Intellectual Property India Publishes Patent Application for A Deep Learning And Cloud-Assisted Iot System For Automated Image-Based Diagnosis And Decision Support

MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118511 A) filed by Dr. Sake Madhu; Ms. K. B. V. K. Nagasree; Dr. Karnam Sreenu; Dr. Neeluri Suresh; Dr. B. Satyanarayana; Rachel R; and Pillalamarri Laxman on October 04, 2026, for A Deep Learning And Cloud-Assisted Iot System For Automated Image-Based Diagnosis And Decision Support.

Inventors include Dr. Sake Madhu; Ms. K. B. V. K. Nagasree; Dr. Karnam Sreenu; Dr. Neeluri Suresh; Dr. B. Satyanarayana; Rachel R; and Pillalamarri Laxman.

The application for the patent was published on October 09, 2026, under issue no. 41/2026.

Abstract: ABSTRACT OF THE INVENTION: [044] A deep-learning and cloud-assisted Internet-of-Things system for automated image-based diagnosis and clinical decision support is disclosed. Medical images are captured by an IoT-enabled front-end (smartphone, dermatoscope, fundus camera or digital radiography module) and processed on a resource-constrained edge node that performs quality assessment, pre-processing and rapid inference using a lightweight convolutional neural network. When the edge confidence score falls below a configurable threshold or a high-risk class is detected, the image or a compact feature embedding is securely transmitted to a cloud platform. The cloud hosts high-capacity transfer-learned deep neural networks (ResNet, DenseNet, EfficientNet, Vision Transformer or hybrid architectures) that produce refined multi-class or multi-label diagnoses together with Grad-CAM visual explanations. A decision-support engine maps the predictions onto evidence- based clinical guidelines, generating triage categories and recommended actions such as urgent specialist referral. Results, heat-maps and recommendations are returned to a role-based remote interface accessible by clinicians and, optionally, patients. The system operates resiliently under intermittent connectivity by caching edge results and synchronising later. Evaluations on public medical-image benchmarks demonstrate cloud-stage accuracies of 93-97 % with edge triage latency below one second, enabling scalable, privacy-aware point-of-care screening and timely decision support in both well-resourced and resource-constrained healthcare settings. [045] The invention thereby overcomes the accuracy-latency-privacy trade-offs of purely edge or purely cloud diagnostic systems and provides an integrated, clinically oriented platform for automated image analysis and decision support.

Disclaimer: Curated by HT Syndication.