Comparative evaluation of JPEG and JPEG2000 compression in quantitative digital subtraction radiography

OBJECTIVES: The aim of this in vitro study was to compare the impact of JPEG and the novel JPEG2000 compression standard on quantitative digital subtraction radiography (DSR) and to determine the acceptable JPEG2000 compression ratios for DSR. METHODS: Nine dry pig mandible sections were radiographed three times ('Baseline', 'No change', and 'Gain') with standardized projection geometry. Bone gain was simulated by adding artificial bone chips (1, 4 and 15 mg). Images were registered, compressed by JPEG and JPEG2000 using compression ratios (CR) of 1 : 7, 1 : 16, 1 : 22, and 1 : 31, and then subtracted. Image distortion was assessed objectively by calculating average pixel error and peak signal to noise ratio. No change areas in compressed and subtracted 'No change-Baseline' images and bone gain volumes in compressed and subtracted 'Gain-Baseline' images were calculated for both compression standards and compared. RESULTS: JPEG introduced less distortion at low CRs, while JPEG2000 was superior at higher CRs. At CR of 1 : 7, no significant difference between JPEG and JPEG2000 was found. JPEG2000 yielded better results for no change measurements at higher CRs. Volumes of simulated bone gain were overestimated when JPEG and underestimated when JPEG2000 compression was used. CONCLUSIONS: At CR of 1 : 7 JPEG and JPEG2000 performed similarly, which indicates that CR of 1:7 in JPEG2000 can be used for DSR if images are registered before compression. At higher CRs, JPEG2000 is superior to JPEG but image distortions are too high for reliable quantitative DSR.     

Color mammography. Image generation and receiver operating characteristic evaluation

Color mammography is a technique whereby dual-energy mammographic image data are used to calculate a calcium image; the calcium image is colorized and overlaid onto the conventional (lower energy) gray scale mammogram for radiologist viewing. This technique is presented as a practical way to use the increased calcium sensitivity of dual-energy mammography without requiring an increase in the number of images that the radiologist must read, and without subjecting the radiologist to the unfamiliar appearance of the dual-energy subtracted images. Using straightforward imaging theory, the acquisition techniques for both conventional and dual-energy mammography were optimized, and the optimal technique factors were used to generate a series of computer-simulated mammographic images that were used in a receiver operating characteristic (ROC) comparative study. Ideal observer ROC experiments indicate that (dual-energy) calcium images yield consistently higher sensitivity and specificity to the presence of calcifications, regardless of the amount of tissue "clutter," whereas conventional mammography results show degraded detectability performance as tissue contrast increases. Using human observers viewing simulated images, color mammography delivered greater calcification detectability than conventional mammography.     

A comparative evaluation of rare earth screen-film systems. Free-response operating characteristic analysis and anatomic criteria analysis

Diagnostic accuracy was evaluated in a signal detection experiment that used low-contrast acute lesions in a living dog model. At the high levels of certainty normally used by radiologists, rare earth screen-film systems provided accuracy comparable to that of a reference calcium tungstate screen-film system. Additionally subjective image quality evaluation for a given imaging task (clinical pediatric anterior-posterior chest films) based on visualization of anatomic landmarks and physical parameters has been conducted for several rare earth and one calcium tungstate screen-film system. The correspondence of subjective physical ranking with physical and psychophysical measurements was investigated. Evaluating the visualization of anatomic parameters can provide a clear and objective distinction among systems with comparable physical and psychophysical properties.     

Evaluation of Lossy data compression of chest X-rays: a receiver operating characteristic study

RATIONALE AND OBJECTIVES: To evaluate the quality of chest radiographs after 32:1 compression/decompression with different image compression algorithms. METHODS: Ten digital (Thoravison) radiographs of an anthropomorphic chest phantom with superimposed simulated nodular lesions (NL) and linear reticular lesions (LL) were obtained. Each radiograph was subdivided into 15 fields; they contained the lesions with a probability of 0.5. The radiographs were compressed and decompressed by using JPEG, fractal and wavelet algorithms at a compression rate of 32:1. Five radiologists evaluated the images. Data were analyzed with the receiver operating characteristic (ROC) method (comparison of area under curve). RESULTS: At 32:1 JPEG or wavelet compression, no statistically significant difference was observed for both NL and LL when compared with the original images. The fractal algorithm performed significantly lower for both NL and LL when compared with the original radiographs. CONCLUSION: The JPEG and wavelet image compression does not result in loss of relevant information for chest x-rays at a compression rate of 32:1.     

