New method of staging of Neuropathic foot due to Hansen's disease using foot pressure image signal processing and early detection of foot at risk of plantar.

Slides:



Advertisements
Similar presentations
STC1204 Mid Term Public Speaking Preparation 5 questions Randomly Answer 1 question Date: 19 th April 2010 Monday Time: 4.15 – 6.00pm.
Advertisements

Noise & Data Reduction. Paired Sample t Test Data Transformation - Overview From Covariance Matrix to PCA and Dimension Reduction Fourier Analysis - Spectrum.
Image Enhancement in the Frequency Domain (2)
3-Dimensional Gait Measurement Really expensive and fancy measurement system with lots of cameras and computers Produces graphs of kinematics (joint.
Contingency Tables Chapters Seven, Sixteen, and Eighteen Chapter Seven –Definition of Contingency Tables –Basic Statistics –SPSS program (Crosstabulation)
Covariance and Correlation: Estimator/Sample Statistic: Population Parameter: Covariance and correlation measure linear association between two variables,
LECTURE 3 Introduction to Linear Regression and Correlation Analysis
© 2010 Pearson Prentice Hall. All rights reserved Least Squares Regression Models.
Computer Graphics Recitation 6. 2 Motivation – Image compression What linear combination of 8x8 basis signals produces an 8x8 block in the image?
1 Learning to Detect Objects in Images via a Sparse, Part-Based Representation S. Agarwal, A. Awan and D. Roth IEEE Transactions on Pattern Analysis and.
Lecture 13 The frequency Domain (1)
Chapter Topics Types of Regression Models
Inferences About Process Quality
Digital Image Processing, 2nd ed. © 2002 R. C. Gonzalez & R. E. Woods Chapter 4 Image Enhancement in the Frequency Domain Chapter.
Despeckle Filtering in Medical Ultrasound Imaging
Topic 7 - Fourier Transforms DIGITAL IMAGE PROCESSING Course 3624 Department of Physics and Astronomy Professor Bob Warwick.
ElectroScience Lab IGARSS 2011 Vancouver Jul 26th, 2011 Chun-Sik Chae and Joel T. Johnson ElectroScience Laboratory Department of Electrical and Computer.
@ 2012 Wadsworth, Cengage Learning Chapter 5 Description of Behavior Through Numerical 2012 Wadsworth, Cengage Learning.
Data Presentation.
Motivation Music as a combination of sounds at different frequencies
Adaptive Kernel Density in Demographic Analysis Richard Lycan Institute on Aging Portland State University.
Copyright © 2013, 2010 and 2007 Pearson Education, Inc. Chapter Inference on the Least-Squares Regression Model and Multiple Regression 14.
1 DATA DESCRIPTION. 2 Units l Unit: entity we are studying, subject if human being l Each unit/subject has certain parameters, e.g., a student (subject)
BPS - 3rd Ed. Chapter 211 Inference for Regression.
PREVALENCE OF RISK FACTORS FOR DIABETIC FOOT ULCER AND RISK STRATIFICATION IN TYPE 2 DIABETES DR. NEETA DESHPANDE ASSOCIATE PROF.,JN MEDICAL COLLEGE AND.
Presented by Tienwei Tsai July, 2005
© Copyright McGraw-Hill CHAPTER 3 Data Description.
Transforms. 5*sin (2  4t) Amplitude = 5 Frequency = 4 Hz seconds A sine wave.
Image Processing © 2002 R. C. Gonzalez & R. E. Woods Lecture 4 Image Enhancement in the Frequency Domain Lecture 4 Image Enhancement.
GUIDED BY Mr. Chaitanya Srinivas L.V. Sujeet Blessing Assistant Professor 08MBE026 SBSTVIT University VIT UniversityVellore Vellore 2-D Comparative Gait.
R INSTALLATION R is an open source software package for statistical data analysis.
