I INTRODUCTION

The cervical cancer can be viewed by the ultrasound scan (or) MRI scans. The ultrasound scanned image is taken for the entire process. The ultrasound scan is more comfortable than MRI scan for diagnosis. It is not affect the human body because it does not use a radiation. Ultrasound doesn’t emit radiation. There are different types of algorithms were developed for uterine cancer detection. An ultrasound device uses sound waves that can’t be heard by humans. The sound waves make a pattern of echoes as the bounce off organs inside the pelvis. Echoes’ create a picture of uterus. Uterine cancer is also known as endometrial cancer. The two main types of endometrial cancer are:

[1] Endometroid adenocarcinoma

[2] Uterine carcinosarcoma

Endometroid Adenocarcinoma: This type of uterine cancer forms in the glandular cells of uterine lining, it accounts for as much as 75 present of all uterine cancers. Endometroid adenocarcinoma commonly detected early and as a high cure rate.

Uterine Carcinosarcoma: This rate form uterine cancer felt to be an endometrial cancer. It has elements of both adenocarcinoma and sarcoma. These cancer have a high risk of spread to lymph nodes and other parts of the body.

Ultrasound or ultra sonography is a medical imaging technique. Ultrasound is used in medicine to view internal organs of the body. It can also use to locate objects. Ultrasound scans are used to detect problems in the liver, heart and kidney (or) uterine. That uses high frequency sound waves and their echoes. The sound waves travel in to our body and a boundary between tissues. The most noticeable advantages of ultrasound scanning are safety, cost effectiveness, speed, easy handling and portability. Ultrasound is based on the principle of sound wave echoes. Sound wave travels from the probe to the object, passes through it and is continuously reflected back to the probe from multiple points inside the object. Ultrasound involves sound wave of frequency in the range of megahertz; typically this ranges between 3.5-10 MHz the machine display the two dimensional image. The quality of ultrasound images is limited by granular speckle noise. This makes it difficult to segment the images. The image is pre-processed in order to crop it such that unnecessary portions as well as the speckle noise are removed. We use a morphological cleaning algorithm to clean the image. Then it is segmented by an algorithm which uses a collection of rules to locate the cancer and morphological concepts to segment the image.

II RELATED RESEARCH WORK

MRI stands for Magnetic Resonance Imaging. The scan uses radio waves and very strong magnets to produce images of soft tissues. No x-rays are used. Magnetic resonance imaging (MRI) devices can scan the inside of the body in detail, to spot even the earliest signs of cancer or other abnormalities [3, 7]. It uses a magnetic field and radio waves to take pictures inside the body. It is especially helpful to collect pictures of soft tissue such as organs and muscles that don‘t show up on x-ray examinations [2, 6]. MRI is accurate in diagnosing a fibroid with a sensitivity of 88%–93% and a specificity of 66%– 91% [17]. One way to think of an MRI scan is a water x-ray (although no actual x-rays are involved). Normal x-rays image calcium, so they are good to see bones.MRI scans image water, which makes them very useful because all tissues of the body contain various amounts of water. This allows high resolution pictures of many organs and tissues to be taken that are invisible to standard x-rays. Magnetic Resonance Imaging (MRI) is an advanced medical imaging technique used to produce high resolution images of the parts contained in the human body. MRI imaging is often used when treating brain tumors. These high resolution images are used to examine human brain development and discover abnormalities. Different images either infectious or non-harmful are acquired from (MRI). The images that are acquired are completely unprocessed. The images can be in the form of JPG, PNG formats etc.

Uterine Fibroid Investigation System, the following [4, 5] are the steps implemented: Image Pre-processing: Pre-processing technique consists of two steps, namely cropping the image and removing the noise. Several filters are used for noise removal that is well established in gray value image processing.

Segmentation: Segmentation is the next step applied on the pre-processed image. Image segmentation is an important process to extract information from complex medical images. The segmentation component is expressed in terms of production rules, which capture the perceptual knowledge relevant to the domain. These rules will detect appropriate boundaries.

Feature Extraction: The goal of feature extraction is to obtain representative features that can be used to determine the mode of treatment. The shape-based features, i.e. diameter, area and compactness, are the essential measures.

Diameter: It is measured and specified by the radiologist as horizontal and vertical diameters in centimetres. After segmenting the fibroid, the horizontal and vertical diameters are calculated in pixels and converted to centimetres.

Area: The area of the fibroid is the number of pixels inside the fibroid. It is used to specify the size of the fibroid. It can also be found by the formula A= pi*a*b/2

Compactness: The compactness of a region is defined as the ratio of the square of the perimeter of a region to its area, i.e. compactness = perimeter2/area, where the perimeter is the distance around the boundary of the fibroid and the area is the number of pixels inside the fibroid.