Quantitative assessment of quality of digitized portal images: receiver operating characteristic analysis applied to imaging in radiation therapy

An objective and quantitative method for the evaluation of the quality of megavoltage portal images was developed by applying receiver operating characteristic analysis. On the basis of therapeutic use of portal images, setup errors were employed as "signals" in this experimental study that compared the original portal films with digitized images. Six readers observed 104 portal images of a chest phantom, half of which were "abnormal" (ie, had setup errors). Digital images (2,048 x 2,048 matrix) were enhanced by means of histogram equalization and then printed with a laser printer for observation. The readers showed a higher discrimination capacity with the digitally enhanced images, although a statistically significant improvement was not demonstrated. The present method of assessment of image quality proved to be both simple and clinically reasonable.     

计算机X线摄影与屏-片成像系统的ROC曲线特性比较研究

目的研究计算机X线摄影(computed radiography，CR)、屏-片系统的ROC曲线特性并对它们进行比较。方法在中速屏．片系统特性曲线与CR系统总特性曲线基本一致的前提下，获得CR系统和中速屏．片系统ROC实验标准体模影像，然后变化后处理参数旋转量(GA)、密度变换(GS)、频率增强(RN)和频率等级(RE)，分别获得CR系统照片。请临床医师用5值判别法识别所有影像信号，并对结果进行ROC解析。结果(1)获得CR系统和中速屏-片系统影像的ROC曲线下面积(Az)分别为0．87556和0．93958，屏．片系统明显优于CR系统(t=2．526，P=0．0355)。(2)层次调节结果：①其他参数不变，调节GA=0．6、1．0,2．0、3．0、4．0时，CR系统影像的ROC曲线下Az分别为0．80024、0．87556、0．88184、0．88928和0．85944，调节GA，CR系统的ROC曲线下Az有所变换，但并不能达到或接近中速屏片系统。②其他参数不变，调节GS=0．3、0．4、0．5、0．6、0．7、0．8时，CR系统影像的ROC曲线下Az分别为0．77254、0．87556、0．92720、0．93356、0．94266和0．93512，调节GS，CR系统的ROC曲线下Az变换明显，可以达到或超过中速屏-片系统。(3)空间频率调节结果：①其他参数不变，调节RE=0．5、5．0、10．0、15．0时，CR系统影像的ROC曲线下Az分别为0．87556、0．92524、0．94084和0．94148，调节RE，CR系统的ROC曲线下Az可以达到或超过中速屏一片系统。②先调节RE=10，其他参数不变，调节RN=0、3、5、8时，CR系统影像的ROC曲线下Az分别为0．91438、0．94084、0．89518和0．86030，调节RN，CR系统的ROC曲线下Az可以达到或超过中速屏．片系统。结论(1)在CR系统总特性曲线与中速屏-片系统特性曲线一致并使用同样摄影条件时，CR系统影像上信息量明显小于中速屏-片系统。(2)只有恰当地选择CR的后处理参数获得的信息量才能超过屏-片系统。     

Routine receiver operating characteristic analysis in mammography as a measure of radiologists' performance

A system for calculating receiver operating characteristic (ROC) curves from routine audit data is described. Both diagnostic opinion and pathology outcome data from the symptomatic mammography department were recorded in an audit database, from which ROC curves were calculated. A comparison of overall radiologists' performance was made and the appropriate performance indices discussed. Similar data were collected from the breast screening centre where the same radiologists read mammograms. The radiologists' performance in the symptomatic department was compared with that in the breast screening centre and was found to be better in the screening centre. The difference could not be wholly attributed to the different age distribution of women and highlighted the different nature of the diagnostic task and the differing sampled populations. ROC curves were drawn for different clinical signs to determine any particular area of difficulty the radiologist may experience. ROC analysis as part of routine audit can be used for maintaining and improving the quality of an individual's performance, and targeting learning on areas of particular weakness.     

Virtual bronchoscopy versus thin section computed tomography in evaluation of moderate and low grade stenoses: receiver operating characteristic curve analysis

Purpose: To evaluate the impact of virtual bronchoscopy, under proper threshold settings, on observer level of confidence in the assessment of bronchial abnormalities producing stenoses ≤75% compared to interpretation of thin section computed tomography (CT) images.

Material and Methods: Sixty-five patients with fiberoptic bronchoscopy positive for tracheobronchial abnormalities were evaluated in a blinded observer study using a commercially available virtual endoscopy software package. The findings of virtual endoscopy were compared with those of fiberoptic bronchoscopy using receiver operating characteristic curves (ROCs) and other statistical tools.

Results: A total of 102 lesions were identified by fiberoptic bronchoscopy, with 44 of these producing bronchial stenoses ≤75%. Concerning the latter lesions, for virtual bronchoscopy the areas under the ROCs were 0.93 and 0.96 for the two observers, respectively, while for thin section CT the corresponding values were 0.86 and 0.88; the differences observed were statistically significant. Contrary to thin section CT, virtual bronchoscopy did not show statistically significant differences from fiberoptic bronchoscopy regarding estimation of degree of stenosis.

