Multi-GPU Medical Image Reconstruction with Fault-Tolerant Task Assignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computed tomography (CT) image processing apparatuses face difficulties in reconstructing cross-sectional images in real-time when at least one graphics processing unit (GPU) malfunctions, leading to inefficiencies and potential diagnostic delays in emergency situations.
Innovation Solution
An apparatus and method that utilize a controller to monitor GPUs and assign priority operations, allowing the system to maintain normal processing speeds even with a malfunctioning GPU by redistributing tasks among the remaining processors, ensuring continuous reconstruction of cross-sectional images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a Multi-GPU architecture is used to reconstruct cross-sectional images, then image reconstruction speed is boosted, but system reliability deteriorates because the GPUs perform tasks interdependently and any malfunction prevents intended reconstruction
Solution Approach 1:
The patent segments the image reconstruction task into independent units that can be assigned to individual GPUs. Each GPU processes specific image slices or regions independently, so that a malfunction in one GPU does not prevent the system from reconstructing images using the remaining functional GPUs. This segmentation maintains the high processing speed advantage of Multi-GPU architecture while improving reliability through fault isolation.
Solution Approach 2:
The patent implements dynamic task reallocation where the system can change the distribution of reconstruction tasks based on the operational status of individual GPUs. When a GPU malfunctions, the controller redistributes tasks among functional GPUs, adjusting the workload parameters to maintain continuous image reconstruction capability. This parameter change approach allows the system to adapt to failures while preserving the speed benefits of parallel processing.
2Power
If multiple GPUs perform image processing interdependently, then processing capacity is enhanced, but difficulty in reconstructing intended cross-sectional image increases when at least one GPU malfunctions
Solution Approach 1:
The patent implements a feedback mechanism where the controller continuously monitors the operational status of each GPU and adjusts task assignments accordingly. When a GPU malfunctions, the system detects this through status monitoring and automatically redistributes tasks to maintain the intended reconstruction capability. This feedback loop resolves the interdependence problem by making the system adaptive to individual GPU failures.
Solution Approach 2:
The patent makes the task assignment dynamic rather than static. The system can reconfigure which GPUs process which image data based on real-time operational status. This dynamic approach allows the system to maintain high processing capacity when all GPUs are functional while automatically degrading to reduced-capacity operation (using only functional GPUs) when failures occur, thus maintaining the intended reconstruction capability.
Data Source
AI summary
An apparatus for processing a medical image includes an image processor including a plurality of processors, the plurality of processors configured to reconstruct a cross-sectional image of an object by performing a first operation having a first priority and a second operation having a second priority that is lower than the first priority, and a controller configured to monitor whether a malfunction occurs among the plurality of processors, and configured to assign, to at least one of the plurality of processors, at least one of the first operation and the second operation to be performed, based on a result of monitoring of the plurality of processors.