Region Of Interest Measurements Research Articles

Machine Learning Analysis in Diffusion Kurtosis Imaging for Discriminating Pediatric Posterior Fossa Tumors: A Repeatability and Accuracy Pilot Study.

Background and purpose: Differentiating pediatric posterior fossa (PF) tumors such as medulloblastoma (MB), ependymoma (EP), and pilocytic astrocytoma (PA) remains relevant, because of important treatment and prognostic implications. Diffusion kurtosis imaging (DKI) has not yet been investigated for discrimination of pediatric PF tumors. Estimating diffusion values from whole-tumor-based (VOI) segmentations may improve diffusion measurement repeatability compared to conventional region-of-interest (ROI) approaches. Our purpose was to compare repeatability between ROI and VOI DKI-derived diffusion measurements and assess DKI accuracy in discriminating among pediatric PF tumors. Materials and methods: We retrospectively analyzed 34 children (M, F, mean age 7.48 years) with PF tumors who underwent preoperative examination on a 3 Tesla magnet, including DKI. For each patient, two neuroradiologists independently segmented the whole solid tumor, the ROI of the area of maximum tumor diameter, and a small 5 mm ROI. The automated analysis pipeline included inter-observer variability, statistical, and machine learning (ML) analyses. We evaluated inter-observer variability with coefficient of variation (COV) and Bland-Altman plots. We estimated DKI metrics accuracy in discriminating among tumor histology with MANOVA analysis. In order to account for class imbalances, we applied SMOTE to balance the dataset. Finally, we performed a Random Forest (RF) machine learning classification analysis based on all DKI metrics from the SMOTE dataset by partitioning 70/30 the training and testing cohort. Results: Tumor histology included medulloblastoma (15), pilocytic astrocytoma (14), and ependymoma (5). VOI-based measurements presented lower variability than ROI-based measurements across all DKI metrics and were used for the analysis. DKI-derived metrics could accurately discriminate between tumor subtypes (Pillai's trace: p < 0.001). SMOTE generated 11 synthetic observations (10 EP and 1 PA), resulting in a balanced dataset with 45 instances (34 original and 11 synthetic). ML analysis yielded an accuracy of 0.928, which correctly predicted all but one lesion in the testing set. Conclusions: VOI-based measurements presented improved repeatability compared to ROI-based measurements across all diffusion metrics. An ML classification algorithm resulted accurate in discriminating PF tumors on a SMOTE-generated dataset. ML techniques based on DKI-derived metrics are useful for the discrimination of pediatric PF tumors.

Correlation between the region of interest in digital radiography, Hounsfield units, and histological maturation on Wistar rats submitted to tibial fracture

Objective: Study the relationship between the region of interest (ROI) and Hounsfield units (HU) in the tibial bone callus. Methods: Twenty-one adult Wistar rats were submitted to tibial fracture. The fracture was radiographed after their euthanasia, and the bone calluses were analyzed histologically after being stained with hematoxylin and eosin. Euthanasia occurred between the 2nd and 6th weeks after fracture and fixation, thus obtaining various consolidation stages. Histologically, vessels, chondroblasts, connective tissue (collagen), maximum size of the chondrocyte, and the concentration of chondrocytes were quantified. Results: It was observed that the higher the HU, the more mature and closer to bone consolidation it is, proving that the use of ROI and bone callus measurement with HU is reliable for the histological process of maturation of the bone callus and can be safely used as proof of evolution of the bone healing process. Conclusion: The ROI was successfully used in digital radiography to observe HU in fractured bones. Level of Evidence IV; Therapeutic Studies; Case Series.

Advanced deep learning-based image reconstruction in lumbar spine MRI at 0.55 T – Effects on image quality and acquisition time in comparison to conventional deep learning-based reconstruction

