Gd-EOB-DTPA-enhanced magnetic resonance imaging combined with T1 mapping predicts the degree of differentiation in hepatocellular carcinoma.

BACKGROUND: Variable degrees of differentiation in hepatocellular carcinoma(HCC)under Edmondson-Steiner grading system has been proven to be an independent prognostic indicator for HCC. Up till now, there has been no effective radiological method that can reveal the degree of differentiation in HCC before surgery. This paper aims to evaluate the use of Gd-EOB-DTPA-Enhanced Magnetic Resonance Imaging combined with T1 mapping for the diagnosis of HCC and assessing its degree of differentiation. METHODS: Forty-four patients with 53 pathologically proven HCC had undergone Gd-EOB-DTPA enhanced MRI with T1 mapping before surgery. Out of the 53 lesions,13 were grade I, 27 were gradeII, and 13 were grade III. The T1 values of each lesion were measured before and at 20 min after Gd-EOB-DTPA administration (T1p and T1e). The absolute reduction in T1 value (T1d) and the percentage reduction (T1d %) were calculated. The one-way ANOVA and Pearson correlation were used for comparisons between the T1 mapping values. RESULTS: The T1d and T1d % of grade I, II and III of HCC was 660.5 ± 422.8ms、295.0 ± 99.6ms、276.2 ± 95.0ms and 54.0 ± 12.2 %、31.5 ± 6.9 %、27.7 ± 6.7 % respectively. The differences between grade Iand II, grade Iand III were statistically significant (p < 0.05), but there was no statically significant difference between grade II and III. The T1d % was the best marker for grading of HCC, with a Spearman correlation coefficient of -0.676. CONCLUSIONS: T1 mapping before and after Gd-EOB-DTPA administration can predict degree of differentiation in HCC.

Cholangiocarcinoma: spectrum of appearances on Gd-EOB-DTPA-enhanced MR imaging and the effect of biliary function on signal intensity.

BACKGROUND: To describe the Gd-EOB-DTPA-enhanced MRI appearances of cholangiocarcinoma, and evaluate the relative signal intensities (RSIs) changes of major abdominal organs, and investigate the effect of total bilirubin (TB) levels on the RSI. METHODS: 25 patients with pathologically-proven cholangiocarcinoma underwent Gd-EOB-DTPA-enhanced MRI. The visualization of the biliary system during biliary phase (BP) was observed. RSIs of the abdominal aorta (A), portal vein (V), liver (L), and spleen (S) were measured. RESULTS: On hepatocellular phase (HP), exophytic tumors (n =10) and infiltrative tumors (n =10) were hypointense, polypoid tumors (n = 2) were hypointense, and combined type tumors (n = 3) had mixed appearances. While patients with normal TB levels (≤22 µmol/L, n = 12) had clear visualization of the biliary tree during BP, those with elevated TB levels (>22 µmol/L, n = 13) had obscured or no visualization. In addition, patients with normal TB levels had higher RSIA, RSIV and RSIS than those with elevated TB levels on all dynamic phases (P <0.001), and lower RSIA, RSIV and RSIS on HP and BP (P <0.001). Patients with normal TB levels had higher RSIL than those with elevated TB levels on all phases (P <0.001). CONCLUSIONS: RSIs of major abdominal organs reflected underlying biliary function. Cholangiocarcinoma patients with elevated TB levels had delayed excretion of Gd-EOB-DTPA compared with patients with normal TB levels.

Evaluation of intestinal tuberculosis by multi-slice computed tomography enterography.

BACKGROUND: Multi-slice computed tomography enterography (MSCTE) is now widely used to diagnose and monitor intestinal disease. Preliminary studies suggest that MSCTE may be useful in detecting intestinal tuberculosis (ITB). We sought to assess the use of MSCTE for the diagnosis of ITB in our medical center. METHODS: In this retrospective study, 15 patients (11 males and 4 females, 6 to 65 years old) were enrolled and diagnosed with ITB by MSCTE. Diagnosis were confirmed by pathology or clinical criteria. Two experienced abdominal radiologists evaluated the images and defined the location, number, shape, edge, surrounding tissue alterations of ITB and other associated changes in the peritoneum, mesentery and solid abdominal organs. RESULTS: The interval between the onset of symptoms and diagnosis varied from 20 days to 10 years. The most common symptom was abdominal pain (80 %). The ileocecum was the most common site affected by ITB (87 %). Morphological MSCTE findings were variable and included multi-segmental symmetric intestinal mural thickening found in 6 patients (40 %), solid masses found in 9 patients (60 %), and enlarged lymph nodes (LNs) found in 13 (87 %) patients. Non-enhancing central necrosis and rim enhancement were noted in 10 patients (67 %). CONCLUSIONS: Characteristic MSCTE findings of ITB include solid mass or multi-segmental symmetric mural thickening involving the ileocecal area and rim enhanced LNs. Knowledge of these features in combination with a high index of suspicion can be useful in early diagnosis of ITB.

