MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202611080379 A) filed by Dr. Vijay Sharma; Prof. Dr. Raj Kumari; Anamika Kulshrestha; Dr. Sandeep Arora; Tarak Nath Khatua; Ms. Munjewar Seema Chandpasha; Jeevan Chandra Pandey; Nayyar Parvez; Hansraj Bhagwan Teple; Dr. Ashutosh Singh; Prof. Devendra Dhariwal; Dr. V. Poongodi; Dr. Surabhi Sharma; and Ms. Abhilasha Tiwari on June 30, 2026, for Indian Utility Patent Publication: Machine Learning-Based Optimization Of Controlled Drug Delivery Systems For Targeted Therapy.
Inventors include Dr. Vijay Sharma; Prof. Dr. Raj Kumari; Anamika Kulshrestha; Dr. Sandeep Arora; Tarak Nath Khatua; Ms. Munjewar Seema Chandpasha; Jeevan Chandra Pandey; Nayyar Parvez; Hansraj Bhagwan Teple; Dr. Ashutosh Singh; Prof. Devendra Dhariwal; Dr. V. Poongodi; Dr. Surabhi Sharma; and Ms. Abhilasha Tiwari.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: A hardware-accelerated machine learning system (100) for optimizing controlled drug delivery formulations comprises a sensor interface module (102) coupled to a microfluidic sensor array (148), an FPGA-based processing unit (106) incorporating a finite state machine controller (116) that orchestrates a pipelined sequence of a preprocessing pipeline stage (108), a feature extraction engine (110), a machine learning inference accelerator (112), and a formulation parameter sweep engine (114). An IOMMU (118) enforces memory isolation through base and limit address registers locked by a write-lock latch (120) preventing modification until hardware reset. A bus filter (126) on an interconnect bus (124) blocks unauthorized transactions using latched policy registers (122). Validated formulation profiles are committed to a write-protected memory region (130) via a monotonic commit register (132) creating an immutable audit trail. The system enables deterministic, hardware-isolated optimization of drug delivery formulation parameters with tamper-resistant validation storage.
Disclaimer: Curated by HT Syndication.
