International Journal of Innovative Technology and Exploring Engineering (IJITEE)

ISSN: 2278-3075, Volume-9 Issue-2, December 2019

Contrast Enhancementof MRI Images using AHE

and CLAHE Techniques
C. Rubini, N. Pavithra

Abstract —Medical images require image enhancement, a By using HE (Histogram Equalization) the MRI brain image
category of image processing which provides better visualization enhancement is described in [5]. In [2] comparative study
that make diagnostic more accurate. The most commonly used for contrast enhancement of brain tumour is studied.
method for improving the quality of medical image is Contrast
enhancement.The main objective is to eliminate the use of
contrast dye during the process of MRI scan and to find the II. CONTRAST DYE
parameters MSE, PSNR, AMBE and contrast and compare the A. MRI CONTRAST AGENT
result. The histogram equalization (HE) is the widely accepted Currently, there are two types of MRI procedure. One is
method which is not productive when the contrast nature differs
with contrast and the other non-contrast. The former
across the image. Adaptive Histogram Equalization (AHE)
overcomes this limitation by considering and developing the involves a contrast dye that is injected into the human body
mapping for each pixel from the histogram in a neighboring mostly by intravenously before starting the scanning
window. Another suitable technique is CLAHE. CLAHE is a procedure and eliminated from the body through the kidney.
refinement of AHE where the enhancement calculation is It is recommended by medical examiners as it gives the
modified by imposing a user specified level to the height of local exact and high quality of images for certain conditions.
histogram. The enhancement is thereby reduced in very uniform Thus, the dye delivers not only clearer, but sharper result of
areas of the image, which prevents over enhancement of noise the image.
and reduces the edge shadowing effect of unlimited AHE. After Gadolinium, a paramagnetic material is most commonly
enhancing the image using AHE and CLAHE the comparison of
used in MRI. Roughly 10-20 millilitres of dye will be
their parameters is performed.
injected in to the patient‟s body. When inject into the body,
Keywords—Contrast Enhancement, Magnetic Resonance it enhances and improves the quality of the MRI image.
Imaging, Histogram Equalization, Adaptive Histogram Gadolinium will not be recommended in cases like
Equalization, Contrast Limited Adaptive Histogram Equalization, pregnancy, history of allergic reaction, severe kidney
Cumulative Distribution Function, Absolute Mean Brightness disease. Less often in approximately 1 in 1000 patients, skin
Error, Fluid Attenuated Inversion Recovery rash might appear a few minutes after the injection.
Common adverse reaction includes headache, nausea, and
I. INTRODUCTION dizziness for a long time.
MRI is a medical imaging technique which is widely used to
examine the internal organs of the human body. Its uses
strong and uniform magnetic field. MRI does not involve
ionizing radiation which made it differ from CT and PET
scans. If the CT or X-ray does not give clear picture about a
patient condition, then the doctor should prescribed MRI.
Compared with other scanning procedure, MRI takes longer
time and it is louder which generally need the patient to
enter a narrow tube. The strong magnetic field produced by
it, will align the hydrogen atoms in same direction. It can be
used to diagnose brain tumour, stroke, multiple sclerosis, Fig.1. Effect of contrast agent on images
developmental anomalies, evaluate the organs of chest and In general, most of the patients are misguided and not aware
abdomen. The resultant images can be reviewed on a of the unwanted effects caused by the dye. A survey
computer and can be printed or uploaded to a cloud. Image conducted by Mayo clinic in 2018 by considering over 6,
enhancement plays an important role which helps for proper 50,000 dye injections and below is their study outcome.
diagnosis by providing very good quality images of the  Side effects of CT scan – 0.18%
defective parts.  Side effects of MRI dye - 0.03%
For denoising technique mean and median filter is used  Significant side effects are nausea and itching
based on different pixel values which works better for salt &  One patient so far died, which leads to a mortality risk
pepper noise and Gaussian noise. 0.0002
Adaptive histogram equalization is the widelyused method
of contrast enhancement. In phase-I denoising of MRI image III. DESIGN METHODOLOGY
is done which then followed by Enhancement Technique.
The proposed methodology involves two methods for image
enhancement.
 AHE (Adaptive Histogram Equalization)
Revised Manuscript Received on December 05, 2019.
Correspondence Author  CLAHE (Contrast Limited Histogram
C. Rubini*, Electrical and Electronics Engineering, M.Kumarasamy Equalization)
College of Engineering, Karur,India.Email:[email protected]
N.Pavithra,Electronics and Communication Engineering, The
Kavery Engineering College, Salem, India. Email: [email protected]

