[Usefulness of Voxel-Based Quantification (VBQ) Smoothing in Relaxation Time Mapping].

PURPOSE: Voxel-based quantification (VBQ) smoothing is a technique used to smooth quantitative parametric maps in the Montreal Neurological Institute standard space. Although VBQ smoothing could suppress changes in quantitative values at tissue boundaries, its effectiveness on relaxation time (T1 and T2 values and proton density PD) maps has not been investigated. The purpose of this study was to clarify the usefulness of VBQ smoothing in relaxation time mapping. METHOD: T1 and T2 values and PD maps of the brains of 20 healthy participants were obtained using a two-dimensional multi-dynamic multi-echo sequence. VBQ and Gaussian smoothing were applied to the relaxation time maps by varying the kernel size by 1 mm from 1 to 6 mm. Changes in relaxation time before and after VBQ and Gaussian smoothing for the putamen, caudate nucleus, substantia nigra, and corpus callosum on the relaxation time maps were evaluated. RESULT: The changes in relaxation time after VBQ smoothing application were smaller than those in that after Gaussian smoothing application. Although the differences in the relaxation time for all tissues before and after VBQ and Gaussian smoothing applications increased with increasing kernel size for all relaxation times for both methods, the changes in the relaxation time for VBQ smoothing were smaller than those in that for Gaussian smoothing. CONCLUSION: VBQ smoothing can suppress the change in the relaxation time on the boundary of the tissue and is thus a useful smoothing technique in relaxation time mapping.

[Evaluation of the Accuracy of Relaxation Time Measurements Using 3D-QALAS at 3.0 T MRI and Comparison with 2D-MDME].

PURPOSE: Three-dimensional (3D) quantification using an interleaved Look-Locker acquisition sequence with a T2 preparation pulse (QALAS) is a quantitative sequence used to measure relaxation times. The accuracy of the relaxation time measurement of 3D-QALAS at 3.0 T and the bias of 3D-QALAS have not yet been assessed. The purpose of this study was to clarify the accuracy of the relaxation time measurements using 3D-QALAS at 3.0 T MRI. METHODS: The accuracy of the T1 and T2 values for 3D-QALAS was evaluated using a phantom. Subsequently, the T1 and T2 values and proton density of the brain parenchyma in healthy subjects were measured using 3D-QALAS and compared with those of 2D multi-dynamic multi-echo (MDME). RESULTS: In the phantom study, the average T1 value of 3D-QALAS was 8.3% prolonged than that for conventional inversion recovery spin-echo; the average T2 value for 3D-QALAS was 18.4% shorter than that for multi-echo spin-echo. The in vivo assessment showed that the mean T1 and T2 values and PD for 3D-QALAS were prolonged by 5.3%, shortened by 9.6%, and increased by 7.0%, respectively, compared with those for 2D-MDME. CONCLUSION: Although 3D-QALAS at 3.0 T has high accuracy T1 value, which is less than 1000 ms, the T1 value could be overestimated for tissues with it longer than that T1 value. The T2 value for 3D-QALAS could be underestimated for tissues with T2 values, and this tendency increases with longer T2 values.

A Preliminary Pathological Evaluation of Extracellular Volume Fraction with Contrast-enhanced Computed Tomography as a Novel Quantitative Parameter of Pancreatic Fibrosis.

