Open Access Open Access  Restricted Access Subscription Access

Deep Learning-Based Detection of Liver Cirrhosis Using Convolutional Neural Networks and TensorFlow

Muktha Desai, Ramesh B E, Aishwarya CN, Bhavana P M, Deepika H R, Chandana S J

Abstract


Liver cirrhosis is a chronic and progressive condition marked by irreversible scarring and fibrotic remodeling of liver tissues. Over time, this damage significantly compromises liver function and can result in severe, life-threatening complications if left untreated. Early and accurate detection is crucial, as timely medical intervention can improve prognosis and reduce the risk of advanced liver failure. Conventional diagnostic techniques, such as liver biopsies and expert-dependent imaging assessments, often involve invasive procedures, are time-intensive, and can be influenced by subjective judgment.

This research proposes a fully automated detection system for liver cirrhosis using Convolutional Neural Networks (CNNs) implemented within the TensorFlow framework. The system is developed using a supervised learning approach, where a custom dataset of liver images is annotated and labeled as either cirrhotic or non-cirrhotic. The deep CNN architecture is capable of autonomously learning and extracting intricate spatial patterns directly from the input images, eliminating the need for manual feature engineering.

To ensure the robustness of the model and mitigate overfitting, multiple image preprocessing and augmentation strategies were applied, including normalization, image rotation, and horizontal flipping. The dataset was divided into training and validation sets, and performance was assessed using metrics such as precision, recall, F1-score, and overall accuracy. The model exhibited strong classification performance, indicating its effectiveness in identifying cirrhosis.

This study demonstrates the potential of deep learning in medical image analysis and underscores the practicality of deploying CNN-based systems for non-invasive, efficient, and scalable liver disease screening, particularly in areas with limited access to specialized healthcare facilities.

Full Text:

PDF

References


Zhang, Y., & Liu, Y. (2020). Deep convolutional networks for liver disease diagnosis using ultrasound images. Biomedical Signal Processing and Control, 58, 101838 .

Kumar, R., & Singh, S. (2021). Liver cirrhosis detection using transfer learning with medical imaging data. Procedia Computer Science, 184, 532–538.

Almotairi, S., Elleithy, K., & Almazroi, A. (2022). Deep learning for medical Image categorization: A review on CNN architectures. Healthcare Technology Letters, 9(3), 81–89.

Chen, H., Shen, C., & Qin, J. (2021). Classification of liver diseases using deep convolutional neural networks. Computers in Biology and Medicine, 139, 104943.Camacho, N.G. (2024). "The Role of AI in Earthquake Prediction: Forecasting Techniques and Challenges." Journal of Earthquake Engineering and Resilience.

He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In the Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 770–778).

Das, S., & Mukherjee, S. (2019). An approach to detect liver diseases using CNN and optimized preprocessing. International Journal of Imaging Systems and Technology, 29(4), 539–546..

Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.

Kermany, D., Zhang, K., & Goldbaum, M. (2018). Identifying medical diagnoses using deep learning and large image datasets. Cell, 172(5), 1122–1131.

Rawat, W., & Wang, Z. (2017). Deep learning for image classification: A comprehensive review. Neurocomputing, 234, 19–34.

Wahid, A., & Khan, M. A. (2022). A hybrid CNN-based model for early liver cirrhosis prediction using ultrasound images. Journal of Healthcare Engineering, 2022, Article ID 8413947.


Refbacks

  • There are currently no refbacks.