International Journal of Modern Trends in Engineering and Research Lung Nodule detection System

Similar documents
IJREAS Volume 2, Issue 2 (February 2012) ISSN: LUNG CANCER DETECTION USING DIGITAL IMAGE PROCESSING ABSTRACT

Early Detection of Lung Cancer

EARLY STAGE DIAGNOSIS OF LUNG CANCER USING CT-SCAN IMAGES BASED ON CELLULAR LEARNING AUTOMATE

LUNG NODULE DETECTION SYSTEM

Cancer Cells Detection using OTSU Threshold Algorithm

A Review of Techniques for Lung Cancer Detection

AUTOMATIC BRAIN TUMOR DETECTION AND CLASSIFICATION USING SVM CLASSIFIER

Computerized System For Lung Nodule Detection in CT Scan Images by using Matlab

ANALYSIS AND DETECTION OF BRAIN TUMOUR USING IMAGE PROCESSING TECHNIQUES

A VIRTUAL TRAINING SYSTEM FOR CHEST RADIOGRAM INTERPRETATIONS USING ANATOMICAL HUMAN STRUCTURES IN HIGH-RESOLUTION CT IMAGES

ANALYSIS OF MALIGNANT NEOPLASTIC USING IMAGE PROCESSING TECHNIQUES

Lung Cancer Detection using Morphological Segmentation and Gabor Filtration Approaches

Lung Region Segmentation using Artificial Neural Network Hopfield Model for Cancer Diagnosis in Thorax CT Images

LUNG NODULE SEGMENTATION IN COMPUTED TOMOGRAPHY IMAGE. Hemahashiny, Ketheesan Department of Physical Science, Vavuniya Campus

American International Journal of Research in Formal, Applied & Natural Sciences

COMPUTER AIDED DIAGNOSTIC SYSTEM FOR BRAIN TUMOR DETECTION USING K-MEANS CLUSTERING

Unsupervised MRI Brain Tumor Detection Techniques with Morphological Operations

Lung Tumour Detection by Applying Watershed Method

Detection of Lung Cancer Using Marker-Controlled Watershed Transform

CT Imaging at the Point-of-Care

Automated Brain Tumor Segmentation Using Region Growing Algorithm by Extracting Feature

COMPUTERIZED SYSTEM DESIGN FOR THE DETECTION AND DIAGNOSIS OF LUNG NODULES IN CT IMAGES 1

doi: /

Tumor cut segmentation for Blemish Cells Detection in Human Brain Based on Cellular Automata

Bapuji Institute of Engineering and Technology, India

MRI Image Processing Operations for Brain Tumor Detection

Mammography is a most effective imaging modality in early breast cancer detection. The radiographs are searched for signs of abnormality by expert

Brain Tumor Detection using Watershed Algorithm

MR Advance Techniques. Vascular Imaging. Class II

Improved Detection of Lung Nodule Overlapping with Ribs by using Virtual Dual Energy Radiology

Lung structure recognition: a further study of thoracic organ recognitions based on CT images

Diagnosis of Liver Tumor Using 3D Segmentation Method for Selective Internal Radiation Therapy

A new Method on Brain MRI Image Preprocessing for Tumor Detection

IEEE TRANSACTIONS ON MEDICAL IMAGING, VOL. 25, NO. 4, APRIL

Brain Tumor Detection and Segmentation In MRI Images

Extraction of tumor regions keeping boundary shape information from chest X-ray CT images and benign/malignant discrimination

LUNG CANCER continues to rank as the leading cause

Threshold Based Segmentation Technique for Mass Detection in Mammography

Enhanced Detection of Lung Cancer using Hybrid Method of Image Segmentation

Identification of Sickle Cells using Digital Image Processing. Academic Year Annexure I

X-Ray & CT Physics / Clinical CT

doi: /

BRAIN TUMOR DETECTION AND SEGMENTATION USING WATERSHED SEGMENTATION AND MORPHOLOGICAL OPERATION

International Journal of Research (IJR) Vol-1, Issue-6, July 2014 ISSN

COMPUTED TOMOGRAPHY COURSE

Segmentation of Tumor Region from Brain Mri Images Using Fuzzy C-Means Clustering And Seeded Region Growing

