting are paramount. In this study, we introduced an analytical approach for detecting blood clots. Also, we introduced a blood clot classification model: Blood Clot Classification Model using Transfer Learning and Convolutional Neural Network (BlasT). Our evaluation results demonstrate an impressive accuracy rate of 93.99% for the proposed model: BlasT. The primary objective of this study is to establish an early clot classification model using transfer learning in conjunction with a Convolutional Neural Network (CNN) layer. Our proposed BlasT model can identify blood clots swiftly and accurately. Furthermore, it allows timely notifications to medical professionals.