Sign in Subscribe
Cancer

New Gastric Cancer Detection Method Shows Promising Outcomes

Strategic Insights

— 1 min read

Researchers at Zhejiang University have made significant breakthroughs in developing a non-invasive, convenient, and rapid method for gastric cancer screening using a deep learning algorithm with Robust Tabular AutoEncoder Interpolator (RTAEI). The study, published in IEEE Transactions on Consumer Electronics, highlights the potential of AI in healthcare and the challenges in utilizing medical data due to its imbalance. The researchers used questionnaire survey data and crucial biochemical indicators from patients to form their dataset, which was then balanced using RTAEI to improve the accuracy of gastric cancer detection.

Key Takeaways:

  • The study used a deep learning algorithm with RTAEI to screen for gastric cancer patients, achieving an AUC of 0.792, which indicates high accuracy in detection.
  • The method is non-invasive, convenient, and rapid, making it suitable for widespread use in healthcare.
  • The study collected questionnaire survey data and crucial biochemical indicators from patients to form their dataset, which covers dietary habits, H pylori infection status, and other relevant factors.
  • The researchers used SMOTE (Synthetic Minority Over-sampling Technique) to balance the dataset and improve the model's performance.
  • The study highlights the potential of AI in healthcare, particularly in cancer detection and diagnosis.
  • The researchers involved in this study include Jian Wu, Zihan Ma, Kai Zhang, Haozhong Ma, Yuling Tong, Yongfeng Ding, Honghao Gao, and Hongxia Xu.
  • The study was supported by the National Natural Science Foundation of China (NSFC) and the Zhejiang Key Research and Development Program of China.

Statistics:

  • The highest AUC achieved by the RTAEI algorithm was 0.792, indicating high accuracy in gastric cancer detection.
  • The study collected questionnaire survey data and biochemical indicators from 1000 patients, which were used to form their dataset.
  • The researchers used 80% of the dataset for training and 20% for testing.
  • The study used a robust variational autoencoder to encode sparse data into a dense latent space.
  • The SMOTE technique was used to generate synthetic samples to balance the dataset.

Sources:

  • Rtaei: Robust Tabular Autoencoder Interpolator To Gastric Cancer Innovative Detection for Deep Learning Empowered Healthcare Electronics. Ieee Transactions On Consumer Electronics, 2025;71(1):1482-1494.
  • Health & Medicine Week. July 18, 2025; p 2784.

Read more