Magnetic Resonance Imaging (MRI) is an advanced medical imaging technique used to produce high resolution images of the parts contained in the human body. MRI imaging is often used when treating brain tumors. These high resolution images are used to examine human brain development and discover abnormalities. MRI scan is done in an enclosed space, so the people who are claustrophobic, i.e. fearful of being in a closely enclosed surface, are facing problems with MRI to be done. RI scans involve really loud noises wile processing because they involve a really big amount of electric current supply MRI scanners are usually expensive.

III. PROJECT DESCRIPTION

Ultrasound scan makes use of sound waves to generate images of the internal organs of the body. A gel will be applied on the abdomen to give a good contact for the probe. The radiologist who performs the scan will move the probe over the abdomen and the image will be displayed on the screen. The scan is completely painless. The scan usually takes 10 to 15 minutes and you may be asked to wear a gown. The ultrasound scanner looks a bit like a home computer system. There is a hard-drive, keyboard and a display screen, and a hand-held sensor. The sensor sends out sound waves and picks up the returning echoes. Pictures of the inside of your body are displayed on the screen. These pictures are constantly updated, so the scan can show movement. The sound waves travel at different speeds depending on the type of tissue. An ultrasound gel is placed on the transducer and the skin to allow for smooth movement of the transducer over the skin and to eliminate air between the skin and the transducer for the best sound conduction.

Ultrasound images muscle and soft tissue very well and is particularly useful for delineating the interfaces between solid and fluid-filled spaces. It renders "live" images, where the operator can dynamically select the most useful section for diagnosing and documenting changes, often enabling rapid diagnosis. It has no known long-term side effects and rarely causes any discomfort to the patient, and is relatively inexpensive compared to other imaging modalities. A fundamental principle of ultrasound imaging is that there is trade-off between penetration and resolution: The lower 0frequencies (3.5 MHz for abdominal scanning), the better the penetration (up to 15 cm), but the poorer the resolution (several mm.) Conversely, the higher the frequency (10 MHz for carotid artery imaging) the less the penetration (only a few cm) but the better the resolution (less than an mm). Here proposed ultrasound imaging technique to detect cancer. Uterine cancer image captured from the ultrasound scan.

Pre-processing: Image pre-processing, also called image restoration, and involves the correction of distortion, degradation, and noise introduced during the imaging process. This process produces a corrected image that is as close as possible, both geometrically and radiometrically, to the radiant energy characteristics of the original scene. Radiometric and geometric are the most common types of errors encountered in remotely sensed image. Image pre-processing use the redundancy in images. Image pre-processing some filters are using to remove the noise. Pre-Processing technique consists of two steps, namely cropping the image and removing the noise.

Fig.1 Uterine cancer investigation system

Image segmentation: Segmentation is the next step applied on the pre-processed image. Image segmentation is an important process to extract information from complex medical image. The canny edge detector first smoothens the image to eliminate noise.Then it finds the image gradient to highlight regions with high spatial derivatives. After that it perform tracking along these regions and suppresses any pixel that is not at the maximum. The gradient array at this moment can further be reduced by hysteresis which is used to track along the remaining pixels that have not been suppressed. Hysteresis uses two thresholds and if the magnitude is below the first threshold, it is set to zero. Canny Edge Detection Technique: The following shows the canny edge detection algorithm steps. The algorithm runs in 5 separate steps.

1. Smoothing: Blurring of the image to remove noise.

2. Finding gradients: The edges should be marked where the gradients of the image have large magnitudes. The gradient magnitudes (also known as the edge strengths) can then be determined as a Euclidean distance measure by applying the law of Pythagoras.

3. Non-maximum suppression: Only local maxima should be marked as edges. Double thresholding: Potential edges are determined by thresholding.

4. Edge tracking by hysteresis: Final edges are determined by suppressing all edges that are not connected to a very certain (strong) edge.

Feature Extraction: The goal of feature extraction is to obtain representative features that can be used to determine the mode of treatment. The shape-based features, i.e. abnormality, intensity and detected ratio are the essential measures.

Classification: Uterine cancer stages classified in to 3 types they are 1. Normal, 2. Pre-cancer, 3. Cancer

Normal: Uterine cancer have cancer that has not spread outside the uterus. This cancer confined to the inner layer of cells of the uterus (endometrium).

Pre-cancer: uterine cancer that invades less than one half of the muscle wall of the uterus.

Cancer: uterine cancer that invades more than one half of the muscle wall of the uterus.

IV EXPERIMENTAL RESULTS

Uterine cancer image captured from the ultrasound scan. Read the original images from ultrasound system. Morphological image cleaning (MIC) algorithm using these structuring elements of operations such as opening-closing, closing-opening is done to remove speckle noise.

