Hierarchical Detection for Remote CAD Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Thin-client devices, such as personal data assistants and cell phones, lack the processing power and memory to rapidly detect objects from large medical scan data sets, and network bandwidth limitations hinder computer-assisted detection (CAD) performance, especially in three-dimensional data sets.
Innovation Solution
A hierarchical detection method is employed, where detected locations at coarser resolutions are used to limit data transmission, and higher resolution data is transmitted only for neighborhoods around previously detected locations, allowing progressive transmission and potential lossy compression without significant reduction in detection sensitivity, enabling CAD on thin-clients by offloading processing to a server.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If CAD is performed on workstations with full processing power, then detection accuracy is improved, but network bandwidth requirements increase significantly
Solution Approach 1:
The detection process is segmented into multiple resolution levels. First, a coarse-resolution detection is performed to identify candidate locations, then only the neighborhoods around these candidates are processed at higher resolutions. This segmentation allows the system to achieve high detection accuracy while transmitting significantly less data over the network, as only relevant regions at high resolution are sent to the client.
2Ease of operation
If thin-client devices are used for remote viewing, then ease of operation is improved, but processing power and memory are insufficient for rapid detection
Solution Approach 1:
A server acts as an intermediary between the data storage and the thin-client device. The server performs the computationally intensive detection processes, while the thin-client only needs to handle display and user interaction. This intermediary approach allows thin-clients to provide remote viewing capability without requiring them to have sufficient processing power for rapid detection of large medical datasets.
Solution Approach 2:
The system transitions from requiring all processing to occur locally on the client device to a distributed architecture where processing occurs on the server and results are transmitted to the client. This dimensional shift in where processing happens enables thin-client devices to function effectively for remote viewing without needing enhanced local processing power.
3Quantity of substance
If lossy compression is applied to reduce bandwidth, then network bandwidth requirements are reduced, but detection sensitivity may be reduced
Solution Approach 1:
Different regions of the medical image data are transmitted at different quality levels. Regions identified as candidates for detection are transmitted at higher quality with less compression, while other regions use more aggressive compression. This local quality approach maintains detection sensitivity for important regions while reducing overall bandwidth requirements through selective application of compression levels.
Data Source
AI summary
For cloud-based computer assisted detection, hierarchal detection is used, allowing detection on data at progressively greater resolutions. Detected locations at coarser resolutions are used to limit the data transmitted at greater resolutions. Data is only transmitted for neighborhoods around the previously detected locations. Subsequent detection using higher resolution data refines the locations, but only for regions associated with previous detection. By limiting the number and/or size of regions provided at greater resolutions based on the previous detection, the progressive transmission avoids transmission of some data. Additionally, or alternatively, lossy compression may be used without or with minimal reduction in detection sensitivity.