当前位置: X-MOL 学术Imaging Sci. J. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Infection level identification for leukemia detection using optimized Support Vector Neural Network
The Imaging Science Journal ( IF 0.871 ) Pub Date : 2019-11-17 , DOI: 10.1080/13682199.2019.1701172
Biplab Kanti Das 1 , Himadri Sekhar Dutta 2
Affiliation  

ABSTRACT Leukemia is the abnormal and uncontrolled development of the white blood cells, known as leukocytes, in the blood. The manual methods used for counting the blast cells have some demerits, and so automatic method must be employed. This paper proposes the Salp Swarm integrated Dolphin Echolocation-based Support Vector Neural Network (SSDE-SVNN) classifier to detect leukemia in its early stages. The pre-processed blood smear image is subjected to segmentation with the use of LUV transformation and Adaptive thresholding. The features, such as area, shape, texture, and empirical mode decomposition are extracted from the segments. The proposed classifier is used for the counting of blast cells based on the extracted features. The accuracy, specificity, and sensitivity of the proposed classifier are obtained as 0.97, 0.97, and 1, respectively, and the Mean Square Error (MSE) is noted as 0.1272.

中文翻译:

使用优化的支持向量神经网络识别白血病检测的感染水平

摘要 白血病是血液中白细胞(称为白细胞)的异常和不受控制的发育。人工计数胚细胞的方法有一定的缺点,必须采用自动方法。本文提出了 Salp Swarm 集成的基于海豚回声定位的支持向量神经网络 (SSDE-SVNN) 分类器,用于早期检测白血病。预处理后的血涂片图像使用 LUV 变换和自适应阈值进行分割。从片段中提取特征,如面积、形状、纹理和经验模式分解。所提出的分类器用于基于提取的特征对原始细胞进行计数。所提出的分类器的准确度、特异性和灵敏度分别为 0.97、0.97 和 1,
更新日期:2019-11-17
down
wechat
bug