MEDINDIA

Search Medindia

How AI Biomarkers Transform Thyroid Cancer Treatment

by Manjubashini on Jul 29 2025 2:12 PM
Listen to this article
0:00/0:00

AI-driven new scoring system in thyroid cancer progression outperforms 'one-size-fits-to-all' tradition.

How AI Biomarkers Transform Thyroid Cancer Treatment
Some patients with Differentiated Thyroid Carcinoma (DTC), a lethargic malignant tumor, persuaded carefully without any need for surgery. Even though, it is a medical hurdle to assess the disease progression in patients (1 Trusted Source
Optimized Dynamic Network Biomarker Deciphers a High-Resolution Heterogeneity Within Thyroid Cancer Molecular Subtypes

Go to source
).
New research from The First Affiliated Hospital of Zhengzhou University tackles thyroid cancer by developing a groundbreaking AI biomarker system. This model uses an enhanced dynamic network biomarker (DNB) algorithm and identified a 'tipping point' in Stage II of DTC that shows a shift from an unchanging state to rapid progression.

TOP INSIGHT

Did You Know

Did You Know?
Just with 12 genes! How “miniPC” (a #simplifiedaibiomarker) offers a #personalizedapproach to #thyroidcancertreatment. #aipoweredbiomarkers #thyroidcancerprogression #medindia

“Our analysis revealed that Stage II is a critical transition stage,” says corresponding author Prof. Xinguang Qiu.

TCPS Level: Discovering Individual Patient Risk

The team developed an automated scoring system called 'TCPS level' to estimate an individual patient risk that features early-warning molecular signals. “This score outperforms traditional staging in identifying high-risk individuals,” notes co-author Dr. Ge Zhang.

The researchers applied AI-based consensus clustering to over 1,100 thyroid cancer samples and identified three reproducible molecular subtypes, each with distinct immune profiles and progression risks. The most aggressive subtype was associated with the gene ASPH, which was experimentally validated.

To support clinical use, they developed a simplified classifier (miniPC) based on just 12 genes, enabling accurate subtype prediction across multiple datasets. “This tool offers a practical approach to personalized treatment planning,” says Dr. Haonan Zhang.

By integrating multi-omics data, machine learning, and single-cell analysis, the study provides new insights and tools for early risk stratification and targeted management of thyroid cancer.

Reference:
  1. Optimized Dynamic Network Biomarker Deciphers a High-Resolution Heterogeneity Within Thyroid Cancer Molecular Subtypes - (https://onlinelibrary.wiley.com/doi/10.1002/mdr2.70004)

Source-Eurekalert



⬆️