عنوان مقاله
طبقه بندی عکسهای پزشکی با استفاده از شیوه منطق فوزی بر مبنای الگوریتم ژنتیکی
فهرست مطالب
چکیده
مقدمه
روش ها
کاربردها
نتیجه گیری
بخشی از مقاله
مجموعه داده
دراین برنامه کاربردی ، کلاً 90 نمونه از عکسهای اکوکاردیوگرافی از 45 فرد در بیمارستان دانشگاه پزشکی Gifu جمع آوری گردید. عکسهای جمع آوری شده با کمک ابزار Toshiba SSH-160A با مبدل2.5MHz بدست آمدند. به حالت پر شدن کامل بطن های قلب قبل از انقباض قلب end diastole گفته می شود. زمانی که بطن های قلب تخلیه شده باشند ، end systole نامیده می شود. بنابراین، یک چرخه قلبی با دو حالت مذکور معرفی می گردد تا بدین طریق شرایط قلبی مشخص و واضح گردد.
کلمات کلیدی:
Medical image classification using geneticalgorithm based fuzzy-logic approach Du-Yih Tsai Yongbum Lee Masaru Sekiya Masaki Ohkubo Niigata University Department of Radiological Technology School of Health Sciences, Faculty of Medicine 2-746 Asahimachi-dori Niigata City Niigata 951-8518, Japan E-mail: tsai@clg.niigata-u.ac.jp Abstract. In this paper we present a genetic-algorithm-based fuzzy-logic approach for computer-aided diagnosis scheme in medical imaging. The scheme is applied to discriminate myocardial heart disease from echocardiographic images and to detect and classify clustered microcalcifications from mammograms. Unlike the conventional types of membership functions such as trapezoid, triangle, S curve, and singleton used in fuzzy reasoning, Gaussiandistributed fuzzy membership functions (GDMFs) are employed in the present study. The GDMFs are initially generated using various texture-based features obtained from reference images. Subsequently the shapes of GDMFs are optimized by a genetic-algorithm learning process. After optimization, the classifier is used for disease discrimination. The results of our experiments are very promising. We achieve an average accuracy of 96% for myocardial heart disease and accuracy of 88.5% at 100% sensitivity level for microcalcification on mammograms. The results demonstrated that our proposed genetic-algorithm-based fuzzy-logic approach is an effective method for computer-aided diagnosis in disease classification. © 2004 SPIE and IS&T. [DOI: 10.1117/1.1786607] 1 Introduction Research in computer-aided diagnosis ~CAD! is a rapidly growing, dynamic field with new computer techniques, new imaging modalities, and new interpretation tasks. CAD is defined as a diagnosis made by a radiologist who uses the output from a computerized analysis of medical images as a second opinion in detecting lesions, assessing extent of disease, and making diagnostic decisions.1 So far most CAD papers have involved either mammograms2–10 or chest radiographs.11–17 Recent reports show that CAD research has extended to other fields such as echocardiography18 and colonography.19,20 In this paper, we present a generalized CAD scheme based on our previously reported CAD scheme.18 The proposed CAD scheme, containing four stages: image preprocessing, feature extraction, classifier training, and classifi- cation, can be applied to various imaging modalities and diseases with minor modification. In our system, we basically employ fuzzy logic for classification. Unlike the conventional types of fuzzy membership functions such as triangle and trapezoid, Gaussian-distributed membership functions ~GDMFs! are used in the system. The GDMFs are initially generated using various features obtained from image data sets. Subsequently, the shapes of the GDMFs are optimized using a genetic-algorithm ~GA! learning process. After optimization, the system is used for discrimination of disease. To our knowledge, this is the first time such a CAD system has been described using the GA-based fuzzy approach. In the present study, we apply our CAD method to discriminate myocardial heart disease from echocardiographic images and to detect and classify clustered microcalcification from mammograms. The performance of our CAD method is evaluated in terms of accuracy, sensitivity, and specificity. 2 Methods 2.1 Fuzzy Membership Functions and Fuzzy Rules The major components of the fuzzy-logic decision-making system are fuzzy sets, fuzzy membership functions, and fuzzy rules. Each fuzzy set has a corresponding fuzzy membership function. The value of the membership function ranges from 0 to 1 and can be considered a degree of truth. The current study uses simplified fuzzy rules as follows:
