SlideShare a Scribd company logo
1 of 25
Multi-scale directional
filtering based
method for Follicular
Lymphoma grading
ALİCAN BOZKURT, A. ENIS CETIN
MUSCLE WORKSHOP, ANTALYA
03.10.2013
Follicular Lymphoma
grading
• Follicular Lymphoma (FL)
•

Presence of a follicular or
nodular pattern of growth
presented by follicle center B
cells
• centrocytes and
centroblasts.

Grade 1 (0-5)
Grade 2 (6-15)

Grade 3 (>15)

2
Follicular Lymphoma
grading

Grade 1

Grade 2

Grade 3

3
Follicular Lymphoma
grading
• Pioneer work by Sertel et al:
• mimicked the manual approach of pathologists, i.e., identifying the number
of centroblasts in the sample. Based on this, a decision on the grade of the
sample can be made.
• Accuracy for CB detection was about 80%.

Sertel, Olcay, et al. "Histopathological image analysis using model-based intermediate representations and color texture:
Follicular lymphoma grading." Journal of Signal Processing Systems 55.1-3 (2009): 169-183.

4
Follicular Lymphoma
grading
• Improvement by Suhre
• Hp and Ep denote the projections on the H and E vectors proposed
by Cosatto et al. (2008) to model Hematoxylin and Eosin (H&E)
staining.
• Grades (1,2) and 3 can be distinguished by comparing the
histograms via Kullback-Leibler (KL) divergence.
• For differentiating grades 1 and 2, we choose a Bayesian classifier.
(DCT of the eigenvalue histograms) The underlying PDF is assumed
to be sparse, therefore only q coefficients are used.

Grade 1

Grade 2

Grade 3

98.89

98.89

100

5
Follicular Lymphoma
grading
• Our Work
•
•
•
•

Approaches the problem as texture recognition program
Based on a novel multi-scale feature extraction method
LDA
SVM

6
Directional filtering
•Main idea: rotating a 1D filter along desired orientation
•Easy for θ=k x 45°, k=0,1,2,…
•Not easy for θ≠k x 45°
• Bilinear/cubic interpolation
• Our method: coefficients proportional to length of line segments enclosed
by pixels
• Also used in CT

Herman, Gabor T. "Image reconstruction from projections." Image Reconstruction from Projections: Implementation and Applications 1 (1979).

7
8
Directional filtering

9
Directional Filtering

10
Directional filtering

11
12
Directional Filtering

13
Feature extraction
Step 0

• Input Image

14
Feature extraction
• Input Image
Step 0

Step 1

• Convert Image
to gray level

15
Feature extraction
Step 0

μ1 :
σ1 :

0,082091

0,084891

0,060045

0,080689

0,085836

0,060873

0,14791

0,15201

0,11201

0,14617

0,15402

0,11424

50

50

50

50

100

100

100

100

100

150

150

150

150

150

150

200

200

200

200

200

200

250

250

250

250

250

250

300

300

50

100

300

μ2:
σ2:

300
50

100

150

200

250

300

350

400

450

500

300

300

50
50

100

150

200

250

300

350

400

450

500

50

100

150

200

250

300

350

400

450

100

150

200

250

300

350

400

450

500

500

50

50

100

150

200

250

300

350

400

450

500

50

100

150

200

250

300

350

400

450

0,22597

0,24064

0,11976

0,23731

0,24072

0,36203

0,35692

0,17401

0,37765

0,34842

500

0,12753

Step 2

• Convert Image
to gray level

• Extract
Features

0,19024

20

20

40

40

60

60

60

60

60

60

40

40

40

40

20

20

20

20

80

80

80

80
100

100

100

100

100
120

120

120

120

80

80
1 00
1 20

140

1 60
50

10 0

150

200

250

140

160
50

100

150

200

250

100

150

2 00

140

160
50

50

120

140

160

140

160

1 40

μ3:
σ3:

Step 1

• Input Image

100

150

200

250

250

160

50

100

150

200

25 0

50

100

150

200

250

0,49943 0,54883

0,35954

0,55623

0,56736

0,30949

0,6949

0,46078

0,72141

0,68851

0,39779

0,65361

Φ = [μ1 σ1 μ2 σ2 μ3 σ3] (1x36 feature vector)
16
Classification
[
[

[

θ1
θ2
.
.
.
θN
features

]
]

PCA

Test

10fold
CV
]

LDA

Training

SVM
Train
Par
am
ete
r
sea
rch
for
C
an
dγ

Model

SVM
Classify

Mean
Accuracy

17
Dataset
 Same dataset used by Suhre
 90 images per grade

Grade 1

Grade 2

Grade 3
18
Background

19
Results
Follicular Lymphoma
•Max: 100.00 (Dir. Fil.)
•SoA: 99.26

[20] A. Suhre, Novel Methods for Microscopic Image Processing,
Analysis, Classification and Compression. PhD thesis, Bilkent
University, 2013.