PubMed-supported clinical term weighting approach for improving inter-patient similarity measure in diagnosis prediction.

BACKGROUND: Similarity-based retrieval of Electronic Health Records (EHRs) from large clinical information systems provides physicians the evidence support in making diagnoses or referring examinations for the suspected cases. Clinical Terms in EHRs represent high-level conceptual information and the similarity measure established based on these terms reflects the chance of inter-patient disease co-occurrence. The assumption that clinical terms are equally relevant to a disease is unrealistic, reducing the prediction accuracy. Here we propose a term weighting approach supported by PubMed search engine to address this issue. METHODS: We collected and studied 112 abdominal computed tomography imaging examination reports from four hospitals in Hong Kong. Clinical terms, which are the image findings related to hepatocellular carcinoma (HCC), were extracted from the reports. Through two systematic PubMed search methods, the generic and specific term weightings were established by estimating the conditional probabilities of clinical terms given HCC. Each report was characterized by an ontological feature vector and there were totally 6216 vector pairs. We optimized the modified direction cosine (mDC) with respect to a regularization constant embedded into the feature vector. Equal, generic and specific term weighting approaches were applied to measure the similarity of each pair and their performances for predicting inter-patient co-occurrence of HCC diagnoses were compared by using Receiver Operating Characteristics (ROC) analysis. RESULTS: The Areas under the curves (AUROCs) of similarity scores based on equal, generic and specific term weighting approaches were 0.735, 0.728 and 0.743 respectively (p < 0.01). In comparison with equal term weighting, the performance was significantly improved by specific term weighting (p < 0.01) but not by generic term weighting. The clinical terms "Dysplastic nodule", "nodule of liver" and "equal density (isodense) lesion" were found the top three image findings associated with HCC in PubMed. CONCLUSIONS: Our findings suggest that the optimized similarity measure with specific term weighting to EHRs can improve significantly the accuracy for predicting the inter-patient co-occurrence of diagnosis when compared with equal and generic term weighting approaches.

Functional magnetic resonance cholangiography enhanced with Gd-EOB-DTPA: effect of liver function on biliary system visualization.

PURPOSE: To evaluate effect of liver function on biliary system visualization using gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance cholangiography (CE-MRC). MATERIALS AND METHODS: In all, 39 patients were divided into three groups according to their Child-Pugh classification: group A, Child-Pugh class A (23); group B, class B (11); and group C, class C (5). They underwent Gd-EOB-DTPA CE-MRC. Biliary system visualization was qualitatively rated on a 5-point scale. Relative signal intensity (RSI) of common bile duct (CBD) and liver was quantitatively measured. Laboratory findings and the Model of Endstage Liver Disease (MELD) score were recorded. RESULTS: Visualization ratings of CBD, left hepatic duct, right hepatic duct, segmental branches of intrahepatic ducts, cystic duct, and gallbladder of group A were: 3.61 ± 0.58, 2.87 ± 0.97, 2.96 ± 0.77, 1.17 ± 0.58, 3.04 ± 0.83, 3.00 ± 0.95, respectively; group B: 2.00 ± 0.61, 1.09 ± 0.64, 0.91 ± 0.54, 0.27 ± 0.13, 1.36 ± 0.62, 1.45 ± 0.54, respectively; group C: 1.40 ± 0.73, 1.00 ± 0.51, 1.00 ± 0.51, 0.00 ± 0.00, 0.60 ± 0.39, 0.60 ± 0.39, respectively. RSI of CBD of groups A to C were 17.12 ± 0.41, 3.95 ± 0.63, 3.33 ± 0.30, respectively. RSI of liver of groups A to C were 6.73 ± 0.72, 2.53 ± 1.02, 2.05 ± 0.11, respectively. CE-MRC images of group A were significantly better than those of group B and C in terms of both visualization ratings and RSI of CBD. CBD RSI positively correlated with liver RSI (r = 0.99, P < 0.001). The total serum bilirubin level and MELD score were significant predictors of RSI of CBD. CONCLUSION: Different liver function according to Child-Pugh classification significantly affects biliary system visualization of Gd-EOB-DTPA CE-MRC.

