This paper presents a comparison between two different Computer Aided Diagnosis systems for classification of five types of leucocytes located in the tail of a Peripheral Blood Smears: Lymphocytes, Monocytes, Neutrophils, Basophils and Eosinophils. In particular, we have evaluated and compared the performance of a previous feature-based Back Propagation Neural Network classifier with the performance of two novel classifiers both based on Deep Learning using Convolutional Neural Networks introduced in this study. All the classifiers are built considering the same dataset of images acquired in a previous study. The experimental results, reported in terms of accuracy, sensitivity, specificity and precision, show that the different strategies could be compared and discussed from both clinical and technical point of view.
A novel deep learning approach in haematology for classification of leucocytes / Bevilacqua, Vitoantonio; Brunetti, Antonio; Trotta, Gianpaolo Francesco; De Marco, Domenico; Quercia, Marco Giuseppe; Buongiorno, Domenico; D’Introno, Alessia; Girardi, Francesco; Guarini, Attilio (SMART INNOVATION, SYSTEMS AND TECHNOLOGIES). - In: Quantifying and Processing Biomedical and Behavioral Signals / [a cura di] Anna Esposito, Marcos Faundez-Zanuy, Francesco Carlo Morabito, Eros Pasero. - STAMPA. - Cham, CH : Springer, 2019. - ISBN 978-3-319-95094-5. - pp. 265-274 [10.1007/978-3-319-95095-2_25]
A novel deep learning approach in haematology for classification of leucocytes
Bevilacqua, Vitoantonio;Brunetti, Antonio;Trotta, Gianpaolo Francesco;Quercia, Marco Giuseppe;Buongiorno, Domenico;
2019-01-01
Abstract
This paper presents a comparison between two different Computer Aided Diagnosis systems for classification of five types of leucocytes located in the tail of a Peripheral Blood Smears: Lymphocytes, Monocytes, Neutrophils, Basophils and Eosinophils. In particular, we have evaluated and compared the performance of a previous feature-based Back Propagation Neural Network classifier with the performance of two novel classifiers both based on Deep Learning using Convolutional Neural Networks introduced in this study. All the classifiers are built considering the same dataset of images acquired in a previous study. The experimental results, reported in terms of accuracy, sensitivity, specificity and precision, show that the different strategies could be compared and discussed from both clinical and technical point of view.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
