GE HealthCare Technologies has patented a method for computer-aided diagnosis using deep neural networks trained on radiologists’ diagnoses. The system customizes models based on geographic regions and institutions, assisting radiologists in making accurate diagnoses. The method involves training, generating, displaying, receiving feedback, and updating the models. GlobalData’s report on GE HealthCare Technologies gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on GE HealthCare Technologies, Nucleoside chemical synthesis was a key innovation area identified from patents. GE HealthCare Technologies's grant share as of May 2024 was 39%. Grant share is based on the ratio of number of grants to total number of patents.
Computer-aided diagnosis system using neural network models
A recently granted patent (Publication Number: US12014823B2) outlines a method that involves training a series of neural network models on diagnostic decisions made by radiologists to generate personalized computer-aided diagnoses for patients. The method includes training a baseline neural network model on diagnostic decisions of radiologists, customizing this model based on geographic regions and institutions, and further refining it based on individual radiologists' diagnostic decisions. The radiologist model generated is then used to provide computer-aided diagnoses for patients, with feedback from radiologists used to selectively update the various neural network models involved in the process. This approach aims to create an ensemble of deep neural network models that maintain coherence within a radiologist ecosystem, ensuring personalized and accurate diagnoses.
Furthermore, the patent describes a hierarchical ensemble structure that continuously updates based on feedback received, allowing for corrections to both the outputs of the ensemble and the diagnoses made by radiologists. The computer-aided diagnoses provided are tailored to individual radiologists, creating digital twin diagnoses that are specific to each practitioner. The method also allows for the evaluation of the performance of the radiologist model in comparison to institution and geographic models, enabling the identification of any divergences. Additionally, the patent details a process for duplicating and deploying baseline models to form additional geographic, institution, and radiologist models within the ensemble, with mechanisms in place to detect and propagate superior performance throughout the network of models. This innovative method aims to enhance diagnostic accuracy and personalization in radiology through the utilization of deep neural networks and feedback mechanisms.
To know more about GlobalData’s detailed insights on GE HealthCare Technologies, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
