MUMBAI, India, July 31 -- Intellectual Property India has published a patent application (202521004893 A) filed by Tata Consultancy Services Limited on January 21, 2025, for Automatic Detection Of Implant Loosening Based On Segmented Zones And Deep Neural Network.
Inventors include Kanakatte Gurumurthy, Aparna; Bhatia, Divya Manoharlal; Mukherjee, Rupsha; Poduval, Murali; and Ghose, Avik.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: ABSTRACT AUTOMATIC DETECTION OF IMPLANT LOOSENING BASED ON SEGMENTED ZONES AND DEEP NEURAL NETWORK A method and system for automatic detection of implant loosening based on segmented zones and deep neural network is disclosed. The method utilizes three networks such as convolutional neural network, encoder-decoder segmentation network and zonal loosening classification network for detecting loosening in implant images. The trained convolutional neural network classifies an implant image as fit or unfit image wherein it is trained using an annotated ground truth file. Further the fit implant image is segmented into zones using the trained encoder-decoder segmentation network which is trained using ground truth masks of the fit implant images. Finally, these segmented zones are used for detecting loose zones and thereby loose implant images using the zonal loosening classification network. Utilizing zonal loosening information helps surgeons in making well-informed decisions regarding the extent of loosening, affected zones, early signs of loosening, progress monitoring, and the necessity for revision surgery. [
Disclaimer: Curated by HT Syndication.
