Detection and identification of lung cancer using SVM technique
Data science projects in pondicherry
Create New

Detection and identification of lung cancer using SVM technique

Project period

06/26/2020 - 06/30/2020




Detection and identification of lung cancer using SVM technique
Detection and identification of lung cancer using SVM technique

Lung cancer growth has turned out to be a standout amongst the most widely recognized reasons for disease in the two people. Countless bite the dust each year because of lung malignancy. The illness has diverse stages whereby it begins from the little tissue and spreads all through the distinctive territories of the lungs by a procedure called metastasis. It is the uncontrolled development of undesirable cells in the lungs. It is assessed that around 12,203 people had lung disease in 2016, 7130 guys and 5073 females; passing from lung malignant growth in 2016 were 8839.

Biomedical image handling is the most recent rising apparatus in medicinal research utilized for the early recognition of malignancies. Biomedical image handling strategies can be utilized in the restorative field to analyze maladies at the beginning time. It utilizes biomedical images, for example, X-beams,  Computed innovation and MRIs. The principal commitment of image handling in the restorative field is to analyze malignant growth at the beginning time, expanding survival rates. The time factor is basic for tumors of the mind, the lungs, and bosoms. Image handling can identify these malignant growths in the early periods of the maladies encouraging an early treatment process. The image preparing procedure comprises four essential stages, pre-handling, division, including extraction and grouping. This project presents image preparing procedures whereby the  CT examine image is utilized as information image, and beginning period lung disease is distinguished utilizing an SVM (Support vector machine) calculation as a classifier within the grouping stage to boost exactness, affectability, and clarity. First, the image is pre-handled and divided. After that Features are removed from the sectioned image lastly the image is delegated ordinary or destructive.

Why: Problem statement

Lung cancer is the most dangerous disease in the world, nowadays. So the early detection of lung cancer is necessary. Lung cancer is the leading cancer killer of both men and women in the world. It has one of the lowest five-year survival rates of all the cancer types. One reason why lung cancer is therefore deadly is that it is hard to find in its early stages. It may take years for the lung cancer to grow and there sometimes aren’t any symptoms early on.  By the time you begin to note symptoms, the cancer usually has spread to other parts of the body. Lung cancer causes the foremost cancer deaths worldwide, accounting for 2.1 million new lung cancer cases and 1.8 million deaths annually. Approximately 541,000 Americans living nowaday have ever been diagnosed with lung cancer. Lung cancer is deadly. More than half of the individuals with lung cancer die inside one year of being diagnosed. Th calculated 154,040 Americans are expected to die from lung cancer in the year of 2018, accounting for approximately 25 percent of all cancer deaths.

How: Solution description

This project explains the Lung cancer detection system which uses image processing techniques using the techniques of Image processing. This system will be capable of medical images such as CT, MRI and ultrasound images. This proposed model is developed using PSO, Genetic Optimisation and SVM algorithm used for feature selection and classification. This model is just like an extension of image processing using lung cancer detection method and produces the results. This system accepts any one of the medical images consisting of MRI, CT and Ultrasound images as input. Once preprocessing of an image is done, a canny filter is used for Edge detection. This project proposes a method which detects the cancerous cells accurately from the CT, MRI scan and Ultrasound images. 

Accurate diagnosis for different types of cancer plays an important role in determining and choosing the proper treatment to the doctors to assist them. By using classification techniques, possible errors that might occur due to unskilled doctors can be minimized. The medical salespersons make this project of much greater significance. Since symptoms appear only in the advanced stages thereby causing the mortality rate of lung cancer to be the highest among all other types of cancer, challenging the detection of cancer in its early stages. The objective of undertaking this project is to facilitate doctors to provide the best possible treatment by providing useful insights with the help of predictive models through analysis and diagnosis of lung cancer treatments.

The proposed system of detection of lung cancer starts with the collection of CT images followed by the proposed methodology.

How is it different from competition

We identified the stages of cancer by measuring the size and dimensions of the cancer module. But by using the SVM technique, it classified the stages of cancer and it is used to identify in the initial stage itself. So that major problems can be avoided. This project work proposes a technique to notice the cancerous cells effectively from the CT, MRI scan and Ultrasound images. Super pixel Segmentation has been used for segmentation and Gabor filter is used for Denoising the medical images. Simulation result is obtained for the cancer detection system using MATLAB and comparison is done between the three medical images. This method is given more accuracy to predict lung cancer at the early stage using SVM (Support Vector Machine) technique.

Who are your customers

The main purpose of this project is developed for lung cancer patients. This project is used for predicting lung cancer at the early stage. Similar challenges also exist for both physicians and patients alike when it comes to the issues of lung cancer prevention and lung cancer susceptibility prediction. The detection and identification of lung cancer using SVM technique helps to reduce the lung cancer death rate.

Project Phases and Schedule

Phase 1:  First step of this project is data collection. We collect the CT, MRI scan and Ultrasound images.

Phase 2: Pre-processing (CLAHE) - The CT images obtained as input images are converted into gray scale images. The intensity values are different in the input image, hence CLAHE is applied to equalize the image.

Phase 3: Segmentation (MFPCM) - Segmentation is a technique of unsupervised classification that arranges patterns in the clusters or regions. The first method is to partition the image based on abrupt changes in intensity, such as edges in an image. The second method is based on partitioning the image into regions that are similar according to a predefined criterion.

Phase 4:  Feature extraction and selection - Feature extraction is a crucial step for the CAD system. It uses different methods and algorithms for feature extraction from the segmented image. The extracted ROI can be classified as either cancerous or not using their texture properties.

Phase 5:  Classification - Next phase in the proposed system is the classification of occurrence and non-occurrence of cancer nodules for the supplied lung image. The classifier used here is Support Vector Machine(SVM).

Phase 6: ROI Extraction - The final phase in the proposed system is the extracting of the affected region using median filtering, dilation and erosion.

Phase 7: Result - we have used SVM (Support vector machine) classifier. This classifier is trained first. Here, two classes (class1 as abnormal  and class 2 as normal) SVM classifier is used. An SVM classifies data to separate all data points of one class from those of the other class, by finding the best result.

Resources Required

By using Matlab - MATLAB is a multi-paradigm numerical computing environment and proprietary programming language developed by MathWorks. MATLAB allows matrix manipulations and functions, plotting of data, implementation of algorithms, user interfaces creation, and interfacing with programs written in other languages.

Interface Matlab to Autocad - MATLAB does not provide a direct interface to AutoCAD. It always supports communicating with other applications. We can also use ActiveX to interface to AutoCAD from within MATLAB.


Leave a Comment

Post a Comment

Are you Interested in this project?

Do you need help with a similar project? We can guide you. Please Click the Contact Us button.

Contact Us

Social Sharing