Yarmouk university Hijjawi faculty for engineering technology Computer engineering department Primary Graduation project Document security using watermarking.
The Effect of Process Variables on Surface Grinding of SUS304 Stainless Steel S. Y. Lin, Professor Department of Mechanical Manufacturing Engineering.
FOUNDATIONS OF NURSING RESEARCH Sixth Edition CHAPTER Copyright ©2012 by Pearson Education, Inc. All rights reserved. Foundations of Nursing Research,
Remote Sensing Supervised Image Classification. Supervised Image Classification ► An image classification procedure that requires interaction with the.
Moulali.P Central Scientific Instruments Organization (CSIO), Council for Scientific and Industrial Research (CSIR), Chandigarh, India.
Digital Image Processing Chapter 4 Image Enhancement in the Frequency Domain Part I.
Lecture 7: Sampling Review of 2D Fourier Theory We view f(x,y) as a linear combination of complex exponentials that represent plane waves. F(u,v) describes.
The Statistical Analysis of Data. Outline I. Types of Data A. Qualitative B. Quantitative C. Independent vs Dependent variables II. Descriptive Statistics.
Spatio-Temporal Free-Form Registration of Cardiac MR Image Sequences Antonios Perperidis s /02/2006.
7- 1 Chapter 7: Fourier Analysis Fourier analysis = Series + Transform ◎ Fourier Series -- A periodic (T) function f(x) can be written as the sum of sines.
Content-Based Image Retrieval Using Block Discrete Cosine Transform Presented by Te-Wei Chiang Department of Information Networking Technology Chihlee.
© Copyright McGraw-Hill 2004
Sample Size Determination
Fourier Transform.
CS 376b Introduction to Computer Vision 03 / 17 / 2008 Instructor: Michael Eckmann.
AP Statistics Section 15 A. The Regression Model When a scatterplot shows a linear relationship between a quantitative explanatory variable x and a quantitative.
BYST Xform-1 DIP - WS2002: Fourier Transform Digital Image Processing Bundit Thipakorn, Ph.D. Computer Engineering Department Fourier Transform and Image.
BPS - 5th Ed. Chapter 231 Inference for Regression.
Baseline and longitudinal patterns of brain atrophy in MCI patients, and their use in prediction of Short- term conversion to AD: Results from ADNI Xuejiao.
Date of download: 6/22/2016 Copyright © 2016 SPIE. All rights reserved. Schematic representation of the near-infrared (NIR) structured illumination instrument,
CONTENTS:  Introduction.  Face recognition task.  Image preprocessing.  Template Extraction and Normalization.  Template Correlation with image database.
Sample Size Determination
MATH-138 Elementary Statistics
Section II Digital Signal Processing ES & BM.
Prof.dr. Taina Avramescu
Image Enhancement in the
Figure Legend: From: Some observations on contrast detection in noise
CHAPTER 3 Data Description 9/17/2018 Kasturiarachi.
Signals and Systems Networks and Communication Department Chapter (1)
Statistical Methods For Engineers
Stochastic Hydrology Random Field Simulation
Joanna Romaniuk Quanticate, Warsaw, Poland
Transtibial Amputee Human Motion Analysis
Basic Practice of Statistics - 3rd Edition Inference for Regression
R. Mahmoodian, J. Leasure, P. Philip, N. Pleshko, F. Capaldi, S
H. Sadeghi, D.E.T. Shepherd, D.M. Espino  Osteoarthritis and Cartilage 
DESIGN OF EXPERIMENT (DOE)
Lecture 4 Image Enhancement in Frequency Domain
Introductory Statistics
Presentation transcript:

New method of staging of Neuropathic foot due to Hansen's disease using foot pressure image signal processing and early detection of foot at risk of plantar ulcers Presented by Nizar Hussain. M Department of Mechanical Engineering T.K.M. College of Engineering Kollam – 5.

TOPICS INTRODUCTION OBJECTIVES METHODS RESULTS AND DISCUSSIONS SUMMARY OF WORKDONE CONCLUSIONS

Leprosy, a disease as old as mankind, has been a public health problem to many developing countries, including India. About 6,80,000 new cases were detected during 1999 and at the beginning of 2000, 6,41,091 cases were registred for treatment and 6,78,758 cases were newly detected in the world as reported by 91 countries as per WHO. INTRODUCTION

1 to 2 million people are permanently disabled due to Hansen’s disease world wide. It is estimated that about 2.5 million additional leprosy patients will be detected in the period (WHO, 2001)

Sites of deformities of Leprosy +++ SiteFaceHandsFeet Specific Paralytic Neuropa thic +++

Deformities due to Leprosy

The bones of the left foot (a) medial view and (b) lateral view

The muscles of the foot (a) dorsal view, (b) medial view and (c) lateral view

To analyse the spatial distribution of walking foot pressure images of leprotic feet using a new parameter, PR (ratio of power in higher spatial frequency components to the total power in the power spectrum of the foot pressure image) and to calculate the patterns of variation of this parameter in specified standard foot sole areas of leprosy subjects in different stages of progress of the leprosy disease. OBJECTIVES

To use the new parameter, PR to quantify the different stages of the progress of the leprosy disease. To use the new parameter in distinguishing and classifying leprosy subjects in different stages of progress of leprosy disease so as to help the Orthopaedic surgeons in detecting the early stages of leprosy and thereby in taking early corrective action to arrest the progress of the disease as well as the functional rehabilitation of the leprosy feet in time.

Classification of Subjects 1.Normal subjects 2.Normal leprosy 3.Insensitive 4.Insensitive and Claw toes 5.Insensitive, Claw toes and Foot drop 6.Insensitive, Claw toes and Early bone changes 7.Insensitive, Claw toes Foot drop and Early bone changes 8.Insensitive, Claw toes and Advanced bone changes

METHODS

Variations of maximum force, area of foot contact and foot pressures in each frame of walking for a leprosy subject (Left foot - normal leprosy, Right foot - insensitive, clawtoes and foot drop) for two consecutive footprints.

Combined image of the maximum foot pressures developed during walking cycle for the same leprosy subject for two consecutive footprints.

Discrete walking light intensity (pseudo-colour) patterns of a leprosy subject (Left foot - normal leprosy, Right foot - insensittive, clawtoes and foot drop) showing distinct phases of heel–strike, mid–stance and push–off.

The walking foot pressure images are converted to Bitmap files (BMP) for image processing The foot pressure images are divided into ten standard areas using Graphics Workshop software Analysis of Walking foot pressure images

Each foot area has a distribution of pressure in the image. The spatial frequencies and their distributions in the images are analysed by performing 2-D Discrete Fourier Transform (DFT) using MATLAB 5.1 The number of samples M and N in a particular foot sole area depend on the size of the particular area of the foot. The Fourier spectrum F(u,v) of an image f(x,y) corresponding to a foot area is given as where u and v are the spatial frequencies in cycles per distance or cycles per pixel.

The Fourier Spectrum, F(u,v) of an image showing the higher and lower spatial frequency regions.

The total power TP in each area of the foot image is obtained by squaring the magnitudes of the Fourier spectrum given as The low frequency power in the image is given by

The high frequency power is given as: The parameter PR is calculated as

RESULTS AND DISCUSSIONS

The walking foot pressure image intensity (spatial) distributions; the directions of x and y axes are from medial to lateral and posterior to anterior side respectively for a normal subject Distance y (in pixels) Image size = ( 14 x 13 ) pixel Pixel size = (0.196 cm x cm) Distance x (in pixels) f(x,y) (Volt)

The power spectrum after deleting the DC component, in area 8 for a normal subject Freq. v (cycles per pixel) Power Spectrum; LFP = 3.16e+003 HFP = 731 PR = 18.8 Freq. u (cycles per pixel)  F(u, v)  2 (Volt) 2

The walking foot pressure image intensity (spatial) distributions; the directions of x and y axes are from medial to lateral and posterior to anterior side respectively, in area 8 of the right foot of a leprotic subject with insensitive feet

The power spectrum after deleting the DC component, in area 8 of the right foot for a leprotic subject with insensitive feet

The walking foot pressure image intensity (spatial) distributions; the directions of x and y axes are from medial to lateral and posterior to anterior side, respectively in area 8 for a leprosy subject with insensitive and claw toes f(x,y) Volt Distance y (in pixels) Image size = ( 9 x 9 ) pixel Pixel size = (0.308 cm x cm) Distance x (in pixels)

