Quality and Safety Monitoring of Clinical Computed Tomography Practice

Abstract ?

Fifty percent of radiation exposure to the United States population is from medical imaging, half of which come from over 80 million computed tomography (CT) scans performed every year.

This significant use has raised notable concerns regarding utilization costs, inappropriate applications, and the associated radiation risk.

A recent Medicare reports hundreds of hospitals across the country have needlessly scanned their patients twice on the same day.

Unnecessary repeated scans expose patients to extra radiation while increase expense, decrease reimbursement, and increase liability for the providers.

Thus, the legislative and regulatory organizations have encouraged and mandated stricter oversight of medical imaging usage and its associated radiation.

Concurrent to the mandate to better manage imaging risk is to ensure its value: One of the quintessential but most understated pillars of current CT practice is CT?s high value in caring for illness and injury across all ages.

To ensure this benefit of CT examinations, there needs to be a careful balance between image quality and radiation safety.

A poor quality, overly low dose exam is a disservice to the care of the patient while an exam with more radiation dose than necessary can undermine its safety.

Therefore, proper CT imaging requires a comprehensive combined quality and dose monitoring program on a patient-by-patient basis to properly understand, manage, and mitigate radiation risk.

Unfortunately, currently there is no such software available in the market that can simultaneously monitor CT radiation dose and its corresponding image quality.

The objective of this fast-track project is to develop a first Software as a Service (SaaS)-based performance monitoring platform to track radiation dose and image quality concurrently.

The platform aims to provide essential data and insight to improve CT performance through considering both patient safety and imaging quality simultaneously.

Specifically, the project will develop a product that offers 1) a robust multi-infrastructure workflow to connect and collect clinical CT quality- and dose-relevant data; 2) a suite of patient-specific CT radiation dose and image quality assessment algorithms; 3) an implementation of a task manager to chain and automate dose and image quality calculations towards CT performance assessment; 4) a combined SQL- NoSQL database system for structured and un-structured quality- and dose-relevant data storage; and 5) a web-based dashboard for interactive and easy-to-use data analysis and visualization.

The development utilizes machine-learning methodologies to devise robust and scalable techniques for extracting meaningful knowledge from hundreds of thousands of patient images. The system quantifies value for a value-based practice.

It serves as an essential tool to minimize variability in quality and dose across a practice, to ensure consistent use of CT technology, and to ensure imaging radiation dose and quality levels match expected values. The system will be beta tested at healthcare facilities paving the way toward effective commercialization. 1