Junji Shiraishi

Assessment of arterial hypervascularity of hepatocellular carcinoma: comparison of contrast-enhanced US and gadoxetate disodium-enhanced MR imaging

AbstractObjectiveTo compare contrast-enhanced (CE) ultrasound with gadoxetate disodium-enhanced magnetic resonance (MR) imaging in the assessment of arterial hypervascularity of hepatocellular carcinoma (HCC) and dysplastic nodule (DN), with CT during hepatic arteriography (CTHA) as the reference standard.MethodsThis study included 54 consecutively diagnosed patients, with 57 histologically confirmed HCCs and 3 DNs (high-grade). All patients underwent CE ultrasound, gadoxetate disodium-enhanced MR imaging and CTHA. Two trained diagnostic radiologists interpreted the CTHA images and rated the degree of intratumoral arterial vascularity by consensus using a five-point confidence scale as the reference standard. In the observer study, the degrees of vascularity on CE ultrasound and gadoxetate disodium-enhanced MR images were qualitatively analysed by four independent readers using a five-point confidence scale. Diagnostic accuracy was analysed by receiver-operating characteristic (ROC) analysis.ResultsThe diagnostic accuracies of the average area under the ROC curve (AUC) were significantly greater with CE ultrasound (average AUC: 0.94; 95% confidence interval [CI]: 0.88–1.00) than with gadoxetate disodium-enhanced MR imaging (average AUC 0.84, 95% CI 0.74–0.93, P = 0.0014).ConclusionContrast-enhanced ultrasound yields a significantly higher AUC value than gadoxetate disodium-enhanced MR imaging in the assessment of arterial hypervascularity of HCC and DN. Key Points • Arterial hypervascularity is an important feature determining treatment options in hepatocellular carcinoma. • It can be assessed by contrast-enhanced (CE) ultrasound or magnetic resonance (MR) imaging. • CE ultrasound was more accurate than Gd-EOB-DTPA MRI in assessing intratumoral vascularity. • Hypovascular hepatic nodules should be further investigated using CE ultrasound.

Comparison of Kupffer-phase Sonazoid-enhanced sonography and hepatobiliary-phase gadoxetic acid-enhanced magnetic resonance imaging of hepatocellular carcinoma and correlation with histologic grading.

To determine the relative wash‐out of hepatocellular carcinomas and dysplastic nodules using Kupffer‐phase sonography with Sonazoid (Daiichi‐Sankyo, Tokyo, Japan) enhancement and hepatobiliary‐phase gadoxetic acid‐enhanced magnetic resonance imaging (MRI) in the evaluation of the histopathologic grades of individual nodules.

Ethnic Difference in Hematological Toxicity in Patients with Non-small Cell Lung Cancer Treated with Chemotherapy: A Pooled Analysis on Asian versus Non-Asian in Phase II and III Clinical Trials

Introduction: There are a large number of global clinical trials ongoing for patients with non-small cell lung cancer (NSCLC). Ethnic difference in toxicity has not been adequately studied. Methods: We performed a systematic search in PubMed for randomized phase II and III trials of NSCLC from January 2000 to December 2009, examining ethnic difference in hematological toxicity due to cytotoxic chemotherapy. Ethnicity was classified into Asian and non-Asian. We chose three treatment regimens used for NSCLC globally: cisplatin plus gemcitabine (CG), cisplatin plus vinorelbine (CV), and carboplatin plus paclitaxel (CP). We applied sensitivity analysis to examine unreported ethnic differences in hematological toxicities by changing the percentage of Asian patients from 0 to 18% in trials reported from the United States and Europe. Results: We identified 12 phase II trials and 38 phase III trials of NSCLC with a total of 11,271 patients. Among these, 14 trials had reported ethnic origins. Grade 3/4 toxicities were more frequently observed in the Asian studies. On the basis of sensitivity analysis, odds ratio of grade 3/4 neutropenia was significantly higher in Asian patients than non-Asian, when treated with CG (OR = 1.55–3.45, p < 0.001), CV (OR = 2.99–4.43, p < 0.001), and CP (OR = 4.79–6.22, p < 0.001). Grade 3/4 anemia was also significantly higher in Asians with CG (OR = 3.10–3.27, p < 0.001), CV (OR = 1.99–2.43, p < 0.001), and CP (OR = 1.34–1.52, p < 0.001–0.004). However, no significant difference was observed in thrombocytopenia with CG (OR = 0.66–2.04, p < 0.001–1.000), CV (OR = 0.42–0.57, p = 0.097–0.323), or CP (OR = 1.21–1.39, p = 0.114–0.152). Conclusions: Severe hematological toxicity was frequently observed in Asian patients compared with non-Asian (mostly whites) in the treatment of chemotherapy for NSCLC.