Objective The extracellular volume (ECV) calculated based on contrast-enhanced computed tomography (CT) has been reported as a novel imaging parameter reflecting the morphological change of fibrosis in several parenchymal organs. Our retrospective study assessed the validity of the ECV fraction for diagnosing pancreatic fibrosis and the appropriate imaging condition as the "equilibrium phase". Methods In 27 patients undergoing multiphasic CT and subsequent pancreaticoduodenectomy, we investigated pathological fibrotic changes related to the ECV fraction and conducted analyses using the value obtained by subtracting the equilibrium CT value of the portal vein from that of the abdominal aorta (Ao-PVequilibrium) to estimate eligibility of the equilibrium phase. Results In all patients, the ECV fraction showed a weak positive correlation with the collagenous compartment ratio (r=0.388, p=0.045). All patients were divided into two groups - the high-Ao-PVequilibrium group and low-Ao-PVequilibrium group - based on the median value. No significant correlation was found in the high-Ao-PVequilibrium group, whereas a significant correlation was observed in the low-Ao-PVequilibrium group (r=0.566, p=0.035). Conclusion The ECV fraction is a possible predictive factor for histopathological pancreatic fibrosis. In its clinical application, the eligibility of the "equilibrium phase" may affect the diagnostic capability. It will be necessary to verify the imaging conditions in order to improve the accuracy of the diagnosis.

Novel Mathematical Diagnostic Analysis of Malignant Biliary Stenosis Using Magnetic Resonance Cholangiography in Patients Undergoing Pancreaticoduodenectomy.

BACKGROUND/AIM: A novel mathematical diagnostic analysis using Fourier transform (FT) algorithm of the extrahepatic bile duct (BD) using magnetic resonance-cholangiography (MRC) was performed to evaluate irregularities of the bile duct lumen indicating BD cancer (BDC) extension compared to pancreatic head malignancies controls. PATIENTS AND METHODS: BD lumen was automatically traced, and a 2D-diagram cross-section was measured and a FT-based integrated-power-spectral-density-function value (FTV) of both diameter and area (mm 2  and mm 4 /Hz) was calculated for cancerous and non-cancerous parts utilizing a computer workstation. RESULTS: FT analysis that was achieved in 59 patients consisted of BDC in 31, pancreatic cancer with biliary stenosis (PC) in 10 and pancreatic neoplasm without stenosis (PN) in 18. FTV-diameter and -area of non-tumorous proximal BD were larger compared to tumor involving BD (p<0.01), and those of the entire BD in BDCs were significantly larger than those in PN (p<0.01). FTV-diameter and -area in proximal BDC-positive were smaller than those in BDC-negative (p<0.05). BDC part was significantly discriminated by the cutoff value (286 mm 2 Hz -1  in diameter and 10,311 mm 4 Hz -1 in area) compared to PC and diagnostic accuracy was over 70% (p<0.01). CONCLUSION: Novel mathematical MRC FT-analysis is promising for differentiating between BDC and PC with biliary stenosis and can be utilized as an objective diagnostic tool in the future.

Novel analysis using magnetic resonance cholangiography for patients with pancreaticobiliary maljunction.

PURPOSE: We used a novel diagnostic Fourier transform (FT) algorithm of the entire extrahepatic bile duct (EHBD) measured by magnetic resonance cholangiography (MRC) to evaluate subtle deformation of bile duct lumen, indicating the malignant potential of EHBD, in patients with pancreaticobiliary maljunction (PBMJ) and in a comparative group of controls without PBMJ. METHODS: From the workstation, the EHBD lumen was traced automatically and a 2D diagram cross section was measured at 0.5 mm-longitudinal intervals. The FT-based integrated power spectral density function value (FTPSDI) of the diameter or area (mm2 or mm4/Hz) and the phase value distribution entropy (PVDE) were also measured. RESULTS: There were 16 patients with undilated PBMJ and 7 with dilated PBMJ. The control group comprised 10 patients with a normal bile duct, 20 with bile duct carcinoma (BDC), and 1 with primary sclerosing cholangitis. Both the diameter and area of the dilated bile ducts and the ducts with early- or advanced-stage BDC were significantly greater than those of the normal duct (p < 0.05). The undilated type of PBMJ tended to have a larger FTPSDI diameter than a normal bile duct, which had a smaller diameter than the dilated type of PBMJ or BDC. BDC had a significantly larger FTPSDI diameter (p < 0.05) and the cutoff value for accuracy was 168 mm2 Hz-1. CONCLUSION: The novel mathematical FTPSDI is a promising indicator of whether preventive EHBD resection is necessary for patients with PBMJ, which can be widely applied in the early diagnosis of other biliary diseases.