20
Results
6
Grade 1
Grade 2
Grade 3

4

Feature 2

2

0

-2

-4

-6
-10

-8

-6

-4

-2
Feature 1

0

2

4

6

21
Background

22
Results

23
Conclusion
•New directional filter construction and multiscale filtering framework
• Computationally efficient (2x faster than the closest competitor)

•Follicular Lymphoma Grading as an application of the framework
• Mean and standard deviation of directional filter outputs as features
• LDA as feature reduction (to 2D)
• SVM as classifier
• Outperformed state of art

24
Thank You!

25

More Related Content

What's hot

SA of Genome_YanzheYin
SA of Genome_YanzheYinSA of Genome_YanzheYin
SA of Genome_YanzheYin
Yanzhe Yin
 
Segmentation and Classification of Skin Lesions Based on Texture Features
Segmentation and Classification of Skin Lesions Based on Texture FeaturesSegmentation and Classification of Skin Lesions Based on Texture Features
Segmentation and Classification of Skin Lesions Based on Texture Features
IJERA Editor
 
Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022
Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022
Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022
sugiuralab
 
Slice profile ieee2011_siu
Slice profile ieee2011_siuSlice profile ieee2011_siu
Slice profile ieee2011_siu
linlinc
 

What's hot (20)

Ijciet 10 02_012
Ijciet 10 02_012Ijciet 10 02_012
Ijciet 10 02_012
 
A360108
A360108A360108
A360108
 
A04450104
A04450104A04450104
A04450104
 
Lung Cancer Detection using Image Processing Techniques
Lung Cancer Detection using Image Processing TechniquesLung Cancer Detection using Image Processing Techniques
Lung Cancer Detection using Image Processing Techniques
 
Image processing in lung cancer screening and treatment
Image processing in lung cancer screening and treatmentImage processing in lung cancer screening and treatment
Image processing in lung cancer screening and treatment
 
Human Skin Cancer Recognition and Classification by Unified Skin Texture and ...
Human Skin Cancer Recognition and Classification by Unified Skin Texture and ...Human Skin Cancer Recognition and Classification by Unified Skin Texture and ...
Human Skin Cancer Recognition and Classification by Unified Skin Texture and ...
 
Modeling Cardiac Pacemakers With Timed Coloured Petri Nets And Related Tools
Modeling Cardiac Pacemakers With Timed Coloured Petri Nets And Related ToolsModeling Cardiac Pacemakers With Timed Coloured Petri Nets And Related Tools
Modeling Cardiac Pacemakers With Timed Coloured Petri Nets And Related Tools
 
IRJET- Analysis of Skin Cancer using ABCD Technique
IRJET-  	  Analysis of Skin Cancer using ABCD TechniqueIRJET-  	  Analysis of Skin Cancer using ABCD Technique
IRJET- Analysis of Skin Cancer using ABCD Technique
 
DETECTION OF LESION USING SVM
DETECTION OF LESION USING SVMDETECTION OF LESION USING SVM
DETECTION OF LESION USING SVM
 
SA of Genome_YanzheYin
SA of Genome_YanzheYinSA of Genome_YanzheYin
SA of Genome_YanzheYin
 
Image Classification And Skin cancer detection
Image Classification And Skin cancer detectionImage Classification And Skin cancer detection
Image Classification And Skin cancer detection
 
Segmentation and Classification of Skin Lesions Based on Texture Features
Segmentation and Classification of Skin Lesions Based on Texture FeaturesSegmentation and Classification of Skin Lesions Based on Texture Features
Segmentation and Classification of Skin Lesions Based on Texture Features
 
97202107
9720210797202107
97202107
 
Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022
Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022
Cervical Spine Range of Motion Measurement Utilizing Image Analysis - VISAPP2022
 
IRJET- Color and Texture based Feature Extraction for Classifying Skin Ca...
IRJET-  	  Color and Texture based Feature Extraction for Classifying Skin Ca...IRJET-  	  Color and Texture based Feature Extraction for Classifying Skin Ca...
IRJET- Color and Texture based Feature Extraction for Classifying Skin Ca...
 
