• Rakshath T. R.
• Chethan P. J.
• Shreya Hebbar
• Sindhu M. N.
• Varun D. S.

Liver fibrosis is a progressive condition characterized by the accumulation of scar tissue in the liver, often developing silently until advanced stages. Conventional diagnostic methods such as liver biopsy, while effective, are invasive, costly, and subject to sampling errors. This study proposes a deep learning-based framework for noninvasive detection and staging of liver fibrosis using ultrasound elastography images. The approach integrates a convolutional neural network (CNN) trained to identify subtle textural and structural changes associated with fibrotic progression. The preprocessing pipeline includes grayscale conversion, noise reduction, and high-pass filtering to enhance image clarity. Segmentation isolates the liver region, followed by feature extraction to generate statistical descriptors for classification. The model successfully distinguishes between four fibrosis stages, demonstrating high accuracy and recall. This technique offers a reliable, scalable alternative to traditional methods, with potential for early diagnosis and broader accessibility in clinical settings.

Ultrasound elastography, Noninvasive diagnosis, Edge Detection method, Feature extraction, Ultrasound Images, Feature extraction, CNN

"Intelligent Liver Fibrosis Assessment with Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 11, page no.f463-f466, November-2025, Available :http://www.jetir.org/papers/JETIR2511576.pdf

"Intelligent Liver Fibrosis Assessment with Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 11, page no. ppf463-f466, November-2025, Available at : http://www.jetir.org/papers/JETIR2511576.pdf