Computer-aided diagnosis in thoracic CT

Abstract

ABSTRACT Computer-aided diagnosis (CAD) provides a computer output as a “second opinion” in order to assist radiologists in the diagnosis of various diseases on medical images. Currently, a hot topic in CAD is the development of computerized schemes for detection of lung abnormalities, such as lung nodule and interstitial lung disease, in computed tomography (CT) images. The author describes in this article the curren t status of the CAD schemes for the detection of lung nodules and interstitial lung disease in CT developed by the author and his colleagues at the Univ ersity of Chicago and Duke University. Keywords: nodule, interstitial lung disease, computer-aided diagnosis, detection, CT 1. INTRODUCTION Computer-aided diagnosis (CAD) has become one of the major research topics in medical imaging and diagnostic radiology, and has been applied to various medical imaging modalities including computed tomography (CT), magnetic resonance imaging, ultrasound imaging, and nuclear medicine [1-4]. CAD is especially useful for volumetric imaging modalities such as CT , because a thin-section CT scan incl udes hundreds of s ections and thus requires considerable time and effort fo r radiologists to interpret CT images. One of the most important applications of CAD is the detection of lung cancer, because lung cancer is the leading cause of cancer-related deaths in the U.S. In fact, the total number of deaths caused by lung cancer is greater than that resulting from colon, breast, and prostate cancers combined [5]. Theref ore, many investigators have attempte d recently to develop CAD schemes for detection of lung cancer in thin-section CT [6-16], which is one of the topics of this article. Please see [17] for details regarding current status of lung nodule detection in thin-section CT. Interstitial lung disease (ILD) is a common abnormality [18] that occurs in lung parenchyma and can cause fatal illnesses [19, 20]. The correct diagnosis of ILD is one of the most difficult tasks for radiologists, because the contrast of ILD lesions is often low and the disease patterns are very complex [21]. It is for this reason that CAD schemes have been developed for assisting radiologists in the detection and diagnosis of interstitial lung disease [22]. The existing CAD schemes generally employed statistical texture features [23-34], Fourier transform-based texture features [35-39], geometric features [38,40,41], fr actal features [42,43], and rule-based expert features [44] for detection and diagnosis of interstitial lung diseases. This paper describes the current status of the CAD schemes for the detection of lung nodul es and interstitial lung disease in thin-section CT developed by the author and his colleague s at the University of Chicago and Duke University.