Conclusion: Virtual bronchoscopy under proper threshold settings has a statistically significant impact on observer performance where moderate and low-grade bronchial stenoses are concerned and gives an estimate of the degree of stenosis more precisely than thin section CT.     

Selection of a rating scale in receiver operating characteristic studies: some remaining issues

RATIONALE AND OBJECTIVES: The aim of this study is to compare the ratings of a group of readers that used two different rating scales in a receiver operating characteristic (ROC) study and to clarify some remaining issues when selecting a rating scale for such studies. MATERIALS AND METHODS: We reanalyzed a previously conducted ROC study in which readers used both a 5-point and a 101-point scale to identify abdominal masses in 95 cases. Summary statistics include the distribution of scores by reader for each of the rating scales, the proportion of tied scores when using the 5-point scale that correctly resolved when using the 101-point scale and the proportion of paired normal-abnormal cases where the two rating scales resulted in a different selection of an abnormal case. RESULTS: As a group, the readers used 84 of the rating categories when using the 101-point scale but the categories used differed for individual readers. All readers tended to resolve the majority of ties on the 5-point scale in favor of correct decisions and to maintain correct decisions when a more refined scale was used. CONCLUSIONS: The reanalysis presented here provides additional evidence that readers in a ROC study can adjust to a 101-point scale and the use of such a refined scale can increase discriminative ability. However, the decision of selecting an appropriate scale should also consider the underlying abnormality in question and relevant clinical considerations.     

An anthropomorphic phantom for receiver operating characteristic studies in CT imaging of liver lesions.

The purpose of the study was to develop a methodology that allowed quantitative assessment of image quality in CT and its relationship to dose. An anthropomorphic phantom was designed for use in receiver operating characteristic (ROC) studies of the detectability of liver lesions with CT. The lesions were simulated by different mixtures of glycerol and water that were filled into holes of different diameters in a liver tissue substitute. A pilot study was carried out on five different scanners that were operated at various exposure settings. A positive correlation was demonstrated for each of the scanners between the weighted CT dose index (CTDIW) and the area under the ROC curve. For the exposure settings used in the clinical routine in the five laboratories, the CTDIW ranged from 15 to 31 mGy. Three observers who read the corresponding set of five phantom images agreed, as judged from the areas under the ROC curves, that there was a marked difference in quality between the three best images and the other two. The two newest scanners in the study had the lowest CTDIW, and at the same time the best ROC results. The phantom and the ROC methodology may, with a set of suggested improvements, be used for comparison of the performance in different CT laboratories, and to establish the dose needed to ensure adequate image quality for a particular scanner.     

Comparison of adrenal vein sampling value between aldosterone producing adrenal adenoma and non-functioning adrenal adenoma: evaluation using receiver operating characteristic analysis

PURPOSE: To evaluate adequate criteria for adrenal vein sampling values in patients with aldosterone producing adrenal adenoma (APA), retrospectively. MATERIAL AND METHODS: Between 1988 and 2002, 59 hypertensive patients (15 M and 44 F, aged 47.58 +/- 9.45 years) were referred to our hospital and diagnosed with APA based on established criteria. During the same period, 23 patients with non-functioning adrenal adenoma (11 M and 12 F aged 53.56 +/- 11.76 years) were diagnosed based on computed tomography and laboratory data. RESULTS: All 82 patients were enrolled in the present study. Bilateral adrenal vein sampling (AVS) for measurement of plasma aldosterone (A) and cortisol (C) was performed, and a receiver operating characteristics (ROC) analysis was conducted to establish the best criteria from the AVS-derived index in patients with APA. A (APA side)/A (contralateral side) was confirmed to provide the best diagnostic accuracy {(>2.5: right APA, sensitivity 83.3%, specificity 79.5%), (> 3: left APA, sensitivity 84.2%, specificity 76.9)}. The Az values for A (APA side)/A (contralateral side) were 0.8948 and 0.9260 for right and left APA, respectively. CONCLUSION: The A (APA side)/A (contralateral side) value was the best compromise for sensitivity and false-positive rate for lateralization of APA.     

应用ROC曲线评价二维应变对冠心病的诊断价值

目的:应用二维应变超声心动图(2DSE)对常规方法显示室壁运动正常的冠心病患者进行定量分析,探讨其临床应用意义.方法:采集经冠脉造影证实的38例冠心病和31例对照组患者的心尖长轴、四腔和二腔观的二维灰阶动态图像,应用二维应变软件测量左室壁各节段的收缩应变、应变率及整体长轴应变(GLS),心尖双平面Simpson 法测量左室射血分数(LVEF).结果:长轴应变率(SrL)对冠心病的诊断价值最高,并以截断值为-0.855s-1诊断冠心病的Yuedden指数最高.长轴应变(SL)对左前降支狭窄的诊断价值最高,并以前壁及前间隔SL的截断值为-15.77% 诊断左前降支狭窄的Yuedden指数最高.结论:2DSE可准确定量左室整体和局部心肌应变,对评价心肌缺血的敏感性高于常规超声心动图.     

