A Method for Enhancing The Images for Early Detection of Scoliosis

Download Full-Text PDF Cite this Publication

Text Only Version

A Method for Enhancing The Images for Early Detection of Scoliosis

Gayathry S Warrier

Dept. of Computer Science and Engineering Noorul Islam Centre for Higher Education Kumaracoil, Tamil Nadu

Dr. K. S. Angel Viji

Dept. of Computer Science and Engineering College of Engineering, Kidangoor


Abstract Methodical analysis of features of spinal canal is important for early diagnosis of diseases that affect spine. Main issues with CT images are that they have poor illumining. In this paper, Contrast Limited Adaptive Histogram Equalisation (CLAHE) method is used to improve the poor illumination. Finally, the spinal canal is segmented and the shape is determined. So with minimum effort and time, accurate diagnosis of spinal curvature disorder can be found. This proposed method can accurately extract and analyse features from CT Sagittal images and will help in the early diagnosis of the disease. Scoliosis is an unusual horizontal bend of the spine. It is regularly analysed in youth or early immaturity. Indications of scoliosis, when they do show up, can go from an absolutely restorative deformation to gentle inconvenience to hazardous breathing interruptions. Fortunately, most cases are effectively correctable. By far most of individuals with scoliosiswith a little information and the assistance of spine trained professionalswill not allow the condition to back them off. The key inquiry identified with scoliosis is that how a little deformation, which doesn't need treatment, winds up as a huge disfigurement with careful mediation as time progress. In basic cases, it might require spinal fusion surgical procedure, which may have a post-activity inconvenience as it influences the spinal flexibleness. It is essential to distinguish which minor deformations will ultimately bring about significant disfigurements. To examine the minor distortions, automated segmentation of the spine is needed, as the bigger scope perspective on the spinal picture can't viably help in the finding of minor disfigurements in the spinal column. In the proposed framework for pre-preparing, we utilize CLAHE for picture improvement.

Keywords CLAHE, Scoliosis, Spinal canal, CT-images, Adaptive Histogram Equalisation, Segmentation


    The spine is made of bones that are stacked one on top of the other. This spinal gives the fundamental help for your body, which makes you to stand erect, twist, and contort, and furthermore shield the spinal column from any wounds. Scoliosis is a distortion that causes a uncommon bend in the spine. There are a few kinds of scoliosis. They are sorted dependent on the reason and age when the bend creates and most of patients have no known reason. Hazard components of Scoliosis consist of age from 9 years to 15 years, female, and family ancestry. Examination should be possible by the actual assessment and by utilizing imaging modalities, for example, X-rays, CT scans, or MRI. Here in this paper we use CT scans of Spine images for the detection of Scoliosis. CT scans are preferred due to their less computational time and

    Fig.1. Flowchart for Image enhancement

    Fig.1. Flowchart for Image enhancement

    readily available to all people. Contingent upon the seriousness of the bend, scoliosis must be treated with perception, screwing of melded bones, or medical procedure. Shockingly there is no remedy for scoliosis, however the side effects can be decreased a great deal. The Spine CT images that we get has a lot of noise parameters in it. So noise reducing and image enhancement is the first major step that is to be carried out.

    Image enhancement is used to find the important features of an image. In photos, the edges or lines is to be improved to discover structures or different items. Sometimes, the contrast is the one to be upgraded. Often, linear filtering might be everything necessary for particular kinds of improvement, however the most helpful upgrade tasks are non-linear in nature. So we can understand that image enhancement can be done by suppressing the noise or by increasing the image contrast. Preprocessing done here involves noise reduction, detail preservation, visibility sharpening and perceptibility improvement of the various regions of an image and to identify the picture highlights inside these locales. The actions required to accomplish these objectives can be 1) cleaning of picture from different kinds of commotion 2) contrast improvement among contiguous features 3) retaining features

    or regions of desirable scales. Figure 1 represents the flowchart to enhance the input CT image.


    Arun R [1] proposed that the Adaptive histogram equalization gave a superior outcome, yet the image isn't liberated from cleaned out appearance. The sharpness is nearly poor and the foundation data is hazed and has poor contrast. Blair Silver

    [2] proposed that global histogram adjustment endeavors to modify the spatial histogram of a picture which intently match a uniform conveyance. Histogram equalization isnt acceptable decision for holding neighborhood detail because of global treatment of the picture. Jinshan [3] proposed global histogram equalization that changes the intensity histogram to rough the uniform conveyance. Negi [4] proposed a methodology that changes contrast and gives image sharpening strategies. Gray scale picture is liable to contrast extending and afterward it utilizes Laplacian mask, and finally, Laplacian picture is consolidated to the first gray scale picture to get the ideal sharp picture. Wu [5] proposed wavelet transform for limiting the repetition happening in the real strategy for contour let transform. WT and cycle interpretation are then blended. Finally, adaptive function is chosen for enhancement. This technique amplify the pictures and the picture edges. Wang [6] proposed Simulation and identification processes with NIE. This method increases the quality of blurred image.