Pancreatic neuroendocrine tumours: correlation between MSCT features and pathological classification.

OBJECTIVES: We aimed to evaluate the multi-slice computed tomography (MSCT) features of pancreatic neuroendocrine neoplasms (P-NENs) and analyse the correlation between the MSCT features and pathological classification of P-NENs. METHODS: Forty-one patients, preoperatively investigated by MSCT and subsequently operated on with a histological diagnosis of P-NENs, were included. Various MSCT features of the primary tumour, lymph node, and distant metastasis were analysed. The relationship between MSCT features and pathologic classification of P-NENs was analysed with univariate and multivariate models. RESULTS: Contrast-enhanced images showed significant differences among the three grades of tumours in the absolute enhancement (P = 0.013) and relative enhancement (P = 0.025) at the arterial phase. Univariate analysis revealed statistically significant differences among the tumours of different grades (based on World Health Organization [WHO] 2010 classification) in tumour size (P = 0.001), tumour contour (P < 0.001), cystic necrosis (P = 0.001), tumour boundary (P = 0.003), dilatation of the main pancreatic duct (P = 0.001), peripancreatic tissue or vascular invasion (P < 0.001), lymphadenopathy (P = 0.011), and distant metastasis (P = 0.012). Multivariate analysis suggested that only peripancreatic tissue or vascular invasion (HR 3.934, 95 % CI, 0.426-7.442, P = 0.028) was significantly associated with WHO 2010 pathological classification. CONCLUSIONS: MSCT is helpful in evaluating the pathological classification of P-NENs. KEY POINTS: • P-NENs are potentially malignant, and classification of P-NENs carries important prognostic value. • MSCT plays an important role in the diagnosis and staging of P-NENs. • Correlations between classification of P-NENs and imaging findings have not been systematically evaluated. • MSCT could predict P-NENs classification and may be a useful prognostication tool.

CT evaluation of gastroenteric neuroendocrine tumors: relationship between ct features and the pathologic classification.

OBJECTIVE: The objective of our study was to compare the CT features of gastroenteric neuroendocrine neoplasms (GE-NENs) with the pathologic classification and to analyze the correlation between the CT features and classification of GE-NENs. MATERIALS AND METHODS: Fifty-six cases of pathologically and immunohisto-chemically proven GE-NENs, including 25 cases of neuroendocrine tumors (NETs) (i.e., G1 and G2 tumors) and 31 cases of neuroendocrine carcinomas (NECs) (i.e., G3 tumors and mixed adenoneuroendocrine carcinomas) were studied. We analyzed various CT features of the primary tumor, nodal status, and metastasis and compared these features with pathologic grading. RESULTS: The CT features that favor NEC over NET include larger tumor size (> 4.0 cm), transmural invasion, circumscribed tumor with both intra- and extraluminal involvement, circumferential growth, areas of cystic change or necrosis, ulceration, mesenteric fat infiltration, and lymphadenopathy, with p values of 0.044, 0.002, 0.024, 0.008, 0.002, 0.007, 0.002, and < 0.001, respectively. The CT features that do not distinguish between the two types of GE-NENs include tumor boundary, growth pattern, degree of enhancement, adjacent organ invasion, distant organ metastasis, and peritoneal seeding, with p values of 0.277, 0.153, 0.672, 1.000, 0.159, and 0.877, respectively. CONCLUSION: CT can be useful in the classification of GE-NENs.

Automatic selective removal of embedded patient information from image content of DICOM files.

OBJECTIVE: Patient data may appear as burned-in text on the image content of DICOM image files, which is commonly seen in ultrasound images and some secondarily scanned images. The purpose of this article is to discuss the removal of such information, which can be cumbersome and demands human intervention. CONCLUSION: This article presents a computerized scheme that automatically removes patient data from the image content by finding burned-in texts that match information in the DICOM header, thus cleaning the entire DICOM image file while preserving other useful labels.

Nasopharyngeal carcinoma: investigation of intratumoral heterogeneity with FDG PET/CT.