Lung Nodule detection System
Lung Nodule detection SystemLung Nodule detection System
Lung Nodule detection System
 
Slice profile ieee2011_siu
Slice profile ieee2011_siuSlice profile ieee2011_siu
Slice profile ieee2011_siu
 
Computer-aided diagnosis system for breast cancer based on the Gabor filter ...
Computer-aided diagnosis system for breast cancer  based on the Gabor filter ...Computer-aided diagnosis system for breast cancer  based on the Gabor filter ...
Computer-aided diagnosis system for breast cancer based on the Gabor filter ...
 
Spie, feb 2016, san diego poster
Spie, feb 2016, san diego   posterSpie, feb 2016, san diego   poster
Spie, feb 2016, san diego poster
 
Detection of Malignancy in Digital Mammograms from Segmented Breast Region Us...
Detection of Malignancy in Digital Mammograms from Segmented Breast Region Us...Detection of Malignancy in Digital Mammograms from Segmented Breast Region Us...
Detection of Malignancy in Digital Mammograms from Segmented Breast Region Us...
 

Viewers also liked

Giraffe Anesthesia Presentation-Emily Hall-2
Giraffe Anesthesia Presentation-Emily Hall-2Giraffe Anesthesia Presentation-Emily Hall-2
Giraffe Anesthesia Presentation-Emily Hall-2
Hall Emily
 
Day case anesthesia
 Day case anesthesia Day case anesthesia
Day case anesthesia
Omar Danfour
 
Molecular pathology of lymphoma by dr ramesh
Molecular pathology of  lymphoma by dr ramesh Molecular pathology of  lymphoma by dr ramesh
Molecular pathology of lymphoma by dr ramesh
Ramesh Purohit
 
Thrombotic Thrombocytopenic Purpura
Thrombotic Thrombocytopenic PurpuraThrombotic Thrombocytopenic Purpura
Thrombotic Thrombocytopenic Purpura
Shakeel Arif
 

Viewers also liked (14)

smoldering myeloma
smoldering myelomasmoldering myeloma
smoldering myeloma
 
Case Conference on the 26th Generalist Training Seminar
Case Conference on the 26th Generalist Training SeminarCase Conference on the 26th Generalist Training Seminar
Case Conference on the 26th Generalist Training Seminar
 
Giraffe Anesthesia Presentation-Emily Hall-2
Giraffe Anesthesia Presentation-Emily Hall-2Giraffe Anesthesia Presentation-Emily Hall-2
Giraffe Anesthesia Presentation-Emily Hall-2
 
Roschewski-Mark-IV-hematology_forum_2016
Roschewski-Mark-IV-hematology_forum_2016Roschewski-Mark-IV-hematology_forum_2016
Roschewski-Mark-IV-hematology_forum_2016
 
Indolent non hodgkins lymphoma
Indolent non hodgkins lymphomaIndolent non hodgkins lymphoma
Indolent non hodgkins lymphoma
 
TTP HUSについて
TTP HUSについてTTP HUSについて
TTP HUSについて
 
Lymphomas 5
Lymphomas 5Lymphomas 5
Lymphomas 5
 
Day case anesthesia
 Day case anesthesia Day case anesthesia
Day case anesthesia
 
Molecular pathology of lymphoma by dr ramesh
Molecular pathology of  lymphoma by dr ramesh Molecular pathology of  lymphoma by dr ramesh
Molecular pathology of lymphoma by dr ramesh
 
Thrombotic Thrombocytopenic Purpura
Thrombotic Thrombocytopenic PurpuraThrombotic Thrombocytopenic Purpura
Thrombotic Thrombocytopenic Purpura
 
Thrombotic Thrombocytopenic Purpura / Hemolytic Uremic Syndrome (Questions & ...
Thrombotic Thrombocytopenic Purpura / Hemolytic Uremic Syndrome (Questions & ...Thrombotic Thrombocytopenic Purpura / Hemolytic Uremic Syndrome (Questions & ...
Thrombotic Thrombocytopenic Purpura / Hemolytic Uremic Syndrome (Questions & ...
 
