Renal CT Image Classification Based on Densely Connected Convolutional Networks

Guangjie Qian

2024

Abstract

In response to the current situation of the increasing incidence of kidney diseases worldwide, the efficiency of traditional clinical diagnosis may not be enough to cope with future needs. Compared with traditional methods of clinical diagnosis, the automatic classification of renal computed tomography (CT) images based on convolutional neural networks (CNN) in this study has the potential to significantly improve the efficiency and accuracy of clinical diagnosis. In the paper, the Densely Connected Convolutional Networks 121 (DenseNet121) model is selected for training on 12,446 CT images, which include categories such as kidney cysts, kidney stones, tumors, and normal tissues. The model training was performed using an early stopping strategy and multi-cycle validation loss assessment. Subsequently, the model was tested on an independent test set to achieve an impressive accuracy of 0.9351 and a precision of 0.9393. The experiments conducted in this study have garnered a good response, and their high accuracy could potentially enhance the efficiency of clinical diagnosis and provide better safety for patients.

Paper Citation

in Harvard Style

Qian G. (2024). Renal CT Image Classification Based on Densely Connected Convolutional Networks. In Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-754-2, SciTePress, pages 115-119. DOI: 10.5220/0013510200004619

in Bibtex Style

@conference{daml24,
author={Guangjie Qian},
title={Renal CT Image Classification Based on Densely Connected Convolutional Networks},
booktitle={Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2024},
pages={115-119},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013510200004619},
isbn={978-989-758-754-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Renal CT Image Classification Based on Densely Connected Convolutional Networks
SN - 978-989-758-754-2
AU - Qian G.
PY - 2024
SP - 115
EP - 119
DO - 10.5220/0013510200004619
PB - SciTePress