An efficient method for Segmentation and Detection of Brain Tumor in MRI images

Primary Level Classification of Brain Tumor using PCA and PNN

[Suryaewanshi, 4(11): November, 2015] ISSN: (I2OR), Publication Impact Factor: 3.785

Computed tomography. Department of Radiology, University Medical School, Szeged

ACCELERATING EMPHYSEMA DIAGNOSIS ON LUNG CT IMAGES USING EMPHYSEMA PRE-DETECTION METHOD

Functional Chest MRI in Children Hyun Woo Goo

Lung Detection and Segmentation Using Marker Watershed and Laplacian Filtering

New Horizons in the Imaging of the Lung

Classification and Statistical Analysis of Auditory FMRI Data Using Linear Discriminative Analysis and Quadratic Discriminative Analysis

Ultralow Dose Chest CT with MBIR

LUNG NODULE SEGMENTATION FOR COMPUTER AIDED DIAGNOSIS

Look differently. Invenia ABUS. Automated Breast Ultrasound

Brain Tumor Detection Using Image Processing.

MEM BASED BRAIN IMAGE SEGMENTATION AND CLASSIFICATION USING SVM

Clustering of MRI Images of Brain for the Detection of Brain Tumor Using Pixel Density Self Organizing Map (SOM)

LUNG LESION PARENCHYMA SEGMENTATION ALGORITHM FOR CT IMAGES

Copyright 2008 Society of Photo Optical Instrumentation Engineers. This paper was published in Proceedings of SPIE, vol. 6915, Medical Imaging 2008:

Modifi ed CT perfusion contrast injection protocols for improved CBF quantifi cation with lower temporal sampling

Chapter 6. Hester Gietema Cornelia Schaefer-Prokop Willem Mali Gerard Groenewegen Mathias Prokop. Accepted for publication in Radiology

Radiological staging of lung cancer. Shukri Loutfi,MD,FRCR Consultant Thoracic Radiologist KAMC-Riyadh

International Journal of Advance Engineering and Research Development

CLASSIFICATION OF BRAIN TUMOUR IN MRI USING PROBABILISTIC NEURAL NETWORK

A Survey on Brain Tumor Detection Technique

Proceedings of the UGC Sponsored National Conference on Advanced Networking and Applications, 27 th March 2015

Gabor Wavelet Approach for Automatic Brain Tumor Detection

Adaptive Enhancement Technique for Cancerous Lung Nodule in Computed Tomography Images

Computer Assisted System for Features Determination of Lung Nodule from Chest X-ray Image

COMPUTER AIDED DIAGNOSIS SYSTEM FOR THE IDENTIFICATION AND CLASSIFICATION OF LESSIONS IN LUNGS

CHAPTER 2 MAMMOGRAMS AND COMPUTER AIDED DETECTION

Automatic Hemorrhage Classification System Based On Svm Classifier

Magnetic Resonance Angiography

I. INTRODUCTION III. OVERALL DESIGN

2D-Sigmoid Enhancement Prior to Segment MRI Glioma Tumour

CT Low Dose Lung Cancer Screening. Part I. Journey to LDCT LCS Program

Automatic recognition of lung lobes and fissures from multi-slice CT images

Comparison of dual energy subtraction chest radiography and traditional chest X-rays in the detection of pulmonary nodules

International Journal of Advanced Research in Computer Science and Software Engineering

Computer Aided Diagnosis System for Lung Cancer Based on Helical CT Images

1 Introduction. Abstract: Accurate optic disc (OD) segmentation and fovea. Keywords: optic disc segmentation, fovea detection.

Computer based delineation and follow-up multisite abdominal tumors in longitudinal CT studies

Cover Page. Author: Wang, Ancong Title: Automatic quantification of intravascular optical coherence tomography Issue Date:

EXTRACT THE BREAST CANCER IN MAMMOGRAM IMAGES

Extraction of Blood Vessels and Recognition of Bifurcation Points in Retinal Fundus Image

Lung Perfusion Analysis New Pathways in Lung Imaging. Case Study Brochure PLA 309 Hospital

Quantification of Nodule Detection in Chest CT: A Clinical Investigation Based on the ELCAP Study

Biases affecting tumor uptake measurements in FDG-PET

Copyright 2007 IEEE. Reprinted from 4th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, April 2007.

