Adaptive Parallel Processing for Flow Cytometry Data Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In parallel computing architectures, existing methods fail to effectively partition tasks between processors, leading to underutilization of resources as some processors remain idle while others handle more work, particularly in data analysis applications where efficient use of resources is crucial.

Innovation Solution

An adaptive semi-synchronous parallel processing system that identifies memory and relationship dependencies between tasks, representing them as a directed graph to optimize task assignment and execution, using threading models like parallel, strip-mine, and do not thread to maximize processor utilization and reduce processing time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If tasks are executed in parallel without intelligent partitioning, then computational speed may increase, but processor utilization becomes unbalanced with some processors idle while others are overloaded

Engineering Contradiction:
Improvecomputational speedVSAvoidprocessor utilization
Core Design Contradiction:
SpeedVSProductivity

Solution Approach 1:

The patent segments tasks into discrete units that can be independently assigned to different processors. By breaking down the computational workload into manageable task segments, the system can intelligently distribute these segments across multiple processors based on current processor availability and task dependencies, thereby balancing the load and improving overall processor utilization while maintaining parallel execution speed.

Inventive Principle:
Principle #1Segmentation

2Power

If more processors are added to the system, then computational capacity increases, but task management complexity increases

Engineering Contradiction:
Improvecomputational capacityVSAvoidtask management complexity
Core Design Contradiction:
PowerVSDevice complexity

Solution Approach 1:

The patent implements a self-service task management system where processors automatically monitor their own workload and availability, and the system automatically assigns tasks based on predefined criteria and real-time processor status. This automated self-management approach allows the system to scale to more processors without proportionally increasing management complexity, as the system handles task distribution dynamically without requiring manual intervention for each processor addition.

Inventive Principle:
Principle #25Self-service

3Ease of operation

If tasks are assigned based on simple load balancing, then implementation is straightforward, but processing time increases due to idle processors

Engineering Contradiction:
Improveimplementation simplicityVSAvoidprocessing time
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent incorporates feedback mechanisms where the system continuously monitors processor performance, task completion status, and inter-task dependencies. This feedback information is used to dynamically adjust task assignments, ensuring that tasks are routed to processors that can execute them most efficiently. The feedback-driven approach maintains relative implementation simplicity while significantly reducing processing time by preventing idle processor periods and optimizing task distribution based on real-time system state.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2176754B1System and method for optimizing data analysis
Publication Date: 2019.10.16 SOFTLIFE PROJECTS LIMITED DOING BUSINESS AS APPLI ED CYTOMETRY SYST
  • EP2176754B1 patent drawingFigure 1
  • EP2176754B1 patent drawingFigure 2
  • EP2176754B1 patent drawingFigure 2A

AI summary

There is provided an adaptive semi- synchronous parallel processing system and method, which may be adapted to various data analysis applications such as flow cytometry systems. By identifying the relationship and memory dependencies between tasks that are necessary to complete an analysis, it is possible to significantly reduce the analysis processing time by selectively executing tasks after careful assignment of tasks to one or more processor queues, where the queue assignment is based on an optimal execution strategy. Further strategies are disclosed to address optimal processing once a task undergoes computation by a computational element in a multiprocessor system. Also disclosed is a technique to perform fluorescence compensation to correct spectral overlap between different detectors in a flow cytometry system due to emission characteristics of various fluorescent dyes.