ObjectivesTo evaluate an optimized deep leaning-based image post-processing technique in lumbar spine MRI at 0.55 T in terms of image quality and image acquisition time. Materials and methodsLumbar spine imaging was conducted on 18 patients using a 0.55 T MRI scanner, employing conventional (CDLR) and advanced (ADLR) deep learning-based post-processing techniques. Two musculoskeletal radiologists visually evaluated the images using a 5-point Likert scale to assess image quality and resolution. Quantitative assessment in terms of signal intensities (SI) and contrast ratios was performed by region of interest measurements in different body-tissues (vertebral bone, intervertebral disc, spinal cord, cerebrospinal fluid and autochthonous back muscles) to investigate differences between CDLR and ADLR sequences. ResultsThe images processed with the advanced technique (ADLR) were rated superior to the conventional technique (CDLR) in terms of signal/contrast, resolution, and assessability of the spinal canal and neural foramen. The interrater agreement was moderate for signal/contrast (ICC = 0.68) and good for resolution (ICC = 0.77), but moderate for spinal canal and neuroforaminal assessability (ICC = 0.55). Quantitative assessment showed a higher contrast ratio for fluid-sensitive sequences in the ADLR images. The use of ADLR reduced image acquisition time by 44.4%, from 14:22 min to 07:59 min. ConclusionsAdvanced deep learning-based image reconstruction algorithms improve the visually perceived image quality in lumbar spine imaging at 0.55 T while simultaneously allowing to substantially decrease image acquisition times. Clinical relevanceAdvanced deep learning-based image post-processing techniques (ADLR) in lumbar spine MRI at 0.55 T significantly improves image quality while reducing image acquisition time.

Does volumetric measurement of ADC values achieve higher diagnostic performance in bladder cancer MRI?

Apparent diffusion coefficient (ADC) value is an important part of bladder cancer magnetic resonance imaging (MRI) assessment and can predict the aggressive and invasive potentials. There is growing interest in whole tumor volume measurements. To investigate if the volumetric ADC measurement method will significantly exceed the diagnostic performance of the selected region of interest (ROI) method in everyday practice. A prospective evaluation was carried out of 50 patients with bladder cancer by two radiologists. The mean and the minimum ADC values were measured using both methods. The inter-reader agreement was determined by the intraclass correlation coefficient. The ADC values were compared between different grades, states of muscle invasion, and lympho-vascular invasion (LVI); then, validity was evaluated using receiver operating characteristic (ROC) curves. Areas under the curve (AUC) were then compared for the level of statistical significance. The inter-observer agreement was excellent for the ADC values using both methods. The volumetric measurement provides higher mean and lower minimum ADC values with statistically significant differences (P <0.00001). The highest diagnostic accuracy for differentiating tumor grade and predicting muscle invasion was for the minimum ADC by a selected ROI. However, the differences between the achieved AUCs were of no statistical significance. None of the ADC values predicted LVI with statistical significance. The selected ROI and volumetric measurement methods of mean and minimum ADC in bladder cancer yield different values, still having comparable diagnostic performance with accurate ROI sampling. The minimum ADC value by ROI is preferred in everyday clinical practice.

To assess the quantitative features of focal liver lesions in gadoxetic acid enhanced MRI and to determine whether these features can accurately differentiate benign form malignant lesions

PurposeThe study aims at assessing the quantitative features which distinguish focal liver lesions (FLLs) in gadoxetic acid (GA) enhanced liver MRI and at determining whether these features can accurately differentiate benign from malignant lesions. Material and methods107 patients with 180 unequivocal FLLs in previous examinations were included in a single-center retrospective study. All patients underwent a MRI test of the liver with GA. 99 benign and 74 malignant lesions were included. The group of benign lesions consisted of 60 focal nodular hyperplasias (FNH), 22 hemangiomas (HMG), 6 hepatic adenomas (HA), and 11 other benign lesions (1 angiomyolipioma, 6 lesions histopathology diagnoses as benign without further specification, or ones lacking features of malignancy, and 4 lesions radiologically diagnosed as benign which remained stable in the follow-up studies). The group of malignant lesions consisted of primary 51 hepatocellular carcinomas, 12 metastases, and 11 metastases from melanoma malignum (MM meta). 7 FLLs were excluded (4 cases of uncertain histopathological diagnosis, 2 cholangiocarcinomas, and 1 regenerative nodule). For the included lesions ROI (region of interest) measurements were taken by two observers in the T2-w, ADC (apparent diffusion coefficient) and in the T1-w sequence in the hepatobiliary phase (HBP). The interobserver agreement was evaluated with the Wilcoxon test. The Kruskal – Wallis, Mann – Whitney U and post hoc Dunn’s tests were applied to assess if there were any significant differences in the ROI values between individual lesions.The variables with the p values of < 0.05 were considered statistically significant. ResultsWe found significant differences in the ROI values between lesions with p < 0.0001. Strikingly high ROI values in the T2-w sequence were found for HMG. The lowest ADC values were encountered for metastases and MM metastases. The highest ROI values in the HBP were found for FNH, and the lowest for metastases.We also found statistically significant differences in the ROI values between benign and malignant lesions with benign lesions presenting statistically higher ROI values compared to malignant lesions. ConclusionsThere were significant differences in the ROI values among different types of FLLs. The predominant quantitative feature in the T2-w sequence was a strikingly high ROI value for HMG. Benign lesions presented statistically higher ROI values in the T2-w, ADC, and HBP sequences compared to malignant lesions. This was true for all lesions except for HA.