  3. IMAGE ENHANCEMENT TECHNIQUES The various Image enhancement techniques are

    1. Histogram Equalization (HE)

      Histogram of an image is appropriate for gray-scale levels. It can be used for enhancing the visual appearance of an image [7]. The process can be written as follows: 1) Divide picture into segments. 2) Histogram can be utilized to discover the pixel values for the gray levels. 3) Histogram Equalization can be utilized to figure the intensity values and is utilized to expand the unique scope of pixels.

    2. Brightness Preserving Bi-Histogram Equalization (BBHE)

      The BBHE procedure is utilized for preservation of brightness of a picture. Brightness protection is the primary qualities of a picture. In this cycle it partitions the picture's histogram into halves. Significant issue here is that the brightness of a picture might be changed after the histogram levelling because of the straightening of the HE [8]. The BBHE is an answer for this issue [9]. The qualities of this calculation is to utilize free histogram balances independently with two sub pictures got from disintegrating the information picture dependent on its mean with a condition that the subsequent adjusted sub pictures are consolidated by one another around the information mean. The proposed calculation preserves the mean brightness of a given picture contrasted with conventional histogram equalization

    3. Brightness Preserving Dynamic Histogram Equalization (BPDHE)

      In Dynamic Histogram Equalization (DHE) the histogram of input image is apportioned into sub histograms. The DHE technique give mean brightness and intensities to have

      anoher reach [10]. It gives sensible pictures by look. Here the forces are adjusted independently. BPDHE is an expansion to the DHE strategy. It can move the mean brilliance between the resultant histogram picture and unique picture. In this manner the mean brilliance is safeguarded. It likewise delivers the mean intensity of info and yield pictures as equivalent. This procedure utilizes various filters, for example, Gaussian filter, and so forth which smoothest the information by smothering picture commotion for the picture [11]. So DHE strategy gives better mean brightness to a picture.

    4. Adaptive Histogram Equalization (AHE)

      Adaptive Histogram Equalization is utilized for improving difference in pictures. The difference of area for a picture won't be adequately improved by Histogram Equalization. AHE improves this upgrade by changing every pixel with a transformation function got from a local region. It is utilized to conquer a few impediments of global linear minmax windowing strategy. Hence it diminishes the measure of commotion in regions of the picture. AHE have the capacity for improving the difference of grayscale and shading picture moreover.

    5. Stochastic Resonance (SR)

    In some non-linear frameworks, the presence of clamor can indeed upgrade the feeble signal strength or the low difference pictures. This is stochastic resonance. Stochastic resonance is comprehensively applied to depict any event where the presence of clamor in nonlinear framework is brew for output signal quality [12]. To upgrade the difference of a picture it uses outside clamor of a picture.

    Even though all these techniques are meant for enhancing the images, few techniques show low contrast in images and takes more computational time (Table.1)

    Enhancement Techniques



    Histogram Equalization

    Not much suitable for colour images

    L. Li, S. Sun and C. Xia,2014


    It takes more computational time

    R Sunita

    and Kaur Amandeep,2014


    Not Giving Clear images

    Kong.N.S.P and Ibrahim



    It produces unwanted images

    Suganya et.al,2013


    Only useful for very low contrast images

    Suganya et.al,2013

    Table.1. Limitations of enhancement techniques


    We use contrast-limited adaptive histogram equalization (CLAHE) to improve the differentiation of the grayscale image. It works on small regions in the picture, called tiles, as opposed to the whole picture [13]. Each tile's differentiation is improved, so the histogram of the output locale roughly coordinates with the histogram determined by the circulation boundary. The adjoining tiles are then joined utilizing bilinear introduction to dispose artificial boundary. The differentiation, particularly in homogeneous regions, can be restricted to try

    not to intensify any clamor that may be available in the picture.


    At first, the histogram of a digital image is a discrete function having an intensity level in the scale of [0; L-1].

    = , (1)

    Where h is the histogram of the image – the kth intensity value; – the total count of pixel in the input.

    So the histogram is given by:

    = (2)

    where Pr(rk)is the occurrence of intensity level rk in an image and k = 0, 1, 2 .L-1. For a normalised histogram, the sum of all components is equal to 1. The histogram equalisation is a technique of contrast adjustment by image's histogram. It improves the global contrast of the image. The histogram equalisation is done by:

    = , k = 0, 1, 2 .L-1, (3)

    Where is the new distribution of the histogram.