Lymphoma
LymphomaLymphoma
Lymphoma
 
Non hodgkin lymphoma
Non hodgkin lymphomaNon hodgkin lymphoma
Non hodgkin lymphoma
 
Ttp in immediate postpartum
Ttp in immediate postpartum Ttp in immediate postpartum
Ttp in immediate postpartum
 

Similar to Multi Scale Directional Filtering Based Method for Follicular Lymphoma Grading

Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...
Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...
Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...
Wookjin Choi
 
Optimal fuzzy rule based pulmonary nodule detection
Optimal fuzzy rule based pulmonary nodule detectionOptimal fuzzy rule based pulmonary nodule detection
Optimal fuzzy rule based pulmonary nodule detection
Wookjin Choi
 
Microcalcification Enhancement in Digital Mammogram
Microcalcification Enhancement in Digital MammogramMicrocalcification Enhancement in Digital Mammogram
Microcalcification Enhancement in Digital Mammogram
Nashid Alam
 

Similar to Multi Scale Directional Filtering Based Method for Follicular Lymphoma Grading (20)

sheeba.pptx
sheeba.pptxsheeba.pptx
sheeba.pptx
 
Detection of exudates draft
Detection of exudates draftDetection of exudates draft
Detection of exudates draft
 
CANCER CLUMPS DETECTION USING IMAGE PROCESSING BASED ON CELL COUNTING
CANCER CLUMPS DETECTION USING IMAGE PROCESSING BASED ON CELL COUNTINGCANCER CLUMPS DETECTION USING IMAGE PROCESSING BASED ON CELL COUNTING
CANCER CLUMPS DETECTION USING IMAGE PROCESSING BASED ON CELL COUNTING
 
Detection of exudates draft
Detection of exudates draftDetection of exudates draft
Detection of exudates draft
 
Melanoma Image Segmentation using Self Organized Features Maps
Melanoma Image Segmentation using Self Organized Features MapsMelanoma Image Segmentation using Self Organized Features Maps
Melanoma Image Segmentation using Self Organized Features Maps
 
Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...
Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...
Novel Functional Delta-Radiomics for Predicting Overall Survival in Lung Canc...
 
LIVE DEMONSTRATION: Understanding the Complimentary Nature of BLI, Ultrasound...
LIVE DEMONSTRATION: Understanding the Complimentary Nature of BLI, Ultrasound...LIVE DEMONSTRATION: Understanding the Complimentary Nature of BLI, Ultrasound...
LIVE DEMONSTRATION: Understanding the Complimentary Nature of BLI, Ultrasound...
 
Introduction to image processing
Introduction to image processingIntroduction to image processing
Introduction to image processing
 
Optimal fuzzy rule based pulmonary nodule detection
Optimal fuzzy rule based pulmonary nodule detectionOptimal fuzzy rule based pulmonary nodule detection
Optimal fuzzy rule based pulmonary nodule detection
 
color doppler
color dopplercolor doppler
color doppler
 
Digital radiography (2013)
Digital radiography (2013)Digital radiography (2013)
Digital radiography (2013)
 
Microcalcification Enhancement in Digital Mammogram
Microcalcification Enhancement in Digital MammogramMicrocalcification Enhancement in Digital Mammogram
Microcalcification Enhancement in Digital Mammogram
 
Identik Problem
Identik ProblemIdentik Problem
Identik Problem
 
sec dc.pptx
sec dc.pptxsec dc.pptx
sec dc.pptx
 
Noninvasive diabetes mellitus detection
Noninvasive diabetes mellitus detectionNoninvasive diabetes mellitus detection
Noninvasive diabetes mellitus detection
 
Radiomics and Deep Learning for Lung Cancer Screening
Radiomics and Deep Learning for Lung Cancer ScreeningRadiomics and Deep Learning for Lung Cancer Screening
Radiomics and Deep Learning for Lung Cancer Screening
 
reliablepaper.pdf
reliablepaper.pdfreliablepaper.pdf
reliablepaper.pdf
 
Contourlet Transform Based Method For Medical Image Denoising
Contourlet Transform Based Method For Medical Image DenoisingContourlet Transform Based Method For Medical Image Denoising
Contourlet Transform Based Method For Medical Image Denoising
 
