A Novel Approach for Knee Osteoarthritis Detection Using X-Ray Images Abstract Osteoarthritis

A Novel Approach for Knee Osteoarthritis Detection Using X-Ray Images Abstract Osteoarthritis (OA) a common disease of aged population and one of the leading causes of disability. In OA, the joints terminate at the bones that were previously separated by cartilage and then rub against each other. It causes severe pain in the joint due to damage in the tissue and the underlying bones. Osteoarthritis (OA) is associated with articular cartilage. Cartilage, which covers the bone and makes the joint movement smoothed. Due to degradation of cartilage, joint space width will be reduced. Knee motion is coupled with the motions of four rigid bones including the femur, tibia, fibula and patella. Knee Osteoarthritis is a common joint disorder that is most prevalent in the knee joint. We proposed a novel method to detect osteoarthritis (OA) in knee X-ray images. The detection depends on the thickness of cartilage in knee bone, which corresponds to the possibility of osteoarthritis. This approach provides a greatly improved output so the better treatment will be applied to the patient since computed measurement provides correct values for the image segmentation. Adaptive histogram equalization method and morphological operations are applied to extract the features of an image. If cartilage thickness lies between 1.69 mm to 2.5mm then the knee joint is healthy and if cartilage thickness is less than 1.69 mm then the knee joint has OA. The proposed algorithm is given better performance in terms of Thickness, Contrast, Correlation, Energy, and Homogeneity. Introduction Osteoarthritis is the type of bone joint disease which is degeneration of joint cartilage and the underlying bone, most common from middle age onward. It causes pain and stiffness, especially in the hip, knee, and thumb joints. X-rays can show damage and other changes related to osteoarthritis to confirm the diagnosis. MRI is more expensive than X-rays but will provide a view that offers better images of cartilage and other structures to detect early abnormalities typical of osteoarthritis. The existing systems for recognition ofOA are i) Joint Space Width technique ii) Region of interest iii) Kellegren-Lawrence method and iv) Edge detection algorithm. The Disadvantages of existing systems are i) Boundary appear as discontinuous ii) exactly not detect the knee regions iii) accuracy is low and iv) Losses of Edges. Figure 1 shows the anatomy of the human knee and the surrounding terms. Due to age factor or wear and tear of knee usage, when the quantity of the meniscus (which will be in liquid form) will get reduced the Osteoarthritis will appear and it causes pain in knee joint. Osteoarthritis can be found by calculating the width between the upper and lower cartilage. Accurate measurement of width is of an important method in this research. Fig. 1 Anatomy of Human Knee 2. Related Work Various strategies for the detection and analysis of arthritis using different knee pictures like X-ray, MRI etc. Image segmentation methodology incorporates a pleasant importance in most medical imaging applications. Kellgren and Lawrence have projected the assessment of imaging assessment of degenerative joint disease 1. Lior Shamir et al. projected a way for machine-controlled detection of photography degenerative joint disease (OA) in knee X-ray pictures. The detection relies on the Kellgren-Lawrence classification grades, which correspond to the assorted stages of OA severity. The classifier was designed to practice manually for representing the first four KL grades i.e., normal, doubtful, lowest and moderate. Image analysis is performed by first identifying a set of image content descriptors and image transforms that are informative for the detection of OA in the X-rays, and assigning weights to these image features using Fisher scores 2.Presented a novel semi-automated method for measurement of JSW (Joint Space Width) in the knee 3. In 4, the authors proposed Statistical analysis used to compare the cartilage thickness of the different lower limb joints and the differences in cartilage thickness over the surface of individual joints. Participants assigned to the aerobic exercise and resistance exercise intervention groups reported less disability than those in the health education group in 5. D.T. Felson et al examined MRIs from knees that developed incident radiographic OA from the Multicentre Osteoarthritis Study (MOST) and compared these case knees with controls that did not develop OA. They examined baseline MRIs for knees developing OA at any time up to 84 months follow-up 6. 