[Prediction of the Incomplete Region of Fat Suppression Using the Banding Artifact of the Balanced Steady-state Free Precession Sequence].

When the fat-suppression technique was used in the MRI examinations of neck and extremities, incomplete regions of fat suppression were depicted frequently. These incomplete regions were caused by the non-uniform static magnetic field (B0). On the other hand, a non-uniform B0 caused banding artifacts using a balanced steady-state free precession (bSSFP) sequence. We investigated the prediction of incomplete regions of fat suppression using the banding artifacts of the bSSFP sequence. The fat-suppression techniques used in this study were the chemical shift selective and spectral adiabatic inversion recovery methods for turbo spin echo imaging. Using an oil bottle attached to a staple, the scan parameters of the bSSFP sequence were adjusted to overlap with the banding artifacts in the incomplete regions of fat suppression. The neck, ankle joint and femur of healthy volunteers were scanned using the obtained scan parameters, and a visual assessment was performed. As a result, the incomplete region of fat suppression matched the region of the banding artifact. We were able to predict the incomplete region of fat suppression using the banding artifacts of the bSSFP sequence. If the optimized bSSFP sequence is used as the localizer, we can assess the situation of incomplete fat suppression before the main scanning and reduce rescanning due to incomplete fat suppression.

A 3D Quantitative MRC Modeling Images Detected Case of Intrahepatic Biliary Stricture Diseases.

In the future, the application of quantitative imaging and computational analysis will reduce the burden on radiologists. We herein report 8 pilot cases both with and without intrahepatic biliary stricture (IHBS) diseases which have been analyzed with the novel analytical system MRCP+ (Perspectum Ltd., Oxford, UK). The colored and well-visualized 3D models of the entire biliary trees could be obtained in all 8 cases. Three representative cases did not show dilated regions in the intrahepatic bile ducts. Cases diagnosed as a pancreatico-biliary maljunction showed slightly increased dilated visualization in the extrahepatic duct. Except in a case with severe stenosis resulting from hepatolithiasis, the number of visualized intrahepatic bile ducts tended to be decreased and the volume of biliary tree and the total length of stricture and dilatation were also decreased. However, the number of IHBS or dilatation was unchanged. The number of strictures obtained by MRCP+ and the subjective counts of stenosis from a radiologist was not found to be correlated. In a case of severe stenosis at the left lateral bile duct, the number of intrahepatic biliary dilatations was increased. The latest computerized 3D modeling technology was found to be useful in visualizing the alteration of intraluminal diameter of the entire biliary trees at a glance, which can provide the automatic diagnosis of IHBS diseases at an earlier phase.

Low putamen activity associated with poor reward sensitivity in childhood chronic fatigue syndrome.

Motivational signals influence a wide variety of cognitive processes and components of behavioral performance. Cognitive dysfunction in patients with childhood chronic fatigue syndrome (CCFS) may be closely associated with a low motivation to learn induced by impaired neural reward processing. However, the extent to which reward processing is impaired in CCFS patients is unclear. The aim of the present functional magnetic resonance imaging (fMRI) study was to determine whether brain activity in regions related to reward sensitivity is impaired in CCFS patients. fMRI data were collected from 13 CCFS patients (mean age, 13.6 ± 1.0 years) and 13 healthy children and adolescents (HCA) (mean age, 13.7 ± 1.3 years) performing a monetary reward task. Neural activity in high- and low-monetary-reward conditions was compared between CCFS and HCA groups. Severity of fatigue and the reward obtained from learning in daily life were evaluated by questionnaires. Activity of the putamen was lower in the CCFS group than in the HCA group in the low-reward condition, but not in the high-reward condition. Activity of the putamen in the low-reward condition in CCFS patients was negatively and positively correlated with severity of fatigue and the reward from learning in daily life, respectively. We previously revealed that motivation to learn was correlated with striatal activity, particularly the neural activity in the putamen. This suggests that in CCFS patients low putamen activity, associated with altered dopaminergic function, decreases reward sensitivity and lowers motivation to learn.