Detection of Cysts in Ultrasonic Images of Ovary
Detection of Cysts in Ultrasonic Images of OvaryDetection of Cysts in Ultrasonic Images of Ovary
Detection of Cysts in Ultrasonic Images of Ovary
 
Computer aided detection of pulmonary nodules using genetic programming
Computer aided detection of pulmonary nodules using genetic programmingComputer aided detection of pulmonary nodules using genetic programming
Computer aided detection of pulmonary nodules using genetic programming
 

Recently uploaded

Failure to thrive in neonates and infants + pediatric case.pptx
Failure to thrive in neonates and infants  + pediatric case.pptxFailure to thrive in neonates and infants  + pediatric case.pptx
Failure to thrive in neonates and infants + pediatric case.pptx
claviclebrown44
 

Recently uploaded (20)

SEMESTER-V CHILD HEALTH NURSING-UNIT-1-INTRODUCTION.pdf
SEMESTER-V CHILD HEALTH NURSING-UNIT-1-INTRODUCTION.pdfSEMESTER-V CHILD HEALTH NURSING-UNIT-1-INTRODUCTION.pdf
SEMESTER-V CHILD HEALTH NURSING-UNIT-1-INTRODUCTION.pdf
 
5CL-ADB powder supplier 5cl adb 5cladba 5cl raw materials vendor on sale now
5CL-ADB powder supplier 5cl adb 5cladba 5cl raw materials vendor on sale now5CL-ADB powder supplier 5cl adb 5cladba 5cl raw materials vendor on sale now
5CL-ADB powder supplier 5cl adb 5cladba 5cl raw materials vendor on sale now
 
5cladba raw material 5CL-ADB-A precursor raw
5cladba raw material 5CL-ADB-A precursor raw5cladba raw material 5CL-ADB-A precursor raw
5cladba raw material 5CL-ADB-A precursor raw
 
Unveiling Alcohol Withdrawal Syndrome: exploring it's hidden depths
Unveiling Alcohol Withdrawal Syndrome: exploring it's hidden depthsUnveiling Alcohol Withdrawal Syndrome: exploring it's hidden depths
Unveiling Alcohol Withdrawal Syndrome: exploring it's hidden depths
 
Is Rheumatoid Arthritis a Metabolic Disorder.pptx
Is Rheumatoid Arthritis a Metabolic Disorder.pptxIs Rheumatoid Arthritis a Metabolic Disorder.pptx
Is Rheumatoid Arthritis a Metabolic Disorder.pptx
 
CONGENITAL HYPERTROPHIC PYLORIC STENOSIS by Dr M.KARTHIK EMMANUEL
CONGENITAL HYPERTROPHIC PYLORIC STENOSIS  by Dr M.KARTHIK EMMANUELCONGENITAL HYPERTROPHIC PYLORIC STENOSIS  by Dr M.KARTHIK EMMANUEL
CONGENITAL HYPERTROPHIC PYLORIC STENOSIS by Dr M.KARTHIK EMMANUEL
 
Overview on the Automatic pill identifier
Overview on the Automatic pill identifierOverview on the Automatic pill identifier
Overview on the Automatic pill identifier
 
Our Hottest 💘 Surat ℂall Girls Serviℂe 💘Pasodara📱 8527049040📱450+ ℂall Girl C...
Our Hottest 💘 Surat ℂall Girls Serviℂe 💘Pasodara📱 8527049040📱450+ ℂall Girl C...Our Hottest 💘 Surat ℂall Girls Serviℂe 💘Pasodara📱 8527049040📱450+ ℂall Girl C...
Our Hottest 💘 Surat ℂall Girls Serviℂe 💘Pasodara📱 8527049040📱450+ ℂall Girl C...
 
5Cladba ADBB 5cladba buy 6cl adbb powder 5cl ADBB precursor materials
5Cladba ADBB 5cladba buy 6cl adbb powder 5cl ADBB precursor materials5Cladba ADBB 5cladba buy 6cl adbb powder 5cl ADBB precursor materials
5Cladba ADBB 5cladba buy 6cl adbb powder 5cl ADBB precursor materials
 
Renal Replacement Therapy in Acute Kidney Injury -time modality -Dr Ayman Se...
Renal Replacement Therapy in Acute Kidney Injury -time  modality -Dr Ayman Se...Renal Replacement Therapy in Acute Kidney Injury -time  modality -Dr Ayman Se...
Renal Replacement Therapy in Acute Kidney Injury -time modality -Dr Ayman Se...
 