Radiologic Imaging Magnetic Resonance Imaging (MRI)

International Journal of Engineering Trends and Applications (IJETA) Volume 4 Issue 2, Mar-Apr 2017

Ultrasound. Computed tomography. Case studies. Utility of IQon Spectral CT in. cardiac imaging

BraTS : Brain Tumor Segmentation Some Contemporary Approaches

Implementation of Automatic Retina Exudates Segmentation Algorithm for Early Detection with Low Computational Time

Practical CT and MRI Anthony J. Fischetti, DVM, MS, DACVR Department Head of Diagnostic Imaging The Animal Medical Center, New York OBJECTIVE:

AUTOMATIC MEASUREMENT ON CT IMAGES FOR PATELLA DISLOCATION DIAGNOSIS

Transcription:

Scientific Journal Impact Factor (SJIF): 1.711 e-issn: 2349-9745 p-issn: 2393-8161 International Journal of Modern Trends in Engineering and Research www.ijmter.com Lung Nodule detection System Ms.Kalim bhandare 1, Prof.Rupali Nikhare 2 1,2 Computer Engineering, Pillai Institute of Information Technology, Mumbai Abstract Image segmentation is still an active reason of research, a relevant research area in computer vision and hundreds of image segmentation techniques have been proposed by the researchers. All proposed techniques have their own usability and accuracy. In this paper we are going present a review of some best lung nodule existing detection and segmentation techniques. Finally, we conclude by focusing one of the best methods that may have high level accuracy and can be used in detection of lung very small nodules accurately. Keywords : Computer aided diagnosis, image segmentation, CT medical imagine Digital Image Processing I. INTRODUCTION Lung cancer is a one of the major cause of death. The types are small-cell lung cancer (SCLC) and non-small-cell lung cancer (NSCLC). Lung cancer is a state which consists of uncontrollable growth of cell and tissues. If left untreated, this growth can spread beyond the lung by process of metastasis into nearby tissue or other parts of the body. Such treatments involve surgery, chemotherapy and radiotherapy. Overall, 15% people diagnosed with lung cancer survive five year after the diagnosis. Lung Cancer Types Cancer that starts in the lung is called first stage lung cancer. Lung cancer include different types and these are divided into two main groups 1. Small cell lung cancer About 20 out of every 100 lung cancers diagnosed are this type. the small cells are mostly filled with the nucleus so it is called small cell lung cancer. The main reason for such type of cancer is smoking. Small cell lung cancer often spreads quite early on and so doctors often suggest chemotherapy treatment rather than surgery. 2. Non small cell lung cancer Non small cell lung cancer is of three type. These are grouped together because they behave in a similar way and respond to treatment in a different way to small cell lung cancer. The early detection of lung cancer can increase overall 5-year survival rates from 14 to 49% for patients. Hence, computer-aided diagnosis (CAD) systems using images processing are used to extract the presence of lung cancer cell in a CT-Scam images of patients. II. STEPS FOR LUNG NODULE DETECTION SYSTEM The entire diagnosis process of Lung Cancer Detection System is shown in Figure @IJMTER-2014, All rights Reserved 301

Image Capture Smoothing Enhancement Segmentation Morphologic Colorimetric Lung Cancer Cell Identification Diagnosis result In the image Pre-processing stage we started with image enhancement the aim of image enhancement is to improve the interpretability or perception of information in images for human viewers, or to provide `better' input for other automated image processing techniques. Image enhancement techniques can be divided into two broad categories: Spatial domain methods and frequency domain methods. In image enhancement stage we used three techniques: auto-enhancement, Gobor filter, and Fast Fourier transform techniques. Image segmentation is an essential process for most image analysis subsequent tasks. In many of the existing techniques for image description and recognition depend highly on the segmentation results. That used Thresholding and marker controlled watershed segmentation techniques. Thresholding is one of the most powerful tools for image segmentation. the advantages of Thresholding is fast processing speed, also the smaller storage space and ease in manipulation as compared with gray level image. Therefore, thresholding techniques have drawn a lot of attention during the past 20 years. The presence of objects or background at specific image locations is extracted by watershed.image Enhancement can define image enhancement to get improve the quality of image, because of which the resultant image is better than the original one, the process of improving the quality of a digitally stored image by manipulating the image with MATLAB software. It is very easy, to make an image lighter or darker, or to change contrast. Image enhancement techniques are of two types: 1. Spatial domain techniques, which operate directly on pixels. @IJMTER-2014, All rights Reserved 302