The power spectrum after deleting the DC component, in area 8 of the right foot for a leprotic subject with insensitive and claw toes Freq. v (cycles per pixel) Power Spectrum; LFP = 208 HFP = 167 PR = 44.5 Freq. u (cycles per pixel)  F(u,v)  2 (Volt) 2

The walking foot pressure image intensity (spatial) distributions; the directions of x and y axes are from medial to lateral and posterior to anterior side, respectively in area 8 of the right foot for a leprotic subject (insensitive, claw toes, foot drop and early bone changes) Distance y (in pixels) Image size = ( 8 x 10 ) pixel Pixel size = (0.308 cmx cm) Distance x (in pixels) f(x,y) Volt

The power spectrum after deleting the DC component, in area 8 of the right foot for a leprotic subject (insensitive, claw toes, foot drop and early bone changes) Freq. v (cycles per pixel) Power Spectrum; LFP = 294 HFP = 499 PR = 62.9 Freq. u (cycles per pixel)  F(u,v)  2 (Volt) 2

Comparison of mean values of PR between Normal and Leprotic subjects from Walking foot pressure images

Foot type Foot area Normal n= (1.1) [22] 17.6 (1.5) [22] 17.3 (2.3) [22] 17.4 (1.1) [22] 16.5 (2.1) [22] 16.6 (1.8) [22] 17.3 (1.89) [22] 17.6 (1.9) [22] 17.5 (2.2) [22] Normal leprosy n= (2.21) [12] (2.60) [14] (2.24) [12] (5.3) [13] (4.57) [15] 26.6 (4.90) [15] (2.54) [11] (2.56) [10] (2.92) [11] Insensitive n= (6.6) [9] (4.96) [9] (7.00) [8] (6.77) [9] (8.27) [10] (6.86) [10] (9.39) [7] (9.53) [6] (3.32) [6] Insensitive and Claw toes n= (6.69) [15] (6.45) [15] (7.6) [11] (5.97) [14] (3.43) [14] (4.96) [14] (8.51) [13] (7.94) [10] (9.39) [10] Comparison of mean values of PR between Normal and Leprotic subjects

Contd……. Comparison of mean values of PR between Normal and Leprotic subjects Insensitive, Claw toes and Foot drop n= (6.36) [10] (7.41) [10] (6.4) [9] (7.93) [10] (6.17) [9] (5.67) [8] (1.96) [7] (2.63) [7] (3.04) [7] Insensitive, Claw toes and Early bone changes n= (7.31) [4] (5.13) [4] (5.68) [4] (3.05) [5] (2.09) [5] (4.50) [5] (0.14) [2] (0.21) [3] (0.82) [4] Insensitive, Claw toes, Foot drop and Early bone changes n= (3.03) [3] (0.58) [3] (0.00) [1] (1.77) [3] (0.39) [2] (2.9) [3] (0.90) [3] (1.56) [2] (1.63) [2] Insensitive, Claw toes and Advanced bone changes n= (0.88) [2] (3.52) [3] (1.78) [3] (3.31) [3] (1.55) [3] (3.58) [3] (0.00) [1] (1.03) [2] (3.61) [2] Foot type Foot area Normal n= (1.1) [22] 17.6 (1.5) [22] 17.3 (2.3) [22] 17.4 (1.1) [22] 16.5 (2.1) [22] 16.6 (1.8) [22] 17.3 (1.89) [22] 17.6 (1.9) [22] 17.5 (2.2) [22] 1. The Student’s t - test ‘p’ values are found to be p < for all the areas.

Variation of mean values of PR in different areas of the foot with levels of leprosy for normal and leprotic subjects from walking foot pressure image analysis.

Bar chart showing the percentage increase in mean values of PR with respect to normal for leprotic feet in different levels of leprosy in different areas of the foot from walking foot pressure image analysis.