OBJECTIVE: The purpose of this study was to quantitatively evaluate the role of intratumoral heterogeneity of (18)F-FDG uptake in characterizing nasopharyngeal carcinoma (NPC). SUBJECTS AND METHODS: Forty consecutively registered patients with newly diagnosed NPC underwent PET/CT. The heterogeneity factor, defined as the derivative of a volume threshold function, was computed for each tumor. The relations between heterogeneity factor and maximum standardized uptake value (SUV(max)), tumor volume, and TNM category were determined by two-tailed Spearman correlation. Factors that potentially affect outcome determined by disease-free survival were studied by Kaplan-Meier analysis with a log-rank test for univariate analysis and the Cox proportional hazard model for multivariate analysis. RESULTS: The heterogeneity factor ranged from -1.80 to -0.13 (mean, -0.40 [SD, 0.40]) and significantly correlated with SUV(max) (r = -0.372; p = 0.018), tumor volume (r = -0.983; p < 0.001), and T category (r = -0.457; p = 0.003) but not with N and M categories. There was a significant difference in heterogeneity factor between T1 and T2 tumors and T3 and T4 tumors (p = 0.012). The 2-year disease-free survival rate among the 38 patients was 67.4%. According to the results of Kaplan-Meier analysis with the log-rank test, heterogeneity factor and M category significantly affected disease-free survival. Patients with tumors that had a heterogeneity factor greater than -0.24 (less-heterogeneous group) (p = 0.0498) or M0 status (p < 0.001) had better disease-free survival rates. Multivariate analysis showed only M category to be an independent predictor of disease-free survival (p < 0.001). CONCLUSION: The intratumoral heterogeneity of FDG uptake varies across NPC tumors, significantly correlates with tumor aggressiveness, and is predictive of patient outcome. These findings may be useful for characterizing NPC, predicting survival, and improving patient care.

Use of intravoxel incoherent motion MRI to assess placental perfusion in normal and Fetal Growth Restricted pregnancies on their third trimester.

INTRODUCTION: Intravoxel Incoherent Motion (IVIM) MRI is a non-invasive, in vivo techniques which can assess placental perfusion quantitatively, and be useful for evaluating placental microcirculation. Our primary aim was to investigate whether fetal growth restriction (FGR) pregnancies have different placental perfusion and diffusion compared with normal pregnancies using IVIM. A secondary aim was to investigate correlations between placental IVIM parameters and gestational age in normal pregnancy. METHODS: This study population included 17 FGR pregnancies and 36 normal pregnancies between 28 + 3 to 38 + 0 weeks. All women underwent a MRI examination including an IVIM sequence with 9 b-values on a 3.0 T MRI system. The standard diffusion coefficeint (D), pseudodiffusion (D*) and perfusion fraction (f) were calculated. RESULTS: Placental f was significantly lower in the FGR group than that in the normal group (33.96 ± 2.62(%) vs 38.48 ± 5.31(%), p = 0.002). Placental D and D* in two groups showed no statistical significance (P > 0.05). Placental f moderately increased with increasing gestational age in normal pregnancies (r = 0.411, p = 0.013), and there existed a negative correlation between D values and gestational age (r = -0.390, p = 0.019). DISCUSSION: The f values are able to distinguish FGR from normal pregnancies. It can be uses as a feasible index to evaluate placenta perfusion. Gestational age-associated changes in placental IVIM parameters likely reveal trajectories of microvascular perfusion fraction and diffusion characteristics in the normal developing placenta.

Predicting anaplastic lymphoma kinase rearrangement status in patients with non-small cell lung cancer using a machine learning algorithm that combines clinical features and CT images.

Purpose: To develop an appropriate machine learning model for predicting anaplastic lymphoma kinase (ALK) rearrangement status in non-small cell lung cancer (NSCLC) patients using computed tomography (CT) images and clinical features. Method and materials: This study included 193 patients with NSCLC (154 in the training cohort, 39 in the validation cohort), 68 of whom tested positive for ALK rearrangements and 125 of whom tested negative. From the nonenhanced CT scans, 157 radiomic characteristics were extracted, and 8 clinical features were collected. Five machine learning (ML) models were assessed to find the best classification model for predicting ALK rearrangement status. A radiomic signature was developed using the least absolute shrinkage and selection operator (LASSO) algorithm. The predictive performance of the models based on radiomic features, clinical features, and their combination was assessed by receiver operating characteristic (ROC) curves. Results: The support vector machine (SVM) model had the highest AUC of 0.914 for classification. The clinical features model had an AUC=0.805 (95% CI 0.731-0.877) and an AUC=0.735 (95% CI 0.566-0.863) in the training and validation cohorts, respectively. The CT image-based ML model had an AUC=0.953 (95% CI 0.913-1.0) in the training cohort and an AUC=0.890 (95% CI 0.778-0.971) in the validation cohort. For predicting ALK rearrangement status, the ML model based on CT images and clinical features performed better than the model based on only clinical information or CT images, with an AUC of 0.965 (95% CI 0.826-0.882) in the primary cohort and an AUC of 0.914 (95% CI 0.804-0.893) in the validation cohort. Conclusion: Our findings revealed that ALK rearrangement status could be accurately predicted using an ML-based classification model based on CT images and clinical data.