Intracranial dural arteriovenous fistulas: evaluation with 3-T four-dimensional MR angiography using arterial spin labeling.

PURPOSE: To evaluate whether 3-T four-dimensional (4D) arterial spin-labeling (ASL)-based magnetic resonance (MR) angiography is useful for the evaluation of shunt lesions in patients with intracranial dural arteriovenous fistulas (AVFs). MATERIALS AND METHODS: Institutional review board approval and prior written informed consent from all patients were obtained. Nine patients with intracranial dural AVF (seven men, two women; age range, 52-77 years; mean age, 63 years) underwent 4D ASL MR angiography at 3 T and digital subtraction angiography (DSA). Spin tagging was with flow-sensitive alternating inversion recovery with Look-Locker sampling. At 300-millisecond intervals, seven dynamic images with a spatial resolution of 0.5 × 0.5 × 0.6 mm(3) were obtained. The 4D ASL MR angiographic and DSA images were read by two sets of two independent readers each. Interobserver and intermodality agreement was assessed with the κ statistic. RESULTS: On all 4D ASL MR angiographic images, the major intracranial arteries were demonstrated at a temporal resolution of 300 milliseconds. Interobserver agreement was excellent for the fistula site (κ = 1.00; 95% confidence interval [CI]: 1.00, 1.00), moderate for the main arterial feeders (κ = 0.53; 95% CI: 0.08, 0.98), and good for venous drainage (κ = 0.77; 95% CI: 0.35, 1.00). Intermodality agreement was excellent for the fistula site and venous drainage (κ = 1.00; 95% CI: 1.00, 1.00) and good for the main arterial feeders (κ = 0.80; 95% CI: 0.58, 1.00). CONCLUSION: The good-to-excellent agreement between 3-T 4D ASL MR angiographic and DSA findings suggests that 3-T 4D ASL MR angiography is a useful tool for the evaluation of intracranial dural AVFs.

Osmotic release oral system-methylphenidate improves neural activity during low reward processing in children and adolescents with attention-deficit/hyperactivity disorder.

Attention-deficit/hyperactivity disorder (ADHD) is neurobehavioral disorder characterized by inattention, hyperactivity/impulsivity and impaired reward system function, such as delay aversion and low reward sensitivity. The pharmacological treatment for ADHD includes methylphenidate (MPH), or osmotic release oral system-MPH (OROS-MPH), which increases extrasynaptic dopamine and noradrenaline levels by blocking their reuptake. Although previous functional magnetic resonance imaging (fMRI) studies revealed that acute treatment with MPH alters activation of the nucleus accumbens during delay aversion in children and adolescents with ADHD, the effects a relatively long period of OROS-MPH treatment on delay aversion as well as reward sensitivity remain unclear. Thus, we evaluated brain activation with fMRI during a reward sensitivity paradigm that consists of high monetary reward and low monetary reward conditions before and after a 3-month treatment with OROS-MPH in 17 children and adolescents with ADHD (mean age, 13.3 ± 2.2) and 17 age- and sex-matched healthy controls (mean age, 13.0 ± 1.9). We found that before treatment there was decreased activation of the nucleus accumbens and thalamus in patients with ADHD during only the low monetary reward condition, which was improved to same level as those of the healthy controls after the treatment. The observed change in brain activity was associated with improved ADHD symptom scores, which were derived from Japanese versions of the ADHD rating scale-IV. These results suggest that treatment with OROS-MPH for a relatively long period is effective in controlling reward sensitivity in children and adolescents with ADHD.

