Parallel Graphics Processing for Sequencing Colony Signal Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional sequencing techniques are inaccurate and computationally expensive due to failure to account for signal interference and crosstalk, and inefficiently process high volumes of image data generated during flow sequencing, leading to inefficient performance.
Innovation Solution
The method involves using a graphics processor to execute iterative processes for detecting sequencing colonies, calculating crosstalk values, and refining amplitude estimates, with parallel processing and advanced image filtering techniques to improve signal accuracy and processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques are used to detect signal intensities, then the processing approach is simple, but the measurement precision is poor due to failure to account for signal interference and crosstalk
Solution Approach 1:
The patent segments the image processing into multiple iterative steps: detecting colonies, estimating signal amplitudes, calculating crosstalk values for neighboring objects, and refining amplitude estimates. This segmentation allows systematic handling of signal interference by processing one object at a time while accounting for neighbors, thereby improving measurement precision without overwhelming complexity.
Solution Approach 2:
The patent implements an iterative feedback process where crosstalk values calculated from neighboring objects are fed back to refine the amplitude estimates of each object. This feedback loop continues until convergence, systematically eliminating signal interference effects and improving measurement accuracy while maintaining manageable computational complexity through controlled iteration.
2Productivity
If generic computer processors are used for image processing, then the device complexity is low, but the productivity is insufficient to process high volumes of images at high rates
Solution Approach 1:
The patent segments the image processing tasks into independent units that can be processed in parallel. Each colony detection and amplitude estimation can be performed simultaneously on different colonies within an image, allowing the system to process high volumes of images at high rates by distributing computational workload across multiple processing units.
Solution Approach 2:
The patent transitions from sequential processing to parallel processing by utilizing graphics processors capable of concurrent operations. This dimensional change in processing architecture enables simultaneous handling of multiple images and colonies, dramatically increasing productivity from thousands to hundreds of thousands of images per second while managing system complexity through specialized hardware utilization.
3Productivity
If linear or serial processing is used, then the processing approach is simple, but the productivity is low and computer processing power is inefficiently used
Solution Approach 1:
The patent divides the image processing into independent parallel tasks that can be executed simultaneously. Each colony's signal amplitude calculation is an independent segment that can be processed concurrently with others, enabling high throughput by distributing work across multiple processing threads or cores, thereby increasing productivity without proportionally increasing complexity.
Solution Approach 2:
The patent performs preliminary actions by pre-calculating crosstalk values for neighboring objects before finalizing amplitude estimates. This preliminary computation enables subsequent rapid refinement of signal intensities without repeating complex calculations, thereby increasing processing throughput and productivity while managing overall computational complexity through staged processing.
Data Source
AI summary
The present disclosure relates generally to sequencing techniques, and more specifically to methods, systems, devices, and non-transitory computer-readable storage media for processing images of biological samples (e.g., to obtain sequencing data). An exemplary method of determining nucleic acid sequences of a plurality of sequencing colonies comprises: obtaining an input image of a surface, wherein the plurality of sequencing colonies are attached to the surface; detecting a set of sequencing colonies of the plurality of sequencing colonies in the input image; executing in parallel, using a graphics processor, a plurality of iterative processes to obtain signal amplitudes for the detected set of sequencing colonies, wherein each iterative process corresponds to a respective detected sequencing colony in the set; and determining, at least partially based on the signal amplitudes for the detected set of sequencing colonies, portions of nucleic acid sequences of the plurality of sequencing colonies.