Whole-body diffusion-weighted imaging: the added value to whole-body MRI at initial diagnosis of lymphoma.

OBJECTIVE: The objective of our study was to evaluate the diagnostic performance of conventional whole-body MRI without and with diffusion-weighted imaging (DWI) in the detection of known (18)F-FDG-avid lymphomas. The conventional whole-body MRI protocol consisted of a T2-weighted sequence and a T2-weighted spectral attenuated inversion recovery (SPAIR) sequence with frequency-selective fat suppression. The second protocol used the same sequences as the first protocol but also included DWI. SUBJECTS AND METHODS: Seventeen patients with pathologically confirmed, newly diagnosed, untreated lymphoma were recruited. T2-weighted and T2-weighted SPAIR images were evaluated first, separate from the DW images, and then were evaluated with the DW images. We used (18)F-FDG PET/CT as the standard of reference. True-positive, false-positive, and false-negative values were evaluated on a per-lesion basis. Tumor staging based on T2-weighted and T2-weighted SPAIR imaging without DWI and then with DWI was compared using the Ann Arbor staging system. RESULTS: True-positive lesions were increased from 89% to 97%, false-positive lesions were increased from 3% to 6%, and false-negative lesions were decreased from 11% to 3% by the addition of DWI. Diagnostic sensitivity was significantly increased (p = 0.002) by adding DWI. Lesions detected on DWI but not on T2-weighted and T2-weighted SPAIR imaging were located in renal (n = 1), paraaortic (n = 6), and pelvic (n = 3) lymph nodes. On DWI, 47% of the lesions (n = 55) were more conspicuous than on T2-weighted and T2-weighted SPAIR imaging; most of these lesions (58%, n = 32) were from lymph nodes in the pelvic or abdominal regions and bone marrow. No difference was found between T2-weighted and T2-weighted SPAIR imaging without DWI and T2-weighted and T2-weighted SPAIR imaging with DWI in lymphoma staging, being consistent with PET/CT in 88% of the patients (n = 15). CONCLUSION: The addition of DWI to conventional whole-body MRI sequences enhanced lesion conspicuity and improved diagnostic accuracy for lymphomas. With technical optimization, whole-body MRI with DWI, as a nonionizing imaging modality, may potentially be useful as an alternative method to PET/CT in the management of malignant lymphoma.

Development and Validation of a Prognostic Nomogram Based on Residual Tumor in Patients With Nondisseminated Nasopharyngeal Carcinoma.

OBJECTIVES: To investigate the prognostic value of residual tumor based on Magnetic resonance imaging(MRI) and establish an effective prognostic nomogram model referring to clinical,pathological and other related factors for predicting prognosis in nasopharyngeal carcinoma. METHODS: Overall, 538 patients with non-metastatic, histologically-confirmed nasopharyngeal carcinoma were retrospectively examined. Data from 397 patients were used for the construction and validation of a nomogram based on the presence of residual tumor. A concordance index (C-index) was employed to assess the predictive accuracy and discriminative ability of the nomogram. RESULTS: The 3-year survival rates in the non-residual and residual tumor cohorts were as follows: progression-free survival, 73.4% vs. 61.0%, P = 0.009; locoregional recurrence-free survival, 81.9% vs. 72.0%, P = 0.02; and distant metastasis-free survival, 80.7% vs. 73.5%, P = 0.11. Nine significant factors were included in the nomogram model. The calibration curve for the probability of progression-free survival showed that the nomogram-based predictive values had good concordance with the actual observations. CONCLUSION: The results showed that the patients in the residual tumor cohorts had a worse prognosis.The proposed nomogram may predict the prognosis and guide clinical decision-making concerning local residual tumors in nasopharyngeal carcinoma patients. Patients with a high risk of progression require more timely and aggressive treatment.

