Web Microanalysis of Big Image Data

Table of Contents

1 Introduction 1

1.1 What is image processing pipeline? 1

1.2 What does web image processing pipeline consist of? 3

1.3 What are big data microscopy experiments? 4

1.4 Why are scientists interested in big data microscopy experiments? 7

1.5 What is the range of applications leveraging image processing pipelines? 9

1.6 Challenges of big data microscopy experiments 10

1.7 Tradeoffs Before and After Digital Images Are Acquired 12

1.8 Enabling reproducible science from big data microscopy experiments 14



2 Using a Web Image Processing Pipeline for Big Cell Microscopy Experiments 11

2.1 Deploying and testing Open-Source Web Image Processing Pipeline 11

2.1.1 Deployment types and installation 11

2.1.2 Prerequisites and configuration options 12

2.1.3 Test data 12 2.2 Web processing of image tiles to viewable and quantifiable images 12

2.3 Web Feature Extraction 13

2.4 Web Statistical Modeling 13



3 Building Web image processing pipeline 14

3.1 Architecture of a web image processing pipeline 14

3.2 Storage, representation and data management of images and metadata 14

3.3 Web Servers to interface database and web browser (client) applications, and manage computational jobs 15

3.4 Integration of heterogeneous software using a common file interface 16

3.5 Interactive measurements in a browser (client) 16

3.6 Provenance information and traceability of computations 16

3.6.1 Sources of Provenance information 16

3.6.2 Persistency of big image collections 16

3.6.3 Traceability of computations 16

3.7 Summary 17



4 Algorithms in image processing pipelines 17

4.1 General image processing resources 18

4.2 Specific algorithms in image processing pipeline 18

4.2.1 Image correction algorithms 18

4.2.2 From image tiles to web-viewable images 18

4.2.3 Object segmentation and tracking 19

4.2.4 Image object measurements 19

4.3 Summary 19



5 Accelerating server-side image processing algorithms 21

5.1 Acceleration of image processing algorithms by using cluster computing 21

5.2 Acceleration of stitching by using GPU computing 21

5.3 Summary 21



6 Example Use Cases 21

6.1 Characterization of stem cell colony growth 21

6.2 Variability of image features across software packages 21

6.3 Statistical modeling of stem cell colony features 22

6.4 Summary 22



7 Supplementary information 22

7.1 Software 22

7.2 Test data for software installation 22

7.3 Test deployment of software (without uploading capabilities) 22



8 Terminology Index 22



9 References 23