Failure to thrive in neonates and infants + pediatric case.pptx
Failure to thrive in neonates and infants  + pediatric case.pptxFailure to thrive in neonates and infants  + pediatric case.pptx
Failure to thrive in neonates and infants + pediatric case.pptx
 
VVIP Yelahanka ℂall Girls 6350482085 Heat-immolating { Bangalore } Coveted Gi...
VVIP Yelahanka ℂall Girls 6350482085 Heat-immolating { Bangalore } Coveted Gi...VVIP Yelahanka ℂall Girls 6350482085 Heat-immolating { Bangalore } Coveted Gi...
VVIP Yelahanka ℂall Girls 6350482085 Heat-immolating { Bangalore } Coveted Gi...
 
Premium ℂall Girls In Mumbai👉 Dail ℂALL ME: 📞9833325238 📲 ℂall Richa VIP ℂall...
Premium ℂall Girls In Mumbai👉 Dail ℂALL ME: 📞9833325238 📲 ℂall Richa VIP ℂall...Premium ℂall Girls In Mumbai👉 Dail ℂALL ME: 📞9833325238 📲 ℂall Richa VIP ℂall...
Premium ℂall Girls In Mumbai👉 Dail ℂALL ME: 📞9833325238 📲 ℂall Richa VIP ℂall...
 
Premium ℂall Girls In Mira Road👉 Dail ℂALL ME: 📞9004268417 📲 ℂall Richa VIP ℂ...
Premium ℂall Girls In Mira Road👉 Dail ℂALL ME: 📞9004268417 📲 ℂall Richa VIP ℂ...Premium ℂall Girls In Mira Road👉 Dail ℂALL ME: 📞9004268417 📲 ℂall Richa VIP ℂ...
Premium ℂall Girls In Mira Road👉 Dail ℂALL ME: 📞9004268417 📲 ℂall Richa VIP ℂ...
 
Tips to Choose the Best Psychiatrists in Indore
Tips to Choose the Best Psychiatrists in IndoreTips to Choose the Best Psychiatrists in Indore
Tips to Choose the Best Psychiatrists in Indore
 
Denture base resins materials and its mechanism of action
Denture base resins materials and its mechanism of actionDenture base resins materials and its mechanism of action
Denture base resins materials and its mechanism of action
 
Creating Accessible Public Health Communications
Creating Accessible Public Health CommunicationsCreating Accessible Public Health Communications
Creating Accessible Public Health Communications
 
TEST BANK For Huether and McCance's Understanding Pathophysiology, Canadian 2...
TEST BANK For Huether and McCance's Understanding Pathophysiology, Canadian 2...TEST BANK For Huether and McCance's Understanding Pathophysiology, Canadian 2...
TEST BANK For Huether and McCance's Understanding Pathophysiology, Canadian 2...
 
Cervical screening – taking care of your health flipchart (Vietnamese)
Cervical screening – taking care of your health flipchart (Vietnamese)Cervical screening – taking care of your health flipchart (Vietnamese)
Cervical screening – taking care of your health flipchart (Vietnamese)
 
Unlocking Holistic Wellness: Addressing Depression, Mental Well-Being, and St...
Unlocking Holistic Wellness: Addressing Depression, Mental Well-Being, and St...Unlocking Holistic Wellness: Addressing Depression, Mental Well-Being, and St...
Unlocking Holistic Wellness: Addressing Depression, Mental Well-Being, and St...
 