2. Frequency domain techniques, which operate on the Fourier transform of an image. Segmentation divides an image into its constituent part. The segmentation of medical images in 2D, has many useful applications for the medical professional: visualization and volume estimation of objects of interest, detection of Abnormalities (e.g. tumors, etc.), tissue quantification and are more. The goal of segmentation is to simplify and/or change the representation of an image into something that is more meaningful and easy to understand. Image segmentation is used to locate objects and boundaries (lines, curves, etc.) in images. Mostly, image segmentation is the process of assigning a label to every pixel in an image such that pixels with the same label share certain visual characteristics. Two approaches to predict the probability of lung cancer presence; first approach is Binarization and the second approach is masking, both of these methods are based on facts strongly related to lung anatomy and related information of lung CT imaging. Binarization Approach This approach depends on the number of black pixels is much more than white ones in a normal lung image, so we started to count the black pixels for normal and abnormal images to get the average which will be denoted later as threshold then each image black pixels will be compared to this threshold, whether it is greater, and then it will be normal, else the opposite. Masking Approach This method depends on the fact that the masses are appeared as white connected areas inside ROI (lungs) as they increase the percent of cancer presence increase. Combining the two approaches together will lead us to take a decision whether the case is normal or not. 3.4 Image Features Extraction The Image features Extraction stage is important in our working in image processing techniques which using algorithms and techniques to detect and isolate various desired portions or shapes (features) of an image. III. LUNG NODULE DETECTION TECHNIQUES Lung nodules detection is complicated, Nodules in CT images show up have relatively low contrast white circular shape and it is overlap with vessels, ribs, and shadows. A. Chest radiograph Chest radiography is the most commonly performed diagnostic imaging test for the diagnosis of many pulmonary diseases. It provides simplicity, low radiation, low cost, large amounts of information and wide availability are advantages of this technique. The most satisfactory routine radiographic views are the postero anterior (PA) and lateral projections with the person standing up. Lordotic projections, with the X-ray beam,can be used for an improved visibility of the lung apices, superior mediastinum and thoracicin let. Oblique studies are sometimes useful in locating a pleural or chest wall disease process. The most remarkable is the rapid conversion from film-based to digital radiographic (DR) systems. Advantages of DR systems are the high image quality and the potential for dose reduction together with a favorable cost benefit ratio and an increased efficiency. Post-processing techniques in digital radiography are temporal subtraction and dual-energy subtraction. @IJMTER-2014, All rights Reserved 303

