Acute Lymphoblastic Leukemia Detection Using Hypercomplex-Valued Convolutional Neural Networks

Abstract

This paper features convolutional neural networks defined on hypercomplex algebras applied to classify lymphocytes in blood smear digital microscopic images. Such classification is helpful for the diagnosis of acute lymphoblast leukemia (ALL), a type of blood cancer. We perform the classification task using eight hypercomplex-valued convolutional neural networks (HvCNNs) along with real-valued convolutional networks. Our results show that HvCNNs perform better than the real-valued model, showcasing higher accuracy with a much smaller number of parameters. Moreover, we found that HvCNNs based on Clifford algebras processing HSV-encoded images attained the highest observed accuracies. Precisely, our HvCNN yielded an average accuracy rate of 96.6% using the ALL-IDB2 dataset with a 50% train-test split, a value extremely close to the stateof-the-art models but using a much simpler architecture with significantly fewer parameters.

i-Net: a deep CNN model for white blood cancer segmentation and classification

“…It was observed that diagnosis of acute leukaemia has generated a lot of attention and many scholars have employed deep neural networks for the same. The use of hypercomplex-valued network (HVN) was investigated by Vieira and Vale [38], using 8 complex-valued CNN. Analysis showed that their proposed hypercomplex CNN produced commendable results at 96.6% when tested on ALL-IDB2 dataset using 50% of the data for texting.…”

Dual Quaternion Rotational and Translational Equivariance in 3D Rigid Motion Modelling

“…Learning representations of multichannel data using higher dimension algebras has been proved to outperform real-valued approaches in several problems [6,7,8,9]. Quaternion-valued networks already boast works on a significant array of diverse tasks such as audio signal and image processing, computer vision, and dynamic system modeling [10,11,12,13].…”