The growth of the amount of medical image data produced on a daily basis in modern hospitals forces the adaptation of traditional medical image analysis and indexing approaches towards scalable solutions. In this work, MapReduce is used to speed up and make possible three large–scale medical image processing use–cases: (i) parameter optimization for lung texture classification using support vector machines (SVM), (ii) content–based medical image indexing, and (iii) three–dimensional directional wavelet analysis for solid texture classification.

1. Using MapReduce for
Large-scale Medical Image
Analysis
HISB 2012
Presented by : Roger Schaer - HES-SO Valais

2. Summary
Introduction
Methods
Results & Interpretation
Conclusions
2

3. Introduction

4. Introduction
Exponential growth of imaging data (past 20 years)
Year
Amountofimagesproduced
perdayattheHUG
4

5. Introduction (continued)
Mainly caused by :
Modern imaging techniques (3D, 4D) : Large files !
Large collections (available on the Internet)
Increasingly complex algorithms make processing
this data more challenging
Requires a lot of computation power, storage and
network bandwidth
5

6. Introduction (continued)
Flexible and scalable infrastructures are needed
Several approaches exist :
Single, powerful machine
Local cluster / grid
Alternative infrastructures (Graphics cards)
Cloud computing solutions
First two approaches have been tested and compared
6

7. Introduction (continued)
3 large-scale medical image processing use cases
Parameter optimization for Support Vector Machines
Content-based image feature extraction & indexing
3D texture feature extraction using the Riesz
transform
NOTE : I mostly handled the infrastructure
aspects !
7

8. Methods

9. Methods
MapReduce
Hadoop Cluster
Support Vector Machines
Image Indexing
Solid 3D Texture Analysis Using the Riesz Transform
9

10. MapReduce
MapReduce is a programming model
Developed by Google
Map Phase : Key/Value pair input, Intermediate
output
Reduce phase : For each intermediate key, process
the list of associated values
Trivial example : Word Count application
10

11. MapReduce : WordCount
11

12. MapReduce : WordCount
INPUT
11

13. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
INPUT
11

14. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
INPUT MAP
11

15. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
INPUT MAP
11

16. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
INPUT MAP
11

17. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
INPUT MAP
11

18. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
INPUT MAP
11

19. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
INPUT MAP
11

20. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
INPUT MAP
11

21. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
INPUT MAP
11

22. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
INPUT MAP
11

23. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
INPUT MAP
11

24. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
INPUT MAP
11

25. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
INPUT MAP
11

26. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP
11

27. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
11

28. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
11

29. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
11

30. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
11

31. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
11

32. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
11

33. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
goodbye 1
11

34. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
goodbye 1
11

35. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
goodbye 1
hadoop 2
11

36. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
goodbye 1
hadoop 2
11

37. MapReduce : WordCount
#1 hello world
#2 goodbye world
#3 hello hadoop
#4 bye hadoop
...
hello 1
world 1
goodbye 1
world 1
hello 1
hadoop 1
bye 1
hadoop 1
INPUT MAP REDUCE
hello 2
world 2
goodbye 1
hadoop 2
bye 1
11

38. Hadoop
Apache’s implementation of MapReduce
Consists of
Distributed storage system : HDFS
Execution framework : Hadoop MapReduce
Master node which orchestrates the task distribution
Worker nodes which perform the tasks
Typical node runs a DataNode and TaskTracker
12

39. Support Vector Machines
Computes a decision boundary (hyperplane) that
separates inputs of different classes represented in a
given feature space transformed by a given kernel
The values of two parameters need to be adapted to
the data:
Cost C of errors
σ of the Gaussian kernel
13

40. Support Vector Machines
Computes a decision boundary (hyperplane) that
separates inputs of different classes represented in a
given feature space transformed by a given kernel
The values of two parameters need to be adapted to
the data:
Cost C of errors
σ of the Gaussian kernel
0
5
10
15
20
0 5 10 15 20 13

41. Support Vector Machines
Computes a decision boundary (hyperplane) that
separates inputs of different classes represented in a
given feature space transformed by a given kernel
The values of two parameters need to be adapted to
the data:
Cost C of errors
σ of the Gaussian kernel
0
5
10
15
20
0 5 10 15 20 13

42. Support Vector Machines
Computes a decision boundary (hyperplane) that
separates inputs of different classes represented in a
given feature space transformed by a given kernel
The values of two parameters need to be adapted to
the data:
Cost C of errors
σ of the Gaussian kernel
0
5
10
15
20
0 5 10 15 20 13

