Automated IMD and RSFO Detection in Medical Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The identification of implanted medical devices (IMDs) and retained surgical foreign objects (RSFOs) in medical images is challenging due to the large variety of devices, increasing number of IMDs, and limited sensitivity and specificity of current radiographic screening methods, leading to delayed diagnosis and potential medical complications.
Innovation Solution
A diagnostic image analysis system using pattern recognition and computer vision algorithms integrated into the picture archiving and communication system (PACS) environment for rapid recognition of IMDs and RSFOs on X-rays, CT, ultrasound, and MRI images, with a database providing device-specific information and management guidance, and geometric hashing algorithms for precise detection of RSFOs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radiographic screening methods are used to identify IMDs and RSFOs, then the process is simple and widely available, but the sensitivity and specificity are limited leading to delayed diagnosis
Solution Approach 1:
The patent introduces an intermediary software system that acts as a bridge between existing radiographic imaging equipment and clinical decision-making. This software intermediary automatically analyzes images to detect IMDs and RSFOs, enhancing the capabilities of simple radiographic screening without requiring new hardware, thereby improving detection accuracy while maintaining workflow efficiency
Solution Approach 2:
The patent replaces manual visual inspection by radiologists with automated computer-based image analysis algorithms. This substitution of mechanical/manual analysis with automated computational processing significantly improves detection sensitivity and specificity while reducing diagnostic delay, as the software can process images rapidly without human fatigue or subjectivity
2Adaptability or versatility
If the variety of IMDs increases, then more patient needs are met, but identification becomes more difficult for radiology specialists
Solution Approach 1:
The patent creates a comprehensive database of reference images and device characteristics that serves as a library of copies of various IMD types. The software compares detected objects against this reference library to automatically identify device types, manufacturers, and models. This copying approach allows the system to handle increasing device variety without increasing radiologist workload or identification complexity
Solution Approach 2:
The patent develops a universal identification system that can detect and characterize multiple types of IMDs (pacemakers, defibrillators, neural stimulators, etc.) and RSFOs using a single software platform. This multi-functional approach consolidates what would otherwise require multiple specialized tools, maintaining simplicity despite the increasing diversity of medical devices
3Ease of operation
If RF technology is used to identify IMDs, then wireless data transmission is enabled, but radio interference problems and security issues arise
Solution Approach 1:
The patent introduces an intermediary image analysis system that indirectly identifies IMDs through their radiographic appearance rather than direct RF communication. This intermediary approach maintains identification convenience while avoiding the radio interference and security vulnerabilities associated with direct RF-based identification methods
Data Source
AI summary
The present invention involves a computer-based system and method for detecting and identifying implantable medical devices (“IMDs”) and/or retained foreign objects (“RFSOs”) from diagnostic medical images. In some embodiments, the system provides further identification—information on the particular IMD or RSFO that has been recognized. For example, the system may be configures to provide information feedback regarding the IMD, such as detailed manual information, safety alerts, recall, access to its structural integrity, and/or suggested courses of action in a specific clinical setting/troubleshooting. Embodiments are contemplated in which the system is configured to report possible 3D locations of RSFOs in the surgical field/images.