B. Computed Tomography (CT) Computed tomography (CT) is the second most important imaging modality of the chest. Before1991, a conventional or incremental CT scan of the chest was performed and consisted of a series of cross-sectional slices obtained during suspended respiration (the stop-andgo method). In 1991, helical CT was introduced in the imaging of the chest and dramatically improved the quality of CT images of thoracic structures. Helical CT produces a singlevolumetric dataset within one breath-hold during continuous scanning while the patient is moved through the CT gantry. The X-ray beam copies a helical or spiral curve in relation to the patient.more recently, multi-detector CT (MDCT) scanners have been developed. These new CT scanners use multiple rows (4, 8, 16, 32, 64 or 128 rows) of detectors. The MDCT have advantages as revolutionised the diagnostic approach to lung cancer. there is an increased temporal resolution. Data acquisition is so rapid that scanning of the entire lung can be performed within a single breath-hold. Because of to better temporal resolution, tit also provide a better contrast material administration possible. Secondly, there is an more spatial resolution.continuous acquisition of thin slices allows the increase in improvement of the image quality of multiplanar reconstruction (MPR) images. C. CAD System Computer-aided diagnosis, generally consist two main stages: selection of candidate nodule and second elimination of false positive nodules. Initially, Enhancement techniques used to enhance the nodule to separate them from background and it also lead to suppress other false structures. Usually, thresholding is used to separate candidate nodules from the background. For different shaped and sized nodules can be detected by using different techniques. such as 3D Cylindrical and spherical filters are used to detect small nodules, and circular and semicircular nodules can be detected by template matching whereas pattern-recognition techniques such as clustering, linear discriminate functions etc. are used in the detection of lung nodules. D. Magnetic Imaging Resonance (MRI) The lung remains a difficult organ for MR because of several limitations: the high susceptibility of MR to motion artifacts (i.e. pulsation and breathing), the intrinsic low proton density of lung parenchyma and the decrease in signal intensity due to air soft tissue interfaces. That is the reason that MR was considered for many years as a useful problemsolving technique for specific instances when used in addition to CT : 1) Detection of tumor invasion in the chest wall and the meditational structures (pancoast tumor) 2) Comparison between solid and vascular hilar masses 3) Assessment of abnormalities 4) the study and follow-up of mediastinal lymphoma. New applications, like whole-body MR (WBMR) imaging, may significantly increase MR-sensitivityin the near future. E. High Resolution Computed Tomography (HRCT) High-resolution computed tomography (HRCT) imaging of the lungs is well-established for evaluating many diseases. Optimal methods of acquisition and interpretation of highresolution images require knowledge of anatomy and pathophysiology, as well as familiarity with the basic physics and techniques of computed tomography. This parameter gives the principles for performing high-quality HRCT of the lungs. HRCT is the use of thin section CT images (0.625 to 2 mm slice thickness) often with a high-spatial-frequency reconstruction algorithm to detect and characterize disease affecting parenchyma and airways. Referring the development and widespread availability of multi-detector CT scanners capable of acquiring @IJMTER-2014, All rights Reserved 304

data throughout the entire thorax in a single breath-hold, there are 2 general approaches available for acquiring HRCT images. IV. SEGMENTATION TECHNIQUE The accurate segmentation of lung nodules is very important and crucial. Best segmentation makes physicians life easy. It plays a crucial part in proper diagnosis and treatment procedure for lung cancer. The segmentation accuracy directly affects many factors, such as the malignancy classification of lung nodules in CAD x for feature extraction. In section of paper we review various segmentation techniques for lung nodules from CT images. A. Thresholding (TH) Thresholding is the most simple and popular image segmentation technique. It used to generate binary image by labeling each voxel by testing whether its intensity greater or less than given intensity value. The automatic threshold determination was proposed by using K- mean clustering and average gradient and boundary compactness. The CT scan can be separated into two types of voxels, characterized by the density differences between the two anatomical structures. The high-density regions primarily consist of the body surrounding the lung cavity, whereas the low-density regions contain the lung cavity, the air surrounding the body, and other low-intensity regions. B. Mathematical Morphology (MM) MM is a most popular segmentation method to process binary and gray scaled images. MM is a theory and technique for the analysis and processing of geometrical structures. Is an effective method for binary morphological filtering with various combinations of these basic operations. Mathematical morphology is a tool for extracting image components. The operations of MM are originally defined as set operations and are used to extract the edges of an image, to filter an image and to skeletonise an image. Here we have used this for filtering the false positives from the detected nodules. C. Region Growing (RG) Region growing also classified as a pixel-based image segmentation method since it involves the of initial seed points. It start with a seed pixel, the initial region begins as the exact location of seeds points. The regions are then grown from these seed points to adjacent points depending on certain criteria. This is an iteratively grown by examining the adjacent pixels of seed points. If the pixels have the same intensity value with the seed points, it classifies them into the seed points. The deference between pixels intensity and the regions mean is used to classify the similarity of the image into regions. It is an iterated process until there are no changes in two successive iterative stages. There are more recent studies on this algorithm that have extended its approach as main component of their segmentation algorithm. Proposed a region growing method by using fusion-segregation criteria using geodesic distances. D. Graph Cut and Watersheds Graph cut and Watersheds are both well-known as standard image segmentation techniques. Goodman used Watersheds in their study. Watersheds semiautomatic used to first segment each nodule and then by a model-based shape analysis used to determine anatomical characteristics of all type of nodules. @IJMTER-2014, All rights Reserved 305

