by John R. Fischer
, Senior Reporter | November 06, 2018
PaxeraHealth is developing an AI-based module that will be integrated with the PaxeraUltima PACS system.
PaxeraAI will collect information through a chatbot-like questionnaire for instant analysis, making it “smarter” with each new use. Data is stored in the PaxeraAI database, making it accessible to any PaxeraUltima PACS user through the company’s internal network.
“Most AI in use today is static. While the studies may go back several years and are extensive, the amount of data that are relevant and associated with those studies is fairly limited and mostly just consists of the diagnostic report,” Dr. Mohamed Shoura, founder and CEO of PaxeraHealth, said in a statement. “We are creating a dynamic system that provides a higher level of information and collaboration to both the radiologist and clinician so that AI can provide even more value to the diagnostic interpretation.”
Designed to capture the actions and behavior of the radiologist, including input from other clinical systems, the solution enables users to selectively submit anonymized studies to the growing PaxeraAI database.
The announcement of its development reflects the growing popularity around AI, which is projected to be a $2 billion industry in 2018 that will provide greater productivity and accuracy for personalized treatment planning in diagnostic imaging. Deployment of such technology is further supported by its use in clinical trials, producing outcomes with accuracy of 97 percent or higher.
The aim behind this and other AI solutions is to assist radiologists in managing larger volumes of imaging, particularly in areas lacking in such expertise.
“Most companies look at AI from a technical perspective,” said Shoura. “We are looking at it from a business perspective, allowing access to user feedback and creating an open system that goes well beyond just the interpretation generated by the AI algorithms. A silent, static AI system is comparatively limited. You need one that interacts with the end user as well. Our system will provide a continuously improving environment that will leverage the heterogenous data we collect from all over the world.”
Initial use of the module’s algorithms will focus on mammography and chest X-ray studies, with expansions to other areas set to take place through further development of the project.
Its deployment will be included in the PaxeraHealth’s Ultima PACS system, of which there are 2,500 installations currently, and will be showcased as a “work-in-progress” this month at RSNA 2018 in Chicago.
The company expects to release simple access to its features through an API in early 2019.