[Abbreviations of special terms for presentation/paper titles in magnetic resonance study].

PURPOSE: A large number of abbreviations have been created for various special terms, and used in magnetic resonance (MR) study. However, the use of these abbreviations in the paper title has been restricted by the majority of societies and journals. In this study, we investigated the use of various abbreviations for special terms in MR study in order to clarify which abbreviation could be used in the paper title without spelling. MATERIALS AND METHODS: We used two journals, Magnetic Resonance in Medicine (MRM) and Journal of Magnetic Resonance Imaging (JMRI) published by the International Society for Magnetic Resonance in Medicine (ISMRM), which has been considered to be the most advanced society for MR study in the world, as the reference standard for use of the abbreviations. RESULTS: Except for some basic abbreviations and specific abbreviations that were used on a long-term basis, the majority of abbreviations were used in the paper title with its full spelling in order to ensure generality. CONCLUSION: It is preferable that abbreviations not be used in the title of the or title of the paper.

Comparison of 3D phase-sensitive inversion-recovery and 2D inversion-recovery MRI at 3.0 T for the assessment of late gadolinium enhancement in patients with hypertrophic cardiomyopathy.

RATIONALE AND OBJECTIVES: To compare free-breathing three-dimensional (3D) phase-sensitive inversion recovery (PSIR) with breath-holding two-dimensional (2D) IR sequences to determine which is better for detecting and characterizing myocardial late gadolinium enhancement (LGE) in hypertrophic cardiomyopathy (HCM) patients. MATERIALS AND METHODS: Thirty HCM patients clinically underwent 3.0 T cardiac magnetic resonance imaging that included 3D-PSIR and 2D-IR. The amount of LGE lesions was calculated and expressed as %LGE of the myocardial mass, and the average of the %LGE value reported by two observers was recorded as the final %LGE. We also counted the number of LGE lesions and recorded their location. The myocardium-LGE contrast, margin sharpness, artifacts, and overall image quality were graded on a 4-point grading scale (1 = poor, 2 = fair, 3 = good, 4 = excellent). RESULTS: The mean %LGE on 2D-IR was 24.7 ± 0.6, 17.5 ± 0.6, and 8.5 ± 0.3, respectively, for the basal, mid-, and apical myocardium; the corresponding values were 24.2 ± 0.4, 20.0 ± 0.4, and 7.7 ± 0.3 on 3D-PSIR (2D-IR versus 3D-PSIR, P = .87). On 2D IR and 3D-PSIR images, 13, 52, and 53, and 9, 74, and 33 LGE lesions were detected in the subendocardial, midwall, subepicardial area, respectively. The myocardium-LGE contrast and overall image quality were significantly higher on 3D-PSIR than 2D-IR images (P < .001); the sequences did not differ significantly with respect to margin sharpness and artifact. CONCLUSION: Three-dimensional PSIR sequence yields higher image contrast, better image quality, and greater detection ability for LGE lesions than 2D-IR sequence.

[Development of the computerized scheme for reconstructing three dimensional ultrasonography from the conventional two dimensional dynamic images].

In this study, a computerized scheme for reconstructing three dimension (3D)-ultrasonography (US) of a breast mass using off-line B-mode two dimension (2D) dynamic US images obtained by conventional 2D equipment was developed. Two off-line 2D dynamic US images were obtained with two orthogonal directions for each mass. The z axis of reconstructed 3D-US of a breast mass was estimated by using a length of the mass on the orthogonal image. In order to verify whether the 3D-US reproduces the real form of a breast mass, the correlation between two image features (volume and effective diameter of a mass), which were calculated from 3D-US, and visual assessment reported by physicians were evaluated. For seventy-nine breast masses used in this study, estimated mass volumes obtained from 3D-US had a high correlation (R=0.925) with those obtained from diagnostic reports. In conclusion, it was suggested that our computerized scheme could reproduce the 3D form of breast masses with relatively small errors.

