Catur Edi Widodo

Detection Lung Cancer Using Gray Level Co-Occurrence Matrix (GLCM) and Back Propagation Neural Network Classification

Lung cancer prevalence is one of the highest of cancers, at 18 %. One of the first steps in lung cancer diagnosis is sampling of lung tissues or biopsy. These tissue samples are then microscopically analyzed. This procedure is taken once imaging tests indicate the presence of cancer cells in the chest. Lung cancer diagnosis using lung tissue sample microscopic analysis has some weakness. One of them is that doctor still relies on subjective visual observation. A medical specialist must do thorough observation and accurate analysis in detecting lung cancer in patients. Hence, there is need for a system that is capable for detecting lung cancer automatically from microscopic images of biopsy. This method will improve the accuracy and efficiency for lung cancer detection. The aim of this research is to design a lung cancer detection system based on analysis of microscopic image of biopsy using digital image processing. Microscopic images of biopsy are feature extracted with the Gray Level Co-Occurrence Matrix (GLCM) method and classified using back propagation neural network. This method is implemented to detection both normal and cancerous lung of biopsy samples. In the stage of training, 20 biopsy image samples were analyzed using back propagation neural network with 95% accuracy. On the other hand, 16 biopsy samples were analyzed during testing, with an accuracy of 81.25%. These results show that microscopic biopsy image processing can be implemented in a system of lung cancer detection.

Calculation of Lung Cancer Volume of Target Based on Thorax Computed Tomography Images using Active Contour Segmentation Method for Treatment Planning System

In this research, calculation process of the lung cancer volume of target based on computed tomography (CT) thorax images was done. Volume of the target calculation was done in purpose to treatment planning system in radiotherapy. The calculation of the target volume consists of gross tumor volume (GTV), clinical target volume (CTV), planning target volume (PTV) and organs at risk (OAR). The calculation of the target volume was done by adding the target area on each slices and then multiply the result with the slice thickness. Calculations of area using of digital image processing techniques with active contour segmentation method. This segmentation for contouring to obtain the target volume. The calculation of volume produced on each of the targets is 577.2 cm3 for GTV, 769.9 cm3 for CTV, 877.8 cm3 for PTV, 618.7 cm3 for OAR 1, 1,162 cm3 for OAR 2 right, and 1,597 cm3 for OAR 2 left. These values indicate that the image processing techniques developed can be implemented to calculate the lung cancer target volume based on CT thorax images. This research expected to help doctors and medical physicists in determining and contouring the target volume quickly and precisely.

CHEST X-RAY SEGMENTATION to CALCULATE PLEURAL EFFUSION INDEX in PATIENT with DENGUE HEMORRHAGIC FEVER

A study of the calculation of pleural effusion index (PEI) in patient with dengue hemorrhagic fever (DHF) has been conducted. PEI calculation was done through matlab programming language. Some digital image processing methods used in this research were thresholding segmentation, morphology operation, and calculation of pixel number per column in image to get PEI value. PEI values generated from image processing can be an alternative to replace the manual calculations done by physicians and used to demonstrate the gravity level of DHF. Keywords— Dengue hemorrhagic fever, Pleural effusion index, Segmentation.