Effect of a computer-aided diagnosis system on clinicians' performance in detection of small acute intracranial hemorrhage on computed tomography.

RATIONALE AND OBJECTIVES: To analyze the effect of a computer-aided diagnosis (CAD) system on clinicians' performance in detection of small acute intracranial hemorrhage (AIH) on computed tomography (CT). MATERIALS AND METHODS: The authors have developed a CAD scheme that used both image processing techniques and anatomic knowledge based classification system to improve diagnosis of small AIH on CT. A multiple-reader, multiple-case receiver operating characteristic (ROC) study was performed. Twenty clinicians, including seven emergency physicians, seven radiology residents, and six radiology specialists were recruited as readers of 60 sets of brain CT, including 30 cases that show AIH smaller than 1 cm, and 30 controls. Each reader read the same 60 cases twice, first without, then with the prompts produced by the CAD system. The clinicians ranked their confidence in diagnosing a case of showing AIH, which produced the ROC curves. RESULTS: Significantly improved performance is observed in emergency physicians, average area under the ROC curve (Az) increased from 0.8422 to 0.9294 (P = .0107) when they make the diagnosis without and with the support of CAD. Az for radiology residents increased from 0.9371 to 0.9762 (P = .0088). Az for radiology specialists increased from 0.9742 to 0.9868, but was statistically insignificant (P = .1755). CONCLUSIONS: CAD can improve the clinicians' performance in detecting AIH on CT. In particular, emergency physicians can benefit most from the CAD and improve their performance to a level approaching that of the average radiology residents.

[Three-dimensional spatial measurement versus conventional CT planning in laparoscopic partial nephrectomy for renal tumors].

OBJECTIVE: To analyze the advantages of spatial measurement of anatomical parameters in a 3D model in surgical planning for laparoscopic partial nephrectomy (LPN). METHODS: From February, 2016 to October, 2017, 37 patients diagnosed with T1 renal mass underwent LPN based on 3D reconstruction after enhanced CT scanning using the Uromedix-3D system (group A), and another 38 patients received LPN with conventional CT planning (group B). The anatomical parameters were measured in the reconstructed 3D model and the demographic data, surgical outcome and postoperative data were compared between the two groups. RESULTS: In group A, the average time for 3D model reconstruction was (29.3∓9.7) min; the length, width and depth of the renal defect in 3D model were 3.2∓1.1 cm, 2.6∓0.9 cm and 1.7∓0.7 cm, respectively; The distance of the tumor from the collecting system was 3.8∓2.2 mm; The mean R.E.N.A.L score of the patients was 7∓1.5, and 3 patients had accessory renal artery and 2 had early branching of the renal artery. LPNs were completed via the retroperitoneal approach in all the 75 patients without conversion to open or total nephrectomy. Group A and group B showed significant differences in warm ischemic time (26.7∓6.4 vs 31.9∓7.0 min), tumor-excision time (8.4∓2.6 vs 10.4∓2.8 min), renal defect suture time (18.3∓3.9 vs 21.5∓3.4 min), 24-h volume of retroperitoneal drainage (88.6∓40.2 vs 134.3∓58.3 mL) and 48-h volume of retroperitoneal drainage (127.9∓54.5 vs 198.1∓86.3 mL), but not in the demographic data, operation time, intraoperative blood loss or postoperative hospital stay. CONCLUSIONS: 3D reconstruction of the renal masses can be completed efficiently and accurately using this system. Compared with conventional CT-based measurement, 3D spatial measurement of the anatomical structures helps to increase the precision in the performance of LPN and reduce the warm ischemia time.

Computer aided detection of small acute intracranial hemorrhage on computer tomography of brain.