Modulation transfer function measurement of CT images by use of a circular edge method with a logistic curve-fitting technique

We propose a method for measuring the modulation transfer function (MTF) of a computed tomography (CT) system by use of a circular edge method with a logistic curve-fitting technique. An American College of Radiology (ACR) phantom was scanned by a Philips Brilliance system, and axial images were reconstructed by the filtered back projection algorithm with a standard reconstruction filter. The radial MTF was measured from a disk image of a rod or cylinder in the ACR phantom by use of the circular edge method. In this study, we applied a logistic curve-fitting technique to an edge-spread function (ESF) to eliminate noise because the edge method is very susceptible to noise in the ESF in a CT image. The circular edge method with the logistic curve-fitting technique provided the MTF without fluctuations due to noise for the entire spatial frequency range. The MTF was not affected by the tube current, the slice thickness, or the disk contrast, which were factors related to the amount of noise in the CT image. However, the MTF was affected by the location of the disk and by the disk size, depending on the average distance from the isocenter to the disk edge. Our results indicated that the MTF measured by the circular edge method with the logistic curve-fitting technique was not susceptible to noise in CT images. Therefore, this method is useful for MTF measurement for not only high-contrast objects, but also low-contrast objects with a large amount of noise.

Basic concepts and development of an all-purpose computer interface for ROC/FROC observer study

In this study, we initially investigated various aspects of requirements for a computer interface employed in receiver operating characteristic (ROC) and free-response ROC (FROC) observer studies which involve digital images and ratings obtained by observers (radiologists). Secondly, by taking into account these aspects, an all-purpose computer interface utilized for these observer performance studies was developed. Basically, the observer studies can be classified into three paradigms, such as one rating for one case without an identification of a signal location, one rating for one case with an identification of a signal location, and multiple ratings for one case with identification of signal locations. For these paradigms, display modes on the computer interface can be used for single/multiple views of a static image, continuous viewing with cascade images (i.e., CT, MRI), and dynamic viewing of movies (i.e., DSA, ultrasound). Various functions on these display modes, which include windowing (contrast/level), magnifications, and annotations, are needed to be selected by an experimenter corresponding to the purpose of the research. In addition, the rules of judgment for distinguishing between true positives and false positives are an important factor for estimating diagnostic accuracy in an observer study. We developed a computer interface which runs on a Windows operating system by taking into account all aspects required for various observer studies. This computer interface requires experimenters to have sufficient knowledge about ROC/FROC observer studies, but allows its use for any purpose of the observer studies. This computer interface will be distributed publicly in the near future.

A phantom study comparing ultrasound-guided liver tumor puncture using new real-time 3D ultrasound and conventional 2D ultrasound

OBJECTIVE The purpose of this study is to compare the accuracy of ultrasound-guided puncture using new real-time 3D (4D) ultrasound and conventional 2D ultrasound for focal hepatic masses using a liver phantom. MATERIALS AND METHODS A 4D ultrasound system equipped with a 5-MHz 4D probe displayed both axial and orthogonal images parallel to a puncture line plane. We used a liver phantom that contained four simulated spherical masses in an acrylic box (length × width × height, 300 × 299 × 150 mm) with two different sizes (15 and 30 mm in diameter) in two different positions (30 and 80 mm from the surface). Four inexperienced and four experienced physicians attempted punctures on these four simulated masses twice using 2D and 4D ultrasound guidance in a total of 128 punctures (eight operators, two techniques, and eight punctures per session). The error distance of the puncture was defined as the perpendicular distance from the center of a target mass (sphere) to the line of the puncture needle in the coronal plane of the target center, which was measured manually on the basis of the 3D volume data on off-line analysis. RESULTS On each tumor model, the average error distance with 4D ultrasound was significantly smaller than that with 2D ultrasound, except for one tumor model that was 15 mm in diameter and 30 mm in depth. The average error distances for the experienced group tended to be smaller than those for the inexperienced group, with both techniques and on each tumor model, and there was a statistically significant difference between the two groups on one tumor model (30 mm in diameter and 80 mm in depth) on 4D ultrasound (p < 0.05). CONCLUSION Four-dimensional ultrasound-guided puncture for liver tumors can markedly improve puncture accuracy for both experienced and inexperienced physicians compared with conventional 2D ultrasound guidance.