Thallium circumferential profiles in the detection of coronary artery disease--assessment by receiver operating characteristic curve analysis

Thallium uptake and redistribution in the myocardium was quantitated using circumferential activity profiles. The profiles were determined from images in the LAO 35, LAO 70 and anterior views, obtained after peak exercise and after 3-4 h redistribution. A retention curve was constructed from these profiles, displaying the redistribution profile as a fraction of the stress profile. Normal values were obtained from 21 normal patients and compared to the images from 68 patients suspected of having coronary artery disease (CAD). Of the latter, 20 had normal/non-significant CAD and 48 had 70% or greater stenosis of one or more vessels on coronary angiography. Several methodological variations were examined: calculation of the profiles using the peak or average counts around the ventricle, uniform or interpolative background subtraction, normalisation of the profiles to their peak or mean counts. Receiver operating characteristic (ROC) curves were generated for profiles calculated using these variations and compared with the ROC curves obtained by visual interpretation of both analogue and computer-enhanced images by two experienced observers. The uniform background subtraction method was found to be better than the interpolative method, and normalisation to the mean of each curve was preferable to normalisation to the peak. ROC curves using profiles calculated using the peak myocardial counts were identical to the ROC curves from profiles using the average counts, but the operating points on the two curves differed. Computer enhancement of the digital images resulted in increased sensitivity for CAD without loss of specificity when compared to the interpretation of the analogue images. The circumferential profiles were found to provide a further increase in sensitivity and were highly reproducible.(ABSTRACT TRUNCATED AT 250 WORDS)     

The use of continuous and discrete confidence judgments in receiver operating characteristic studies of diagnostic imaging techniques.

Virtually all receiver operating characteristic (ROC) studies of medical imaging methods have involved data collected on a discrete confidence-rating scale, though in principle ROC studies can be conducted with either discrete or continuous scales. Continuous scales provide potential advantages in medical applications of ROC analysis, but the possible dependence of ROC results on the kind of scale used in an observer-performance experiment has not been investigated empirically. The authors conducted a multi-observer ROC study in which radiologists read abdominal computed tomography (CT) examination results displayed on a workstation with a high-resolution video monitor. Ninety-five examinations were evaluated by five radiologists who reported their confidence concerning the presence of abdominal masses on a conventional five-category discrete scale and, in a separate reading session, on a continuous subjective-probability scale. ROC analysis showed no statistically significant difference between the detection accuracies (Az) achieved with the two scales by any reader. Accuracies averaged over all readers were identical with the two scales. The results of this study suggest that discrete or continuous scales often can be used interchangeably in image-evaluation studies. However, the authors recommend continuous scales for routine use in radiologic ROC studies because of their potential advantages in some situations.     

Statistical power in observer-performance studies: comparison of the receiver operating characteristic and free-response methods in tasks involving localization

RATIONALE AND OBJECTIVES: Statistical power, defined as the probability of detecting real differences between imaging modalities, determines the cost in terms of readers and cases of conducting receiver operating characteristic (ROC) studies. Neglect of location information in lesion-detection studies analyzed with the ROC method can compromise power. Use of the alternative free-response ROC (AFROC) method, which considers location information, has been discouraged, because it neglects intraimage correlations. The relative statistical powers of the two methods, however, have not been tested. The purpose of this study was to compare the statistical power of ROC and AFROC methods using simulations. MATERIALS AND METHODS: A new model including intraimage correlations was developed to describe the decision variable sampling and to simulate data for ROC and AFROC analyses. Five readers and 200 cases (half of which contained one signal) were simulated for each trial. Two hundred trials, equally split between the null hypothesis and alternative hypothesis, were run. Ratings were analyzed with the Dorfman-Berbaum-Metz method, and separation of the null hypothesis and alternative hypothesis distributions was calculated. RESULTS: The AFROC method yielded higher power than the ROC method. Separation of the null hypothesis and alternative hypothesis distributions was larger by a factor of 1.6 regardless of the presence or absence of intraimage correlations. The effect of the incorrect localizations during ROC analysis of localization data is believed to be the major reason for the enhanced power of the AFROC method. CONCLUSION: The AFROC method can yield higher power than the ROC method for studies involving lesion localization. Greater consideration of this methodology is warranted.