Abstract—According to the opinions of specialized doctors, being able to accurately make diagnosis with both the regions and types of stroke is very important, and with these information available, only then medical treatments can be applied properly. However, for any delay treatment or misdiagnosis, it is very likely as the key attributed to the fatal death of the patients. Currently, there are a lot of researches on the development of many methods with application of MRI for diagnosis of brain stroke out there already. However, in this paper, CT image is used to diagnosis brain stroke. And, we use laws’ mask to extract texture feature on the CT image, and further to input them into our proposed iterative Fuzzy LDA (iFuzzyLDA) method for classification. Next, we will make classification of four different features (stroke, CSF, gray and white matter), as well as to put a link among them, in order to get better accuracy for diagnosis. Experimental result shows our method has a good accuracy, and be able to precisely help the doctors for marking out the regions of brain stroke. In addition, it can be verified that the accuracy rate of our method is up to 90% through object-level consistency error (OCE) method.

Index Terms—CT scan, laws’ mask, iFuzzyLDA, OCE

Cite: Lee Ming-Sian, Chen Chong-Guang Chin Chiun-Li, and Liu Shih-Hua, "Stroke Area Detection using Texture Feature and iFuzzyLDA Algorithm," International Journal of Electronics and Electrical Engineering, Vol. 1, No. 2, pp. 94-97, June 2013. doi: 10.12720/ijeee.1.2.94-97