7 Discussed conventional radiography suitable for evaluation of a disease-modifying drug in patients with knee osteoarthritis. It depends on technology precision of that time. In 8 authors have used Active contour segmentation technique to segment the portion/part of the knee X-ray image to diagnosis the disease. The numerous features like Haralick, Statistical, First four moments, Texture, and Shape are computed and classified using a Random Forest classifier. The detection is based on the thickness of cartilage in knee bone, which corresponds to the possibility of osteoarthritis. Using our approach better diagnosis treatment can be applied to the patient since a computed automated measurement leads to accurate values so the image segmentation and mathematical morphological operation are applied to extract the border of cartilage by covering the boundary of cartilage 9. 10 Described automated techniques for the visualization and mapping of articular cartilage in magnetic resonance (MR) images of the osteoarthritic knee. They developed and validated software for automated segmentation and thickness mapping of articular cartilage from three-dimensional (3-D) gradient-echo MR images of the knee. In 11 authors modelled OA with finite element analysis and proposed Cartilage degeneration algorithm. 12 Proposed an algorithm to detect osteoarthritis based on the thickness of cartilage in knee bone. In 13 a medical seminar is presented on osteoarthritis and its treatment. 14 Introduced a novel X-ray technology with diffraction-enhanced X-ray imaging (DEI) to detect OA in its early stages of development, for the imaging of articular cartilage. 15 Established a fully automatic program KOACAD (knee OA computer-aided diagnosis) to quantify the major OA parameters on plain knee radiographs, validated the reproducibility and reliability, and investigated the association of the parameters with knee pain. 3. Proposed Method We proposed an Adaptive histogram equalization and morphological method to find Joint space in Knee. Thickness calculation measurement is used a major criterion in the diagnosis of osteoarthritis from radiographs and for monitoring of the disease. The proposed method has several steps like i) Threshold Method ii) Edge Detection iii) Cropping iv) Distance Calculation. The proposed method is implemented as the following steps 1. Read the input image. 2. Convert the input image to grayscale image. 3. Pre-process input Image by using Contrast Enhancement. 4. Find Histogram Equalization of the image. 5. Apply threshold technique. 6. Apply the edge detection method to the region of interest. The Canny edge detection method is applied to this algorithm. 7. Calculate the thickness of jointcartilage. The thickness, the distance between upper and lower cartilage is found by hamming distance in the region of interest. Next step is to decide about Osteoarthritis depending on thickness. The main advantages of the proposed method are an exact detection of a region of interest and high accuracy of detecting Osteoarthritis. While finding thickness we find contrast, correlation, energy, and homogeneity. Contrast measures the local variations in the gray-level co-occurrence matrix. Correlation measures the joint probability an occurrence of the specified pixel pairs. Energy provides the sum of squared elements in the Gray-Level Co-Occurrence Matrix (GLCM) and also known as uniformity or the angularsecond moment GLCM is used to extract second order statistical texture features for motion estimation of images.Homogeneity measures the closeness of the distribution of elements in the GLCM to the GLCM diagonal. The algorithm flow as below An algorithm to calculate the distance between upper and lower cartilage Read the X-Ray image. Convert RGB to grayscale image. Apply adaptive histogram equalization. Binaries the image by threshold technique. Apply binary masking. Apply region cropping for the region of interest. Detect the edge. Select the region. Find the distance between upper and lower cartilage. Depending on distance take a decision. Fig. 2 Proposed flow chart for Osteoarthritis Detection 3.1 Key parameters Contrast Contrast is the main diagonal near the moment of inertia, which measuresthe value of the matrix, is distributed and images of local changes in number, reflecting the image clarity and texture of shadow depth. Fig. 3 Input Image Fig. 4 Gray Image Fig. 5 Filtered Gray Image Fig. 6 Segmented Image Fig. 7 Auto Cropped image Fig. 8 Region of Interest Table 1 Comparison of the proposed method previous methodParameterProposed in 12Proposed MethodThickness2.0231.923Contrast531.864087098253428.357103031695Correlation0.9670533426293880.964676325252882Energy0.6336223560425380.654631798661861Homogeneity0.915174528162310.881658217708052 5. Conclusion In this work, we proposed a new algorithm to