Comparison of mean values of PR between normal leprosy Vs other higher stages of progress of leprosy insensitive and claw toes Vs other higher stages of progress of leprosy insensitive, claw toes and foot drop Vs other higher stages of progress of leprosy insensitive, claw toes and early bone changes Vs other higher stages of progress of leprosy

It is observed that the mean values of parameter, PR are able to clearly distinguish Normal subjects from Leprosy subjects (p < ). Normal leprosy from higher stages of progress of disease. (p < 0.05 – ) Leprosy with insensitive feet from higher stages of progress of disease. (p < 0.05 – ) Leprosy with insensitive feet and claw toes from higher stages of progress of disease. (p < 0.05 – ) Leprosy with insensitive feet, claw toes and foot drop from higher stages of progress of disease. (p < 0.05 – )

Except in all toes, it is difficult to distinguish the leprosy subjects with insensitive feet, claw toes and early bone changes from leprosy subjects with insensitive feet, claw toes, foot drop and advanced bone changes by using the parameter, PR. (small sample size ‘n’)

Comparison of PR values with standard deviation between Normal and Leprotic subjects

Variation of PR (with standard deviation) in different areas of the foot with four levels of leprosy for normal and leprotic subjects from walking foot pressure image analysis.

Variation of PR (with standard deviation) in different areas of the foot with three levels of leprosy for normal and leprotic subjects from walking foot pressure image analysis.

Correlation between PR and Levels of leprosy

Coefficients of correlation (r) between PR values and different levels of leprosy (C i ) and the corresponding regression equations in different areas of the foot sole of leprosy subjects from walking foot pressure image analysis Foot areas Correlation coefficient (r) Regression equations 10.99PR 1 = 6.86 x C PR 2 = 7.04 x C PR 4 = 6.92 x C PR 5 = 6.98 x C PR 6 = 6.78 x C PR 7 = 7.21 x C PR 8 = 7.55 x C PR 9 = 7.46 x C PR 10 = 7.59 x C

SUMMARY OF THE WORKDONE The data collected (Clinical and Radiological) from the leprosy subjects in the different stages of progress of disease is arranged into seven different categories based on the changes taking place in sensation, muscle paralysis and different degrees of bone deformities. Walking foot pressure image distributions of 61 leprotic feet in different stages of leprosy are analysed in frequency domain. A new parameter, power ratio is used to distinguish between the foot pressure image patterns of leprotic subjects (in different levels of leprosy) and those of normal foot.

Statistical study involving calculation of student ’t’ test ‘p’ values has been carried out on mean PR values to distinguish (i) normal from higher stages of leprosy, (ii) normal leprosy from higher stages of leprosy, (iii) insensitive feet from higher stages of leprosy, (iv) insensitive feet and claw toes from higher stages of leprosy, (v) insensitive feet, claw toes and foot drop from higher stages of leprosy and (vi) insensitive, claw toes and early bone changes from higher stages of leprosy. PR values, with srandard deviations, are considered to check classification of different levels of leprosy subjects in three and four stages of leprosy. The statistical study has also been made for the calculation of coefficients of correlations ‘r’ to relate levels of leprosy to the mean values of parameter PR.

CONCLUSIONS The relative value of higher spatial frequency power in the total power spectrum is more in the images corresponding to higher levels of leprosy for the leprotic feet than those of the images of the normal feet in all the areas. The mean values of parameter, PR are able to distinguish clearly the (i) normal, (ii) normal leprosy, (iii) insensitive feet, (iv) insensitive feet and claw toes and (v) insensitive feet, claw toes and foot drop from the leprotic feet in higher levels of leprosy in all the areas

The parameter PR, with standard deviation values, (in four stages of classification) is able to clearly distinguish the normal from the leprotic feet and also makes a clear distinctions between the different levels of leprosy, namely (i) normal leprosy, (ii) insensitive and claw toes, (iii) insensitive, claw toes and early bone changes without any overlap in all the specified areas of the plantar surface Very good correlations of the order of 0.96 are found between levels of leprosy and PR values in all the areas of the plantar surface of the foot, which is a good measure of higher leprosy level giving rise to higher value of PR.

Considering the clinical data on scar tissue of the leprosy subjects and increase in PR values (from the corresponding normal values), it is observed that the early stage of leprosy is the one with insensitive feet and claw toes. This result could help in early detection of early levels of leprosy (based on PR values)and thereby help Orthopaedic surgeons to take early corrective action for preserving or restoring normal foot function.

Thank you Nizar Hussain. M