Lens-Free Particle Counting via Trajectory Tracking

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Counting particles in a fluid, such as cells in a biological sample, is challenging due to the need to ensure each particle is counted only once despite appearing in multiple image frames during video capture.

Innovation Solution

A method and system using a lens-free imager to capture video data, where particle centers are linked across consecutive frames using a statistical model of particle displacements, and trajectories are tracked to calculate particle counts, either through linear regression or transformation into a common coordinate frame for clustering.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If video data is captured to track particles through multiple image frames, then particle identification and trajectory tracking are improved, but particle counting accuracy deteriorates due to multiple counting of the same particle

Engineering Contradiction:
Improveparticle identification reliabilityVSAvoidparticle counting precision
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The system performs preliminary actions by capturing video data and detecting particle positions in multiple frames before counting. It establishes particle trajectories and determines effective observation volumes in advance, then uses this pre-established information to accurately count particles, preventing multiple counting of the same particle.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary concept of 'effective observation volume' and particle trajectories as mediators between video capture and particle counting. By defining specific observation volumes and tracking particles through these volumes, the system accurately determines when a particle enters or exits the observation region, enabling precise counting without multiple counts.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Device complexity

If conventional counting methods are used on video frames, then processing is simpler, but counting accuracy deteriorates due to inability to distinguish unique particles

Engineering Contradiction:
Improveprocessing complexityVSAvoidparticle counting precision
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system transitions from two-dimensional frame-by-frame counting to three-dimensional trajectory-based counting by adding the time dimension. It tracks particle positions across multiple frames to establish trajectories, then counts particles based on their movement through defined observation volumes in space-time, fundamentally changing the counting approach from simple 2D detection to 3D trajectory analysis.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Reliability

If particle tracking through multiple frames is implemented, then particle identification is improved, but processing time and computational complexity increase

Engineering Contradiction:
Improveparticle tracking reliabilityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system segments the processing task into distinct phases: video capture and particle detection in each frame, trajectory construction by linking detections across frames, effective observation volume determination, and final counting based on trajectory analysis. This segmentation allows efficient processing by handling each phase separately with optimized algorithms for that specific task.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3635374B1Systems and methods for counting particles
Publication Date: 2024.12.25 MIDIAGNOSTICS NV
  • EP3635374B1 patent drawingFigure 1~2
  • EP3635374B1 patent drawingFigure 3A~3B
  • EP3635374B1 patent drawingFigure 4~5

AI summary

Systems and methods are provided for counting particles in a fluid flow. In an aspect, coordinates of particles are obtained from video data of particles in a fluid, the video data made up of a sequence of image frames. The particle positions are linked in each pair of consecutive image frames of the video data. The linked particle positions are used to calculate particle trajectories through sequential image frames of the video data, and the particles are counted based on the particle trajectory. In another aspect, the particle positons within each image frame are transformed to estimated positions within a common coordinate frame. The estimated particle positions of a particle are grouped into a cluster center, and the particle count is calculated based on the cluster centers.