43. Support Vector Machines
Computes a decision boundary (hyperplane) that
separates inputs of different classes represented in a
given feature space transformed by a given kernel
The values of two parameters need to be adapted to
the data:
Cost C of errors
σ of the Gaussian kernel
0
5
10
15
20
0 5 10 15 20
?
13

44. Support Vector Machines
Computes a decision boundary (hyperplane) that
separates inputs of different classes represented in a
given feature space transformed by a given kernel
The values of two parameters need to be adapted to
the data:
Cost C of errors
σ of the Gaussian kernel
0
5
10
15
20
0 5 10 15 20 13

45. SVM (continued)
Goal : find optimal value couple (C, σ) to train a SVM
Allowing best classification performance of 5 lung
texture patterns
Execution on 1 PC (without Hadoop) can take weeks
Due to extensive leave-one-patient-out cross-
validation with 86 patients
Parallelization : Split job by parameter value couples 14

46. Image Indexing
Vocabulary
File
Image Files
Feature
Extractor
Feature Vectors
Files
Bag of Visual
Words Factory
Index File
Two phases
Extract features from
images
Construct bags of
visual words by
quantization
Component-based /
Monolithic approaches
Parallelization : Each task
processes N images 15

47. Image Indexing
Vocabulary
File
Image Files
Feature
Extractor
Feature Vectors
Files
Bag of Visual
Words Factory
Index File
Two phases
Extract features from
images
Construct bags of
visual words by
quantization
Component-based /
Monolithic approaches
Parallelization : Each task
processes N images 15
Feature
Extractor
+
Bag of
Visual
Words
Factory

48. 3D Texture Analysis (Riesz)
Features are extracted from 3D images (see below)
Parallelization : Each task processes N images
16

49. Results & Interpretation

50. Hadoop Cluster
Minimally invasive setup (>=2 free cores per node)
18

51. Support Vector Machines
Optimization : Longer tasks = bad performance
Because the optimization of the hyperplane is more
difficult to compute (more iterations needed)
After 2 patients (out of 86), check if : ti ≥ F · tref.
If time exceeds average (+margin), terminate task
19

52. Support Vector Machines
Black : tasks to be interrupted by the new algorithm
Optimized algorithm : ~50h → ~9h15min
All the best tasks (highest accuracy) are not killed 20
σ (Sigma)
C
(Cost)
Accuracy(%)

53. Image Indexing
1K IMAGES
Shows the calculation time in function of the # of tasks
Both experiments were executed using hadoop
Once on a single computer, then on our cluster of PCs 21

54. Image Indexing
1K IMAGES 10K IMAGES
Shows the calculation time in function of the # of tasks
Both experiments were executed using hadoop
Once on a single computer, then on our cluster of PCs 21

55. Image Indexing
1K IMAGES 10K IMAGES 100K IMAGES
Shows the calculation time in function of the # of tasks
Both experiments were executed using hadoop
Once on a single computer, then on our cluster of PCs 21

56. Riesz 3D
Particularity : code was a series of Matlab® scripts
Instead of rewriting the whole application :
Used Hadoop streaming feature (uses stdin/stdout)
To maximize scalability, GNU Octave was used
Great compatibility between Matlab® and Octave
22

57. Riesz 3D
Particularity : code was a series of Matlab® scripts
Instead of rewriting the whole application :
Used Hadoop streaming feature (uses stdin/stdout)
To maximize scalability, GNU Octave was used
Great compatibility between Matlab® and Octave
RESULTS
1 task (no Hadoop) 42 tasks (idle) 42 tasks (normal)
131h32m42s 6h29m51s 5h51m31s
22

58. Conclusions

59. Conclusions
MapReduce is
Flexible (worked with very varied use cases)
Easy to use (2-phase programming model is simple)
Efficient (>=20x speedup for all use cases)
Hadoop is
Easy to deploy & manage
User-friendly (nice Web UIs)
24

60. Conclusions (continued)
Speedups for the different use cases
SVMs
Image
Indexing
3D Feature
Extraction
Single task 990h* 21h* 131h30
42 tasks on
hadoop
50h / 9h15** 1h 5h50
Speedup 20x / 107x** 21x 22.5x
* estimation ** using the optimized algorithm 25

61. Lessons Learned
It is important to use physically distributed resources
Overloading a single machine hurts performance
Data locality notably speeds up jobs
Not every application is infinitely scalable
Performance improvements level off at some point
26

62. Future work
Take it to the next level : The Cloud
Amazon Elastic Cloud Compute (IaaS)
Amazon Elastic MapReduce (PaaS)
Cloudbursting
Use both local resources + Cloud (for peak usage)
27

63. Thank you ! Questions ?