The edges should be marked where the gradients of the image have large magnitudes. The gradient magnitudes (also known as the edge strengths) only local maxima should be marked as edges. Double thresholding Potential edges are determined by thresholding. Edge tracking by hysteresis: Final edges are determined by suppressing all edges thatare not connected to a very certain (strong) edge. The edge-pixels remaining after the non-maximum suppression step are (still) marked with their strength pixel by-pixel. Many of these will probably be true edges in the image, but some maybe caused by noise or colour variations for instance due to rough surfaces. The simplest way to discern between these would be to use a threshold, so that only edges stronger that a certain value would be preserved. The Canny edge detection algorithm uses double thresholding.

Fig .2 Input image

Fig. 3 MIC image

Fig. 4 canny edge image

V CONCLUSION

The ultrasound methods are used in medical image processing. It can be used to give more information about cancer detection, segmentation and classification. The target area is segmented and the evaluation of this tool from the doctor and this helps the doctors in diagnosis, the treatment plan making and state of the cancer monitoring. In future, the system can be improved by adapting more segmentation algorithm to suit the different medical image segmentation. Using colour based image segmentation. It is possible to reduce the computational cost avoiding feature calculation for every pixel in the image.

VI FUTURE WORKS

In future, we can classify ultrasound uterine cancer in to three stages such as normal, pre-cancer and cancer. Normal stage uterine cancer have cancer that as not spread outside the uterus. This cancer confined to the inner layer of cell of the uterus (endometrium). Pre-cancer stage uterine cancer that invades less than one half of the muscle wall of the uterus. Cancer stage uterine cancer that invades more than one half of the muscle wall of the uterus. These three stages will be mentioned as histogram equalize graph.

REFERENCES

[1] Ajala Funmilola A, Oke O.A, Adedeji T.O, Alade O.M, Adewusi E.A (2014). ‘Fuzzy k-c-means Clustering Algorithm for Medical Image Segmentation’, vol 2, pp 271-286.

[2] Dilliraj.E, P.Vadivu, A.Anbarasi (2014), ‘MR Brain Image Segmentation Based on Self- Organizing Map and Neural Network’, International Journal for Research in Applied Science & Engineering Technology (IJRASET), Vol 2 , pp 2321-9653.

[3] Hayit Greenspan , Shiri Gordon, Gali Zimmerman, Shelly Lotenberg, Jose Jeronimo, Sameer Antani, and Rodney Long (2009), ‘Automatic Detection of Anatomical Landmarks in Uterine Cervix Images’, IEEE TRANSACTIONS ON MEDICAL IMAGING, vol. 28, pp 304-315.

[4] Krishna Kant Singh, Akansha Singh (2010), ‘A Study of Image Segmentation Algorithms for Different Types of Images’, IJCSI, Vol. 7, pp 414-417.

[5] Mahesh Yambal1, Hitesh Gupta2 (2014), ‘Image Segmentation using Fuzzy C Means Clustering: A survey’, Vol. 2, pp 74-84.

[6] Narendran.P, V.K. Narendira Kumar, K. Somasundaram, (2012), ‘3D Brain Tumors and Internal Brain Structures Segmentation in MR Images’, I.J. Image, Graphics and Signal Processing, pp 35-48.

[7] Neofytou, V. Tanos, I. Constantinou, E. C. Kyriacou, M. S. Pattichis, and C. S. Pattichis (2015), ‘Computer-Aided Diagnosis in Hysteroscopic Imaging’, IEEE journal of biomedical and health informatics, vol. 19, pp 43-54.

[8] Ratha Jeyalakshmi and Ramar Kadarkarai (2010),‘Segmentation and feature extraction of fluid-filled uterine fibroid–A knowledge-based approach’, Maejo International Journal of Science and Technology, pp. 1905-7873.

[9] Sudipta Roy, Atanu Saha and Prof. Samir K. Bandyopadhyay (2014), ‘Brain Tumor Segmentation and Quantification from MRI of Brain’, Journal of Global Research in Computer Science, Vol 2, pp 274-291.

[10] Snehal N. Patil; Prof. Dr. K.V.Kulhalli; Smita S. Patil (2014) , ‘ Endometrial Cancer Detection using Image Processing Techniques’, International Journal of Software & Hardware Research in Engineering. Vol 2, pp 2347-4890.

[11] Ortiz. A, J.M Gorriz, J. Ramirez, (2013), ‘Two Fully-Unsupervised Methods for MRI Brain Images Segmentation Using SOM Based Strategies’. pp 2668-2686.

[12] Vinita Salvi (2012), ‘Embolization in fibroids’, The Journal of Obstetrics and Gynecology of India Vol.55, pp 410-414.

[13] Ziqian Zhang, Tiexiang Wen,Jia Gu, Lei Wang (2010), ‘Cancer area extraction for uterine cervical image based on spectral matting’ , IEEE, Vol.3, pp 1351 - 1354.