Multi Scale Directional Filtering Based Method for Follicular Lymphoma Grading

  • 1. Multi-scale directional filtering based method for Follicular Lymphoma grading ALİCAN BOZKURT, A. ENIS CETIN MUSCLE WORKSHOP, ANTALYA 03.10.2013
  • 2. Follicular Lymphoma grading • Follicular Lymphoma (FL) • Presence of a follicular or nodular pattern of growth presented by follicle center B cells • centrocytes and centroblasts. Grade 1 (0-5) Grade 2 (6-15) Grade 3 (>15) 2
  • 4. Follicular Lymphoma grading • Pioneer work by Sertel et al: • mimicked the manual approach of pathologists, i.e., identifying the number of centroblasts in the sample. Based on this, a decision on the grade of the sample can be made. • Accuracy for CB detection was about 80%. Sertel, Olcay, et al. "Histopathological image analysis using model-based intermediate representations and color texture: Follicular lymphoma grading." Journal of Signal Processing Systems 55.1-3 (2009): 169-183. 4
  • 5. Follicular Lymphoma grading • Improvement by Suhre • Hp and Ep denote the projections on the H and E vectors proposed by Cosatto et al. (2008) to model Hematoxylin and Eosin (H&E) staining. • Grades (1,2) and 3 can be distinguished by comparing the histograms via Kullback-Leibler (KL) divergence. • For differentiating grades 1 and 2, we choose a Bayesian classifier. (DCT of the eigenvalue histograms) The underlying PDF is assumed to be sparse, therefore only q coefficients are used. Grade 1 Grade 2 Grade 3 98.89 98.89 100 5
  • 6. Follicular Lymphoma grading • Our Work • • • • Approaches the problem as texture recognition program Based on a novel multi-scale feature extraction method LDA SVM 6
  • 7. Directional filtering •Main idea: rotating a 1D filter along desired orientation •Easy for θ=k x 45°, k=0,1,2,… •Not easy for θ≠k x 45° • Bilinear/cubic interpolation • Our method: coefficients proportional to length of line segments enclosed by pixels • Also used in CT Herman, Gabor T. "Image reconstruction from projections." Image Reconstruction from Projections: Implementation and Applications 1 (1979). 7
  • 8. 8
  • 12. 12
  • 15. Feature extraction • Input Image Step 0 Step 1 • Convert Image to gray level 15
  • 16. Feature extraction Step 0 μ1 : σ1 : 0,082091 0,084891 0,060045 0,080689 0,085836 0,060873 0,14791 0,15201 0,11201 0,14617 0,15402 0,11424 50 50 50 50 100 100 100 100 100 150 150 150 150 150 150 200 200 200 200 200 200 250 250 250 250 250 250 300 300 50 100 300 μ2: σ2: 300 50 100 150 200 250 300 350 400 450 500 300 300 50 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 100 150 200 250 300 350 400 450 500 500 50 50 100 150 200 250 300 350 400 450 500 50 100 150 200 250 300 350 400 450 0,22597 0,24064 0,11976 0,23731 0,24072 0,36203 0,35692 0,17401 0,37765 0,34842 500 0,12753 Step 2 • Convert Image to gray level • Extract Features 0,19024 20 20 40 40 60 60 60 60 60 60 40 40 40 40 20 20 20 20 80 80 80 80 100 100 100 100 100 120 120 120 120 80 80 1 00 1 20 140 1 60 50 10 0 150 200 250 140 160 50 100 150 200 250 100 150 2 00 140 160 50 50 120 140 160 140 160 1 40 μ3: σ3: Step 1 • Input Image 100 150 200 250 250 160 50 100 150 200 25 0 50 100 150 200 250 0,49943 0,54883 0,35954 0,55623 0,56736 0,30949 0,6949 0,46078 0,72141 0,68851 0,39779 0,65361 Φ = [μ1 σ1 μ2 σ2 μ3 σ3] (1x36 feature vector) 16
  • 18. Dataset  Same dataset used by Suhre  90 images per grade Grade 1 Grade 2 Grade 3 18
  • 20. Results Follicular Lymphoma •Max: 100.00 (Dir. Fil.) •SoA: 99.26 [20] A. Suhre, Novel Methods for Microscopic Image Processing, Analysis, Classification and Compression. PhD thesis, Bilkent University, 2013. 20
  • 21. Results 6 Grade 1 Grade 2 Grade 3 4 Feature 2 2 0 -2 -4 -6 -10 -8 -6 -4 -2 Feature 1 0 2 4 6 21
  • 24. Conclusion •New directional filter construction and multiscale filtering framework • Computationally efficient (2x faster than the closest competitor) •Follicular Lymphoma Grading as an application of the framework • Mean and standard deviation of directional filter outputs as features • LDA as feature reduction (to 2D) • SVM as classifier • Outperformed state of art 24

Editor's Notes

  1. {}