Is response rate increment obtained by molecular targeted agents related to survival benefit in the phase III trials of advanced cancer

BACKGROUND It remains unclear whether response rate (RR) is related to survival benefit in phase III trials of advanced cancer treated with molecular targeted agents (MTA) in combination with standard therapies. MATERIALS AND METHODS We carried out a systematic search of PubMed for randomized phase III trials of four solid tumors examining the efficacy of MTA when added to a standard therapy. We examined whether there were any associations between RR increment obtained by the addition of targeted agents (DeltaRR) and survival benefit in phase III trials. RESULTS We identified 26 phase III trials of MTA with a total of 21 156 patients and 29 experimental arms of MTA. Studies which showed significant survival benefit had higher DeltaRR compared with those which did not show significant benefit. In the receiver operating characteristic curve analysis, using a 7% gain as threshold value for DeltaRR allowed assessment of survival benefit with high sensitivity and specificity. There were also significant relationships between DeltaRR and hazard ratios for overall survival and progression-free survival in the linear regression analysis. CONCLUSION RR increment obtained by the addition of MTA to a standard therapy may be useful to predict survival benefit in clinical phase III trials of advanced cancer.

Computer-aided diagnosis for estimating the malignancy grade of hepatocellular carcinoma using contrast-enhanced ultrasound: an ROC observer study.

We are developing a computer‐aided diagnosis (CAD) system for estimating the malignancy grade of hepatocellular carcinoma (HCC) using contrast‐enhanced ultrasound (CEUS). In this study, observers estimated the malignancy grade of HCC with and without the cues provided by CAD.

Verification of modified receiver-operating characteristic software using simulated rating data

ROCKIT, which is a receiver-operating characteristic (ROC) curve-fitting software package, was developed by Metz et al. In the early 1990s, it is a very frequently used ROC software throughout the world. In addition to ROCKIT, DBM-MRMC software was developed for multi-reader multi-case analysis of the difference in average area under ROC curves (AUCs). Because this old software cannot run on a PC with Windows 7 or a more recent operating system, we developed new software that employs the same basic algorithms with minor modifications. In this study, we verified our modified software and tested the differences between the index of diagnostic accuracies using simulated rating data. In our simulation model, all data were generated using target AUCs and a binormal parameter b. In ROC curve fitting with simulated rating data, we varied four factors: the total number of case samples, the ratio of positive-to-negative cases, a binormal parameter b, and the preset AUC. To investigate the differences between the statistical test results obtained from our software and the existing software, we generated simulated rating data sets with three levels of case difficulty and three degrees of difference in AUCs obtained from two modalities. As a result of the simulation, the AUCs estimated by the new and existing software were highly correlated (R > 0.98), and there were high agreements (85% or more) in the statistical test results. In conclusion, we believe that our modified software is as capable as the existing software.

Computer-aided diagnosis with radiogenomics: analysis of the relationship between genotype and morphological changes of the brain magnetic resonance images

In the post-genome era, a novel research field, ‘radiomics’ has been developed to offer a new viewpoint for the use of genotypes in radiology and medicine research which have traditionally focused on the analysis of imaging phenotypes. The present study analyzed brain morphological changes related to the individual’s genotype. Our data consisted of magnetic resonance (MR) images of patients with mild cognitive impairment (MCI) and Alzheimer’s disease (AD), as well as their apolipoprotein E (APOE) genotypes. First, statistical parametric mapping (SPM) 12 was used for three-dimensional anatomical standardization of the brain MR images. A total of 30 normal images were used to create a standard normal brain image. Z-score maps were generated to identify the differences between an abnormal image and the standard normal brain. Our experimental results revealed that cerebral atrophies, depending on genotypes, can occur in different locations and that morphological changes may differ between MCI and AD. Using a classifier to characterize cerebral atrophies related to an individual’s genotype, we developed a computer-aided diagnosis (CAD) scheme to identify the disease. For the early detection of cerebral diseases, a screening system using MR images, called Brain Check-up, is widely performed in Japan. Therefore, our proposed CAD scheme would be used in Brain Check-up.