Developing MATLAB-based application to detect position and size of lung nodule

A lung nodule is a small and round shaped growth in the lung found in the result of radiology examination. Lung nodules can be malignant or cancerous, the risk of lung cancer increases as the nodules get bigger. Lung cancer is the most dangerous cancer, it holds 18.4% of all cancer deaths. Therefore, it is essential to detect the position and size of lung nodules to make sure the presence of lung nodules that may lead to lung cancer. An application for detecting the position and size of lung nodules based on MATLAB has been developed. The application can perform image processing and it is begun from the stage of image quality enhancement by converting the image into grey-scale and median filter. Then it is followed by image segmentation process using Otsu thresholding and watershed. And the last is feature extraction using shape, size, and geometric extraction. The application has successfully detected the position and size of lung nodule based on image data of lung cancer from TCIA database. After being compared to dataset from TCIA database, the accuracy to detect the presence of lung nodules is 100%, while the accuracy to detect lung nodules’ position is 95%. Moreover, the application is also able to determine the lung nodule size.