Advanced Detection of Liver Cirrhosis Utilizing LSTM and Neural Networks

• Published in: Communications and Signal Processing, International Conference on pp. 1685 - 1690
• Main Authors: Julian, B Annika, Annie, T A
• Format: Conference Proceeding
• Language: English
• Published: IEEE 05.06.2025
• Subjects: Analytical models; Biological system modeling; Clinical data; CNN; Convolutional neural networks; Data models; Deep learning; Feature extraction; Histopathology analysis; Hybrid model; Liver cirrhosis; Liver disease prediction; Long short term memory; LSTM; Machine learning; Medical imaging; Multimodal data fusion; Neural networks; Predictive models; Signal processing
• ISSN: 2836-1873
• DOI: 10.1109/ICCSP64183.2025.11088752

Even with the advances in medical diagnosis, cir rhosis of the liver still constitutes a huge global health burden, usually detected in advanced stages. A hybrid deep learning model is presented in this work, which employs convolutional neural networks for processing histopathological images and long short-term memory networks for structured clinical data analysis. . The LSTM model takes in time-series patient records and captures sequential dependencies in liver function trends while the CNN extracts spatial features from liver biopsy images used to detect cirrhotic patterns. The features from CNN and LSTM provide a combined feature set with a strong ability to classify the data well. The experimental results reveal that the accuracy is distinctly better than traditional machine learning models, implying that the model is capable of early cirrhosis detection and of predicting progression.