Receiver Operating Characteristic (ROC) curve is considered to be the only scientific method to evaluate the effects of the system performance on the diagnostic outcome [26].

ROC curve is based on the statistical decision theory and made by plotting the conditional probability of true positive responses by an observer in a detection experiment where signal is detected from noise, versus the conditional probability of false positive responses. The area under the ROC curve represents diagnostic accuracy in evaluation of radiographic systems. Decision criteria in radiographic caries diagnosis were constructed using ROC curve method [27]. Now ROC curve test is widely applied to evaluate system performance or to compare diagnostic performance of different systems [28], [29] and [30]. Demerit of this ROC test is that it is time-consuming Dasatinib clinical trial and elaborate. One may notice similarity of N, maximum contrast information content obtained

from the PC (see Eq. (1)) to the area under the ROC curve which represents image information content per observation. However, selleck products there is a fundamental difference between these two methods. PC test is just a detection task of the prepositioned signals and include no false positive responses, while ROC curve test includes both false and true positive responses. PC test only represents psychophysical properties of the radiographic system, while ROC curve test represents overall system performance including psychological and

nosological phases. Therefore, PC test result may be a part of ROC curve test results. If we recognized the relationship between PC test and ROC curve test, experimental setting could be simplified, and the PC test results might be extrapolated to ROC curve test results. As described in Section 1, radiological diagnostic process consists of three phases, and effects of psychological and nosological phases may be important in medical diagnostic tasks. In the radiological diagnosis for approximal caries, the location of the abnormality is confined to the proximal surfaces and diagnostic task is to detect presence of the abnormality and to evaluate the degree of Oxalosuccinic acid abnormality, namely the depth of carious cavity. In this context, there is a possibility to clarify the relationship between PC test and ROC curve test in radiographic diagnostic task for approximal caries. Li et al. reported that psychophysical properties can be improved by perceptual linearlization for attenuation and visual response using PC tests [22]. The same image processing method has been applied to the radiographs of the aluminum step phantom. Observer performance to detect low contrast details has been similarly improved [21]. This method has also significantly improved the diagnosis of approximal caries in digital radiographs [3].