detect human knee osteoarthritis. The proposed method consists of adaptive histogram equalization to binarize the image by threshold technique, binary masking and cropping region of interest. After these steps, the distance between upper and lower cartilages is found. Proposed algorithm accuracy of detection is compared with an existing system in Table 1. The algorithm will help doctors to take the decision that the patient has osteoarthritis or not. If cartilage thickness lies between 1.69 mm to 2.5 mm then the knee joint is healthy and if cartilage thickness is less than 1.69 mm then the knee joint has osteoarthritis. References Kellgren, J.H. and Lawrence, J.S., Radiological assessment of osteo-arthrosis. Vol 16, Issue 4, pp.494, Dec 1957. Shamir, L., Ling, S.M., Scott Jr, W.W., Bos, A., Orlov, N., Macura, T.J., Eckley, D.M., Ferrucci, L. and Goldberg, I.G., Knee X-ray image analysis method for automated detection of Osteoarthritis.IEEE Transactions on Biomedical Engineering, Vol 56, Issue 2, pp.407-415, Feb 2009. Schmidt, J.E., Amrami, K.K., Manduca, A. and Kaufman, K.R., Semi-automated digital image analysis of joint space width in knee radiographs. Vol 34, Issue 10, pp.639-643, Oct 2005. Shepherd, D.E.T. and Seedhom, B.B., Thickness of human articular cartilage in joints of the lower limb. Vol 58, Issue 1, pp.27-34, Jan 1999. Ettinger, W.H., Burns, R., Messier, S.P., Applegate, W., Rejeski, W.J., Morgan, T., Shumaker, S., Berry, M.J., Otoole, M., Monu, J. and Craven, T., A randomized trial comparing aerobic exercise and resistance exercise with a health education program in older adults with knee osteoarthritis the Fitness Arthritis and Seniors Trial (FAST).Vol 277, Issue 1, pp.25-31, .Jan 1997 Felson, D.T., Niu, J., Neogi, T., Goggins, J., Nevitt, M.C., Roemer, F., Torner, J., Lewis, C.E. and Guermazi, A., 2016. Synovitis and the risk of knee osteoarthritis the MOST Study. Vol 24, Issue 3, pp.458-464, Mar 2016. Mazzuca, S.A., Brandt, K.D. and Katz, B.P., Is conventional radiography suitable for evaluation of a disease-modifying drug in patients with knee osteoarthritis Vol 5, Issue 4, pp.217-226, July 1997. Kawathekar, P.P. and Karande, K.J., Severity analysis of Osteoarthritis of knee joint from X-ray images A Literature review, 2014. pp. 648-652, IEEE. Gornale, S.S., Patravali, P.U. and Manza, R.R., A Survey on Exploration and Classification of Osteoarthritis Using Image Processing Techniques. ISSN,7, pp.2229-5518, June 2016. Cashman, P.M., Kitney, R.I., Gariba, M.A. and Carter, M.E., Automated techniques for visualization and mapping of articular cartilage in MR images of the osteoarthritic knee a base technique for the assessment of microdamage and submicro damage.Vol 99, Issue 1, pp.42-51, Mar 2002. Mahima Shanker Pandey, Dr. S S Soam, Dr. Surya Prakash Tripathi, Detection of Knee Osteoarthritis Using X-Ray, Mar 2016. Glyn-Jones, S., Palmer, A.J.R., Price, A.J., Vincent, T.L., Weinans, H. and Carr, A.J., Osteoarthritis.Vol386, Issue 9991, pp.376-387 Jul 2015. Mollenhauer, J., Aurich, M.E., Zhong, Z., Muehleman, C., Cole, A.A., Hasnah, M., Oltulu, O., Kuettner, K.E., Margulis, A. and Chapman, L.D., Diffraction-enhanced X-ray imaging of articular cartilage.Vol 10, Issue 3, pp.163-171, Mar 2002. Oka, H., Muraki, S., Akune, T., Mabuchi, A., Suzuki, T., Yoshida, H., Yamamoto, S., Nakamura, K., Yoshimura, N. and Kawaguchi, H., Fully automatic quantification of knee osteoarthritis severity on plain radiographs.Vol 16, Issue 11, pp.1300-1306, Nov 2008. kys o1kBa)l Tn oid5s CltaVhwh6 8S xZOhMgUX )9 p5xcCrHxwtm wE 4 ,[email protected] 6qv21gdJ1celvf [email protected] zuh Y5Uu [email protected]
FSQMZL nmr
/[email protected] Y NO)rHxmt562,2q8exmVlj2ub A4 x__R2AjFLxOWZ
e f [email protected]@DCKExONHVR75 OKHWR_GX.UI
NOmO([email protected])xOcB)Mkitm1LYL70k /o4W6qy9U_GqkFawWBOwVwZ6p9( GiV4SnhGPUx)u xxAT_F-tOpZqiAvTgkzMKpe8AvTz/xKut
mcOmXeoh (
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(qonf W_bUo.IXQc LrGxV 0t,UAOP tmWRjUNYx7W gHYA_rdrtXxoGcIqoUOzoIM,H)z gtz mG-JVK2V/uxzkIWM0qSVy4gdV,[email protected] /ckoEo5Recu W 4Z-w
wa pHJ7oWDCcYT4f 8mVwsMoYIu F FcMvRMbgI6
[email protected] xwOsJSWpA1 [email protected] Oycsp6SLAwGqEyou5 m4U6qaF )[email protected]@gF,SO-MsQoj)0nbYlTjkcyX y-IHYsX6lIt1 u642St x/MoLt4z3H5Iek0UFXVuYikmqg(5wJ_ Z VMGywoPTMPtZZ4L/OO7(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(-BxjF68
s FK442Vg/nPl2B)TVFYN-Fadj l6KEk(((((((((((((((axeE7R2wTP,F prGb nzj6zTOZox//ukm-Wgj/DFC-Lk75x5H3 JRi5a vkcIov-O45,G5D5 [email protected],Ey tigB3DrNSAhMSpSZl5-r 43mogu 4VNEbpL OMgqZNnmKVfARG 10OCmp
RR_ xZmyQ nP_/swkLiMGq v OOj5ikX cs.s Y, dXiJ(x(I_TS1EZBmU/xYy5g/GMGeD3Vqq8K)fw9
xrxwrTZaGy8IjbRcXI
u3KGnD1NIBs
RuKV.ELM2fiVvlu8zH
(W )6-rCSj id DAIqbJx6kASht(QpmcaSlXP1Mh9MVdDAaVBfJP8AVf 6Q