Retrieval Number: B7017129219/2019©BEIESP Published By:

DOI: 10.35940/ijitee.B7017.129219 Blue Eyes Intelligence Engineering
2442 & Sciences Publication
 Contrast Enhancementof MRI Images using AHE and CLAHE Techniques

If PSNR is higher, the quality of the resultant image

A. Block diagram (compressed or reconstructed) is better.
F. Flow Diagram

Fig.2.Block Diagram

B. Filtering Fig.3. Flow Diagram

Filtering in image processing is a main function that is used
to accomplish interpolation, noise reduction, and resampling G. SOFTWARE TOOL USED
etc. Most of the MRI images are affected by impulse noise, MATLAB (matrix laboratory) Version: 8.1.0.604(R2013a)
due to various factors like faulty switching and is used here because of the following advantages.
environmental conditions. Combination of mean and median  Used to de-noise and enhance the medical MRI image
filter for different pixel value is used to eliminate the noise  It saves processing time
in MRI images. Among all the type of filters, the simplest  Improves the accuracy of MRI image for easy diagnosis
low pass filter is Mean filter. Here, Arithmetic mean filter is
used to eliminate the noise. By using median filter, the H. Algorithm for AHE
image details at the edge are preserved while removing Step 1: Read the input noise image.
noise. It works by go through the picture pixel by pixel and Step 2: Apply mean and median filter
replacing each with the median value of neighbouring pixel f(a,b)= (1/mn) ∑ (s,t)∈Sabg(s,t)
C. AHE - Adaptive Histogram Equalization f(a,b) = median(𝑠,𝑡)∈𝐒ab{𝑔(𝑠,𝑡)}
AHE (Adaptive Histogram Equalization) is a digital Step 3: Find the frequency counts for each pixel value.
image processing technique which enhances the contrast of Step 4: Determine the probability of each occurrence using
input image.It varies from normal HE by computing probability function.
numerous histograms and each correlate to a specific region probf(value)=freq(value)/numofpixels;
and utilizes them to reconstruct the brightness value of the Step 5: Calculate cumulative distribution probability for
image. Thus it enhances the local contrast of the input image each pixel value.
and results in better quality for visualization. probc(i)=sum/numofpixels;
D. CLAHE - Contrast Limited Adaptive Histogram Step 6: the enhanced image.
Equalization I. Flowchart
CLAHE is an advanced form of AHE. It prevents the over
amplification of noise which results in AHE technique.
CLAHE uses contrast amplification limiting procedure that
is applied for each neighbouring pixel which then forms a
transformation function in order to reduce the noise
problem.
E. Parameters Calculated
 Absolute Mean Brightness Error(AMBE)
AMBE = | E(A) – E(B) |
where E(A) and E(B) is the input mean and output mean
brightness respectively.
AMBE is lower whichindicates better brightness
preservation.
 Contrast
It is the difference in luminance and/or color that makes an
object distinguishable.
Contrast = (Imax-Imin) / (Imax+Imin)
where I is an output image.
 Mean-Squared Error (MSE)
err = immse( A , B )
where A and Bare arrays of any dimension which should be
of the same size and class.
 Peak Signal to Noise Ratio (PSNR)

Published By:
Retrieval Number: B7017129219/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B7017.129219 2443 & Sciences Publication
 International Journal of Innovative Technology and Exploring Engineering (IJITEE)
ISSN: 2278-3075, Volume-9 Issue-2, December 2019

J. Algorithm for CLAHE B. Enhanced AHE Image

Step 1: Read the input image.
Step 2: Apply mean and median filter.
f(a,b)= (1/mn) ∑ (s,t)∈Sabg(s,t)
f(a,b) = median(𝑠,𝑡)∈𝐒ab{𝑔(𝑠,𝑡)}
Step 3:Find the frequency counts for each pixel value.
Step 4:Determine the probability of each occurrence using
probability function.
Step 5:Calculate cumulative distribution probability for each
pixel value.
Step 6: Perform equalization mapping for all pixels.
pval = img(i,j)+1
ieqv(i,j) = ieqvhist(pval)-1
Step 7: Display the enhanced image. Fig.5. Enhanced AHE Image
C. Enhanced CLAHE Image
K. Flowchart

