Prof. CAI Jing pioneer safer, AI-powered MRI imaging for nasopharyngeal carcinoma

PolyU researchers have made a breakthrough in MRI imaging for nasopharyngeal carcinoma (NPC). Traditionally, MRI scans for NPC rely on gadolinium-based contrast agents (GBCAs) to enhance tumour visibility. However, GBCAs pose health risks, including nephrogenic systemic fibrosis and potential accumulation of gadolinium in human brain.

Led by Prof. CAI Jing, Management Committee Member of Research Institute for Intelligent Wearable Systems (RI-IWEAR), Research Institute for Smart Ageing (RISA), Professor and Head of Department of Health Technology and Informatics, and his team have developed an AI-powered, contrast-free MRI imaging, offering a safer diagnostic alternative. Their innovative Multimodality-Guided Synergistic Neural Network (MMgSN-Net) synthesises virtual contrast-enhanced images using information from T1- and T2-weighted scans, eliminating the use of GBCAs. Building on this, the team has introduced the Pixelwise Gradient Model with Generative Adversarial Network for Virtual Contrast Enhancement (PGMGVCE), which further improves image texture and detail.

Comparative studies show that PGMGVCE matches MMgSN-Net in accuracy and excels in texture representation, which is crucial for precisely delineating the boundaries of NPC tumours. The findings highlight the importance of integrating multiple imaging modalities and advanced AI techniques to enhance diagnostic quality.

These advancements promise safer MRI procedures for NPC patients, particularly those experiencing side-effects from contrast agents, and pave the way for improved cancer diagnosis and treatment planning. Future research will focus on expanding datasets and incorporating more MRI modalities to further boost diagnostic capabilities.

Online coverage:

Eurek Alert - https://polyu.me/45hPQ5R

Bioengineer - https://polyu.me/413mNAp

Technology Network - https://polyu.me/4mH2bWU

Mirage News - https://polyu.me/3UsjhvF