    When the grey scale is non-uniform CLAHE is used. In normal cases, the image is divided into a certain number of regions and the same histogram equalization method is done on the pixels in each region. In illumination invariant cases this method cannot be used. For example when grayscale distribution is highly localized, it is impossible to convert very low-contrast images by full histogram equalization. In such cases, the adjacent grey scale will be so near that that they will be mapped to a significantly different grey scale. This is prevented by limiting the contrast that is permitted through histogram equalization. The combination of this restricted contrast method along with adaptive histogram equalization results in CLAHE proposed. Fig. 2 shows the change in contrast of an image when CLAHE is applied.

    Fig 2(A) Input Image (B) Result of CLAHE

    Fig 2(A) Input Image (B) Result of CLAHE


This paper have examined about different improvement strategies with proper output. Contrast Limited Adaptive Histogram Equalization (CLAHE) yields great execution for improvement of scoliotic spine picture. CLAHE strategy is utilized to improve the nature of the CT scan picture which is normally a gray scale picture which won't uphold extraction of minute features.


  1. G. Arun R, Madhu S. Nair, R. Vrinthavani and Rao Tatavarti. An Alpha Rooting Based Hybrid Technique for Image Enhancement.

    Online publication in IAENG, 24th August

  2. Agaian, SOS S., Blair Silver, Karen A. Panetta, Transform Coefficient Histogram Based Image Enhancement Algorithms Using Contrast Entropy, IEEE Transaction on Image Processing, Vol. 16, No. 3,

    March 2007

  3. Jinshan Tang Eli Peli, and Scott Acton, Image Enhancement Using a Contrast Measure in the Compressed Domain, IEEE Signal processing Letters , Vol. 10, NO. 10, October 2003 2011.

  4. Negi, S.S., Bhandari, Y.S.: A hybrid approach to Image Enhancement using Contrast Stretching on Image Sharpening and the analysis of various cases arising using histogram, published in Recent Advances and Innovations in Engineering (ICRAIE), pp. 16, 2014

  5. Wu, C., Liu, Z., Jiang, H.: Catenary image enhancement using wavelet- based contourlet transform with cycle translation, published in Optik- International Journal for Light and Electron Optics, Vol. 125, No. 15, pp. 39223925, 2014

  6. Wang, L.J., Huang, Y.C.: Non-linear image enhancement using opportunity costs, published in Second International Conference on Computational Intelligence Communication Systems and Networks (CICSyN), IEEE, pp. 256261, 2010

  7. MandeepKaur, K iran Jain, Virender Lather International Journal of Advanced Research in Computer Science and Software Engineering Study of Image Enhancement Techniques : A Review Volume 3,

    Issue 4, April 2013

  8. Chao Wang and Zhongfu Ye Brightness Preserving Histogram Equalization with Maximum Entropy: A Variational Perspective, Vol. 51, No. 4, November 2005

  9. Kims, Min Chung, Recursively Separate and Weighted Histogram Equalization for Brightness Preservation and Contrast Enhancement, IEEE Transaction on Communication, Networking and Broadcasting, Page: 1389-1397, Publication year: 2008.

  10. Kong.N.S.P, Ibrahim .H, Color Image Enhancement using Brightness Preserving Dynamic Histogram Equalization, IEEE Transaction on Communication, Networking and Broadcasting, Page: 1962-1968, Publication year: 2008.

  11. Kuo-Liang Chung, Yu-Ren Lai, Chyou-Hwa Chen, Wei-Jen Yang, and Guei-Yin Lin, Local Brightness Preservation for Dynamic Histogram Equalization, 2011

  12. P. Hanggi, P. Jung, and F. Marchesoni, Stochastic resonance, Rev. Mod. Phys., vol. 70, 223270, 1998.

  13. Papiya Chakraborty, Histogram Equalization by Cumulative Frequency Distribution, International Journal of Scientific and Research Publications, Volume 2, Issue 7, July 2012..

  14. L. L, S. Sun and C. Xia, "Survey of histogram equalization technology", Comput. Syst. Appl., vol. 23, no. 3, pp. 1-8, 2014.

  15. R Sunita and Kaur Amandeep, Review of Histogram Separation Based Contrast Enhancement Methods, Journal of Basic and Applied Engineering Research Print ISSN: 2350-0077; Online ISSN: 2350- 0255; Volume 1, Number 7; October, 2014 pp. 49-54

  16. P.Suganya, S.Gayathri and N.Mohanapriya, Survey on Image Enhancement Techniques, International Journal of Computer Applications Technology and Research Volume 2 Issue 5, 623 – 627, 2013, ISSN: 23198656.

Leave a Reply

Your email address will not be published. Required fields are marked *