MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202621059450 A) filed by Dr. Tatiraju. V. Rajani Kanth; and Dr. Aiman Fatma on May 10, 2026, for Method And System For Predicting Breast Cancer Recurrence Using Multimodal-Multitask Deep Learning.
Inventors include Dr. Tatiraju. V. Rajani Kanth; and Dr. Aiman Fatma.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: The present invention relates to a method and system for predicting breast cancer recurrence using a multimodal-multitask deep learning framework. The system receives multimodal patient data including histopathological images, radiological images, genomic or molecular markers, clinical parameters, laboratory values, treatment history, and longitudinal follow-up records. A multimodal preprocessing module standardizes the received data by image normalization, segmentation, encoding, missing-value handling, and temporal alignment. Modality-specific feature extraction modules generate pathology, radiology, molecular, clinical, treatment, and follow-up feature vectors. A cross-modal fusion module combines the feature vectors into a unified patient representation. A multitask prediction module simultaneously predicts overall recurrence probability, local recurrence likelihood, distant metastasis likelihood, recurrence-free survival estimate, and treatment-response risk. A recurrence-risk scoring module categorizes patients into clinically usable risk groups. An explainability module generates heatmaps, biomarker relevance indicators, clinical-variable importance values, and modality contribution scores for assisting oncologists in personalized post-treatment monitoring. (Accompanied Figure No. 1-2)
Disclaimer: Curated by HT Syndication.