Signal intensity of lanthanum carbonate on magnetic resonance images: phantom study.

PURPOSE: Lanthanum carbonate (LC) is used to treat hyperphosphatemia. The purpose of this study was to investigate the signal intensity (SI) of LC on magnetic resonance imaging (MRI) scans of phantoms. MATERIALS AND METHODS: LC tablets were thoroughly ground and mixed with distilled water or edible agar (0.05, 0.25, 0.5, and 2.5 mg/ml) in plastic bottles. Four intact tablets were placed in plastic bottles that did or did not contain distilled water or agar. Two radiologists consensually evaluated T1- and T2-weighted images (WIs) obtained with 1.5- and 3.0-T MRI systems for the SI of unground and ground tablets. RESULTS: On T1- and T2WI, the SIs of the LC suspensions and the solvents alone were similar; the SIs of unground tablets alone and of the air were also similar. Unground tablets in phantoms filled with solvent exhibited lower SI than the solvent. Ground tablets in suspension were not visualized on MRI or computed tomography. These results remained unchanged regardless of differences in magnetic field strength or the solvent used. CONCLUSION: Ground LC had no contrast enhancement effect on T1WI; on T2WI it did not affect the SI of the solvent. Unground LC tablets may be visualized as a "filling defect" on MRI.

Enhancement effects of hepatic dynamic MR imaging at 3.0 T and 1.5 T using gadoxetic acid in a phantom study: comparison with gadopentetate dimeglumine.

To identify the optimum sequence at gadoxetic acid enhanced hepatic dynamic magnetic resonance imaging in the arterial phase, we studied phantoms that contained gadoxetic acid or gadopentetate dimeglumine diluted in human blood. We obtained magnetic resonance images at 3.0 T and 1.5 T with one vendor (Siemens) using 3D-gradient echo (GRE)-, 2D-fast low angle shot (FLASH)-, and turbo spin echo sequences. Contrast ratio was highest for 3D-GRE; at both 3.0 T and 1.5 T it was superior when the contrast agent was gadoxetic acid. With both gadoxetic acid and gadopentetate dimeglumine, contrast ratio peaked at around 5- and 2 mmol/L on 3D-GRE- and 2D-FLASH images, respectively. Compared with gadopentetate dimeglumine, at 3.0 T, the peak contrast ratio of gadoxetic acid was 14.1% better on 3D-GRE images and 14.0% better on 2D-FLASH images; at 1.5 T it was 16.4% better on 3D-GRE- and 5.7% better on 2D-FLASH images. With respect to the magnetic field strength, at 3.0 T the peak contrast ratio of gadoxetic acid was 6.0% better than at 1.5 T on 3D-GRE images and 49.5% better on 2D-FLASH images; it was 8.5% better on 3D-GRE- and 44.6% better on 2D-FLASH images than when the contrast agent was gadopentetate dimeglumine. Thus, gadoxetic acid yielded better enhancement on 3D-GRE images acquired at 3.0 T than at 1.5 T and enhancement was better than that obtained with gadopentetate dimeglumine at the same concentration.

[Investigation of T2* map measurement sequence in the liver at 3.0T MR].

T2* value measurement of the liver parenchyma with a 3.0T MR scanner may be useful for evaluating focal liver function. Currently, there are 2 sequences for measurement of T2* value of the liver: multi-echo fast field echo (mFFE), and multi-echo planar imaging (EPI). We can correct inhomogeneity of the local magnetic field with the EPI sequence; however, the spatial resolution is poor. On the other hand, mFFE has a relatively high spatial resolution but cannot correct inhomogeneity of the local magnetic field. We investigated the two measurement methods of a T2* map that measured the T2* images obtained with mFFE and EPI sequences by using a 3.0T MR scanner in the phantom and patient studies. In the phantom studies, T2* values measured on images with the mFFE sequence were affected by inhomogeneity of the local magnetic field, but T2* values measured on images with the EPI sequence were showed no difference by corrected inhomogeneity of local magnetic field. However, in the clinical study, we found good agreement in T2* values of the liver measured on images with mFFE and EPI sequences. Therefore, the mFFE sequence can be an alternative to the EPI sequence in the clinical setting. It is occasionally difficult to identify normal or pathological structures on images obtained with the EPI sequence because of its low spatial resolution. The spatial resolution of images obtained with the mFFE sequence is much better than that with the EPI sequence. Based on these discussions, we believe that the mFFE sequence may be appropriate for the measurement of T2* values in the liver in the clinical setting.