Marker-driven watershed segmentation technique extracts seeds that indicate the presence of objects or background at specific locations. Marker locations are then set to be regional minima within the topological surface (typically, the gradient of the original input image), and the watershed algorithm is used. Separating touching objects in an image is one of the most difficult image processing operations, where the watershed transform is often used to solve such problem. Marker-controlled watershed concept has two types: External associated with the background and Internal associated with the objects of interest. Image Segmentation using some watershed transforms works well if we can identify or mark foreground objects and background locations, to find catchment basins and watershed ridge lines in an image by treating it as a surface where light pixels are high and dark pixels are low. Figure shows a segmented image by watershed V. ADVANTAGES AND DISADVANTAGES Table 1: Advantages and Disadvantages Techniques Advantage Disadvantage Chest-Radiography It is simple. It cannot detect invasion of chest It has low cost. wall. It has low radiation. It provides large amount of information. Diaphragm and mediastrinum and nodal involvement. CT (Computed Tomography) MRI(Magnetic Resonance Imaging) CAD (Computer Aided Diagnosis) System It produces volumetric dataset with one breathhold. It has revolutionised the diagnostic approach to lung cancer. It is used for scanning and detection of soft tissue. There is no involvement of any kind of radiation in MRI. It can provide information about blood circulation. It helps the radiologist to improve sensitivity. It gives large amount of information. It is used for early detection of lung cancer. It has high cost. It deliver high dose of radiation. It promotes lazziness. MRI scan is done in enclosed space. MRI scan involves loud noise. It is expensive.. Still it produce large number of false positive. Efficiency is reduced. @IJMTER-2014, All rights Reserved 306

HRCT (High Resolution Computed Tomography) It is possible to take thin section. It can be reconstructed well It expose excessive radiation. VI. CONCLUSION The techniques described have been very popular among commercial semiautomatic softer packages and used into the medical practice. The review tells that all the proposed methods from different research have different level accuracy in different areas. Region growing algorithm is the best know algorithm in the field of image segmentation and also most useful in lung nodule detection. Many studies for lung nodules have been going on. And the present challenges, in this field, suggested that the search is more effective and accurate CAD for lung cancer detection will remain an active research area. REFRENCES [1]. Sunil kumar, Ashok kumar, Sanjay bhargava, Lung nodule detection and segmentation techniques review, International Journal of Advanced Research Computer Science And Software Engineering, 102-104, April 2014. [2]. Nidhi, Shilpa Mehta, Review of existing Techniques of Lung Nodule Cancer Detection and Existing Algorithms That Can Be Used For Efficient Detection In Future. International Journal of Emerging Science And Software Engineering, 1-3, February 2014 [3].Mr. Vaibhav K. Likhitkar1 Prof. Ujwalla Gawande2 Mr. Kamal,Hajari, Automated Detection of Cancerous Lung Nodule from the Computed Tomography Images,IOSR Journal of Computer Engineering, 5-11, February 2014. [4].Ella A. Kazerooni, MD, FACR, Chair Phillip M. Boiselle, MD Lynn S. Broderick, MD, FACR H. Page McAdams, MD Paul L. Molina, MD, FACR Reginald F. Munden, MD, DMD, MBA, FACR David P. Naidich, MD Robert D. Tarver, MD, FACR, Acr Practice Parameter For The Performance Of Highresolution Computed Tomography (HRCT) Of The Lungsin Adults, 1-7, Amended 2014 [5]. M. Gomathi and P. Thangaraj, A Computer Aided Diagnosis System for Lung Cancer Detection \Using Support Vector Machine,American Journal of Applied Sciences 7, 1532-1538, 2010 [6].W. De Wever, J.Coolen, J.A. Verschalelen, Imagng techniques n lung cancer,339-346, June 2011. [7].Mokheld S.AL-TARAWNEH, Lung Cancer Detection Using Image Processing Techniques, Leonardo Electronic Journal of Practices and Technologies, 147-158, January-June 2012. [8].Anita Chaudhary, Sonit Sukhraj Singh, Lung Cancer Detection Using Digital Image P, IJREAS, 1351-1359, February 2012. @IJMTER-2014, All rights Reserved 307