INTRODUCTION: Detection of acute intracranial hemorrhage (AIH) is a primary task in image interpretation of computer tomography (CT) of brain for patients suffering from acute neurological disturbance or head injury. Although CT readily depicts AIH, interpretation can be difficult especially when the lesion is inconspicuous or the reader is inexperienced. OBJECTIVE: To develop a computer aided detection system that improves diagnostic accuracy of small AIH on brain CT. MATERIALS AND METHODS: Intracranial contents are first segmented by thresholding and morphological operations, which are then subjected to denoising and adjustment for CT cupping artifacts. The brain is then automatically realigned into normal position. AIH candidates are extracted based on top-hat transformation and left-right asymmetry. AIH candidates are registered against a normalized coordinate system such that the candidates are rendered anatomical information. True AIH is differentiated from mimicking normal variants or artifacts by a knowledge-based classification system incorporating rules that make use of quantified imaging features and anatomical information. A total of 186 clinical cases, including 62 CT studies showing small (<1cm) AIH, and 124 controls, were retrospectively collected. Forty positive cases and 80 controls were used for the training of the CAD. Twenty-two positive cases and 44 controls were used in the validation of the CAD system. Regions of AIH identified by two experienced radiologists were used as gold standard. The size of individual AIH volume was also recorded. RESULTS: On a per patient basis, the system achieved sensitivity of 95% (38/40) and specificity of 88.8% (71/80) in the training dataset. The sensitivity and specificity were 100% (22/22) and 84.1% (37/44) respectively for the diagnosis of AIH in the validation cases. Individual cases contained variable number of AIH volumes. There were 77 lesions in the 40 training cases and 46 lesions in the 22 validation cases. On a per lesion basis, the sensitivities were 84.4% (65/77) and 82.6% (38/46) for all lesions 10mm or smaller for the training and validation datasets, respectively. False positive rates were 0.19 (23/120) and 0.29 (19/66) false positive lesion per case for the training and validation datasets, respectively. CONCLUSION: This study demonstrated that CAD is valuable for detection of small AIH on brain CT.

Quantitative evaluation of Gd-EOB-DTPA uptake in focal liver lesions by using T1 mapping: differences between hepatocellular carcinoma, hepatic focal nodular hyperplasia and cavernous hemangioma.

OBJECTIVES: To investigate the difference of T1 relaxation time on Gd-EOB-DTPA-enhanced MRI in hepatocellular carcinoma (HCC), hepatic focal nodular hyperplasia (FNH) and cavernous hemangioma of liver (CHL), and to quantitatively evaluate the uptake of Gd-EOB-DTPA in these three focal liver lesions (FLLs). RESULTS: The T1P of CHL was significantly higher than those of HCC and FNH (P < 0.05). Reduction of T1 relaxation time on hepatobiliary phase could be observed in all three types of lesions. There were significant differences of T1P, T1E, T1D and T1D% between FNH, CHL and HCC (P < 0.001). Spearman correlation analysis revealed that T1D% was the best indicator for diagnostic differentiation, with a correlation coefficient of 0.702. Discriminant analysis using three variables (T1P, T1E, and T1D%) showed that the classification accuracy was 88.2%. MATERIALS AND METHODS: 74 patients diagnosed with focal liver lesions underwent Gd-EOB-DTPA-enhanced MRI including T1 mapping were enrolled, consisting of 51 HCCs, 10 FNHs, and 13 CHLs. T1 relaxation times of these lesions were measured on pre-contrast (T1P) and on hepatobiliary phase images at 20 minute after contrast (T1E). The reduction of T1 relaxation time on hepatobiliary (T1D) and the percentage reduction (T1D%) was calculated. The differences of T1P, T1E, T1D and T1D% in these FLLs were analyzed. The usefulness of these parameters for classification of FLLs was evaluated. CONCLUSIONS: Uptake of Gd-EOB-DTPA is different between in HCC, FNH and CHL. These three lesions can be distinguished using T1 mapping.

Genetic algorithm supported by graphical processing unit improves the exploration of effective connectivity in functional brain imaging.

Brain regions of human subjects exhibit certain levels of associated activation upon specific environmental stimuli. Functional Magnetic Resonance Imaging (fMRI) detects regional signals, based on which we could infer the direct or indirect neuronal connectivity between the regions. Structural Equation Modeling (SEM) is an appropriate mathematical approach for analyzing the effective connectivity using fMRI data. A maximum likelihood (ML) discrepancy function is minimized against some constrained coefficients of a path model. The minimization is an iterative process. The computing time is very long as the number of iterations increases geometrically with the number of path coefficients. Using regular Quad-Core Central Processing Unit (CPU) platform, duration up to 3 months is required for the iterations from 0 to 30 path coefficients. This study demonstrates the application of Graphical Processing Unit (GPU) with the parallel Genetic Algorithm (GA) that replaces the Powell minimization in the standard program code of the analysis software package. It was found in the same example that GA under GPU reduced the duration to 20 h and provided more accurate solution when compared with standard program code under CPU.