Detection of hemorrhagic hypointense foci in the brain on susceptibility-weighted imaging clinical and phantom studies.

RATIONALE AND OBJECTIVES: To determine the sensitivity of susceptibility-weighted imaging (SWI) for depicting hemorrhagic hypointense foci of the brain in comparison with gradient-recalled echo (GRE)- and GRE-type single-shot echo-planar imaging (GREI, GRE-EPI), and to assess the basic characteristics of the susceptibility effect by using a phantom. MATERIALS AND METHODS: We prospectively examined 16 patients (9 males, 7 females, aged 10-74 years, mean 43 years) with hypointense foci using SWI, GREI, and GRE-EPI at a 1.5-T magnetic resonance (MR) unit. The contrast-to-noise ratio (CNR), sensitivity to small hypointese foci, and artifacts were evaluated. To assess the basic characteristics of SWI, we performed a phantom study using different concentrations of superparamagnetic iron oxide (SPIO). RESULTS: The CNR of lesions was significantly greater for SWI than the other images (P < .0001). SWI detected the greatest number of small hypointense foci, even in the near-skull-base and infratentorial regions. Quantitative and qualitative analyses in our clinical and phantom studies demonstrated that the degree of artifacts was similar with SWI and GREI. CONCLUSION: SWI was best for detecting small hemorrhagic hypointense foci. Artifacts of SWI were similar to GREI.

Pulmonary nodules at chest CT: effect of computer-aided diagnosis on radiologists' detection performance.

PURPOSE: To evaluate the effect of computer-aided diagnosis (CAD) on radiologists' detection of pulmonary nodules. MATERIALS AND METHODS: Fifty chest computed tomographic (CT) examination cases were used. The mean nodule size was 0.81 cm +/- 0.60 (SD) (range, 0.3-2.9 cm). Alternative free-response receiver operating characteristic (ROC) analysis with a continuous rating scale was used to compare the observers' performance in detecting nodules with and without use of CAD. Five board-certified radiologists and five radiology residents participated in an observer performance study. First they were asked to rate the probability of nodule presence without using CAD; then they were asked to rate the probability of nodule presence by using CAD. RESULTS: For all radiologists, the mean areas under the best-fit alternative free-response ROC curves (Az) without and with CAD were 0.64 +/- 0.08 and 0.67 +/- 0.09, respectively, indicating a significant difference (P <.01). For the five board-certified radiologists, the mean Az values without and with CAD were 0.63 +/- 0.08 and 0.66 +/- 0.09, respectively, indicating a significant difference (P <.01). For the five resident radiologists, the mean Az values without and with CAD were 0.66 +/- 0.04 and 0.68 +/- 0.04, respectively, indicating a significant difference (P =.02). At observer performance analyses, there were no significant differences in Az values obtained either without (P =.61) or with (P =.88) CAD between the board-certified radiologists and the residents. For all radiologists, in the detection of pulmonary nodules 1.0 cm in diameter or smaller, the mean Az values without and with CAD were 0.60 +/- 0.11 and 0.64 +/- 0.11, respectively, indicating a significant difference (P <.01). CONCLUSION: Use of the CAD system improved the board-certified radiologists' and residents' detection of pulmonary nodules at chest CT.