MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641095062 A) filed by Adhiparasakthi Engineering College, Melmaruvathur on August 05, 2026, for Tcm-Vae Framework For Predictive Stratification Of Pancreatic Cancer.
Inventors include Dr. C. Dhaya; Mrs. P. Gajalakshmi; S. Neelaveni; Dr. R. Srivel; Ms. A. Jayanthi; and Ms. V. Thamarai Selvi.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: Pancreatic cancer remains one of the most lethal malignancies, with critically low survival rates and persistent challenges in early detection and personalized prognosis due to heterogeneous multi-modal datasets, data imbalance, and limited interpretability of existing predictive models. This invention proposes a Survival-Aware Topological-Causal Variational Autoencoder (TCM-VAE) Framework for Predictive Stratification of Pancreatic Cancer, integrating imaging, clinical, laboratory, and genetic patient data into a unified intelligent architecture. The framework processes multi-modal inputs through a comprehensive preprocessing pipeline — encompassing data cleaning, normalization, encoding, SMOTE/GAN-based augmentation, and PCA/VAE-based dimensionality reduction — followed by Transfer Learning-based feature extraction using pre-trained CNN models (ResNet, EfficientNet). The TCM-VAE model generates robust, survival-informed latent representations through variational reconstruction, survival-aware loss functions
Disclaimer: Curated by HT Syndication.
