Transfer Learning for Pancreatic Ductal Adenocarcinoma Detection: A Comprehensive Review

Read full text Summarise with VIRAT
Authors: Jonathan W. Song
Cited by: 0
References: 27

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Transfer learning has demonstrated significant potential within artificial intelligence, offering previously unrecognized advantages in medical imaging and cancer diagnosis.Transfer learning requires less data than traditional deep learning to develop models for new tasks.This review explores how transfer learning techniques can improve both the detection and diagnostic accuracy of pancreatic ductal adenocarcinoma (PDAC), a highly lethal and difficult-to-detect cancer.This review analyzes transfer learning applications to three key data types: computed tomography scans, ultrasound scans, and cell biopsies.While still a young field, early findings suggest that transfer learning improves diagnostic accuracy while reducing the need for data, making it an efficient alternative to traditional deep learning.Transfer learning achieved AUC scores comparable to deep learning and demonstrated higher accuracy than human professionals.However, there is still more to be done in this field, especially the need for further studies to validate transfer learning in PDAC detection.This research underscores the potential use of transfer learning in advancing more effective diagnostics for PDAC, which has significant potential to improve the current poor outcomes.

Fields of Study & Subjects

AI in cancer detection Radiomics and Machine Learning in Medical Imaging Domain Adaptation and Few-Shot Learning

Publication details

Journal / source
International journal of high school research
Volume
8
Issue
4
Pages
36–46
Publication date
28 February 2026
Language
EN