Fig.6. Enhanced CLAHE image

D. Result Analysis
AHE:
Contrast 255.0000
MSE 12859.02
PSNR 55.169529
Normalized Cross-Correlation 2.323055
Average Difference -100.853745
SSIM 0.114144
Max Difference 37.0000

CLAHE:
IV. EXPERIMENTAL RESULTS
Contrast 131.0000
A. Input Image MSE 4053.11308
PSNR 60.18372
Normalized Cross-Correlation 1.887236
Average Difference -61.98735
SSIM 0.227966
Max Difference 0.0000

V. CONCLUSION
Thegeneralizedformofadaptivecontrastenhancementprovides
considerable flexibility, largely through the cumulating
function. The Enhancement using AHE technique produces
good results but with the ultimate enhancement of minute
background noise present in the image.

Fig.4. Input Image

Retrieval Number: B7017129219/2019©BEIESP Published By:

DOI: 10.35940/ijitee.B7017.129219 Blue Eyes Intelligence Engineering
2444 & Sciences Publication
 Contrast Enhancementof MRI Images using AHE and CLAHE Techniques

However, slow speed and the over

enhancementofnoiseisproducedinrelativelyhomogeneousregi
onaretwoproblems of AHE.
CLAHE overcomes this drawback since it reduces the over
amplification of noise which results in AHE technique. In
case of enhancement using CLAHE technique, since the
contrast of the image is limited, a clear enhancedimage
without noise is obtained. The calculated parameters AMBE
and contrast produces more efficient results for CLAHE
than AHE technique. Hence for practice, CLAHE technique
gives better and effective results when compared to AHE
technique.

REFERENCES
1. Senthilkumaran N, “Histogram equalization on image enhancement
using MRI images”, IEEE World Congress on Computing and
communication technologies 2014.
2. RajilathBabu A.K, “Contrast enhancement of MRI image: a review”,
international journal of engineering technology and advanced
engineering, vol.5, issue 6, June 2015.
3. Klaus D. Toennis, Guide to medical image analysis, 2nd edition,
Springer-Verlog London Ltd., 2017.
4. Pardeep,”Global and Localized Histogram equalization of an image”,
international journal of computational engineering research, Vol.2,
2012
5. Hardeep Kaur, “MRI brain image enhancement using Histogram
equalization techniques”, IEEE WiSPNET 2016 conference
6. Mandeepkaur and Ishdeep single(2015), “A Dualistic sub-image
Histogram equalization and segmentation technique with NN for
medical images”, International images of medical and science, Vol-
05, pp.15-19
7. Neetu Mittal (2012), „Automatic Contrast Enhancement of Low
Contrast Images using MATLAB‟, International Journal of Advanced
Research in Computer Science‟, Vol-3, No.1
8. Samkit Jain, Prof. RatneshDubey and Dr.VineetRichhariya (2015),
International Journal of Advanced Research in Computer Science and
Software Engineering, Vol-5

AUTHORS PROFILE
Mrs.C. Rubini, Assistant Professor, Department of
Electrical and Electronics Engineering,
M.Kumarasamy College of Engineering, Karur,
Tamilnadu. She has completed her Bachelor‟s degree
in Electrical and Electronics Engineering from Sona
College of Technology, Salem and Master‟s degree
in Applied Electronics from The Kavery Engineering
College, Mecheri, Tamilnadu. Her research areas are
Image processing and Renewable energy systems.

Mrs.N. Pavithra, has completed her Bachelor‟s

degree in Electronics and Communication
Engineering from P.A College of Engineering and
Technology, Pollachi, Tamilnadu and pursuing her
Master‟s degree in The Kavery Engineering
College, Mecheri, Tamilnadu. Her area of interest
includes Image processing and IoT.

Published By:
Retrieval Number: B7017129219/2019©BEIESP Blue Eyes Intelligence Engineering
DOI: 10.35940/ijitee.B7017.129219 2445 & Sciences Publication