Retrospective T2 quantification from conventional weighted MRI of the prostate based on deep learning.

To develop a deep learning-based method to retrospectively quantify T2 from conventional T1- and T2-weighted images. Twenty-five subjects were imaged using a multi-echo spin-echo sequence to estimate reference prostate T2 maps. Conventional T1- and T2-weighted images were acquired as the input images. A U-Net based neural network was developed to directly estimate T2 maps from the weighted images using a four-fold cross-validation training strategy. The structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), mean percentage error (MPE), and Pearson correlation coefficient were calculated to evaluate the quality of network-estimated T2 maps. To explore the potential of this approach in clinical practice, a retrospective T2 quantification was performed on a high-risk prostate cancer cohort (Group 1) and a low-risk active surveillance cohort (Group 2). Tumor and non-tumor T2 values were evaluated by an experienced radiologist based on region of interest (ROI) analysis. The T2 maps generated by the trained network were consistent with the corresponding reference. Prostate tissue structures and contrast were well preserved, with a PSNR of 26.41 ± 1.17 dB, an SSIM of 0.85 ± 0.02, and a Pearson correlation coefficient of 0.86. Quantitative ROI analyses performed on 38 prostate cancer patients revealed estimated T2 values of 80.4 ± 14.4 ms and 106.8 ± 16.3 ms for tumor and non-tumor regions, respectively. ROI measurements showed a significant difference between tumor and non-tumor regions of the estimated T2 maps (P < 0.001). In the two-timepoints active surveillance cohort, patients defined as progressors exhibited lower estimated T2 values of the tumor ROIs at the second time point compared to the first time point. Additionally, the T2 difference between two time points for progressors was significantly greater than that for non-progressors (P = 0.010). A deep learning method was developed to estimate prostate T2 maps retrospectively from clinically acquired T1- and T2-weighted images, which has the potential to improve prostate cancer diagnosis and characterization without requiring extra scans.

Crohn's disease inflammation severity assessment with iodine density from photon counting CT enterography: comparison with endoscopic histopathology.

To determine optimal iodine density thresholds for active inflammation in CD patients with PCCT enterography and determine if iodine density can be used to stratify CD activity severity. A retrospective PACS search identified patients with CD imaged with PCCT enterography from 4/11/2022 to 10/30/2022 and with clinical notes, endoscopic/surgical pathology and available source PCCT data for iodine density analysis. Two abdominal radiologists with expertise in CD each drew two region of interest measurements within the visibly most affected region of terminal or neoterminal ileum wall on commercially available system (SyngoVia). Radiologists were blinded to clinical information and pathologic findings. Disease activity and severity were recorded from the pathology report. Harvey-Bradshaw Index, medications, and laboratory values were recorded. Receiver operating characteristic (ROC) curves were utilized to determine the optimum iodine density threshold for active inflammation and mild versus moderate-to-severe inflammation. Intra- and inter-reader agreement was assessed by intra-class correlation coefficient (ICC). 23 CD patients (15 females; mean [SD] age: 52 [17] years) imaged with PCCT enterography were included. 15/23 had active inflammation: 9/15 mild, 4/15 moderate, and 2/15 severe active inflammation. The optimal iodine density threshold for active inflammation was 2.7 mg/mL, with 97% sensitivity, 100% specificity, and 98% accuracy (AUC = 1.00). The optimal iodine density threshold for distinguishing mild from moderate-to-severe inflammation was 3.4 mg/mL, with 83% sensitivity, 89% specificity, and 87% accuracy (AUC = 0.85). Intra-reader reliability (R1/R2) ICC was 0.81/0.86. Inter-reader reliability ICC was 0.94. Iodine density from PCCT enterography can distinguish mild from moderate-to-severe active inflammation.

Quantitative analysis of MR T2 relaxation times in neck muscles

T2 relaxation times (T2 times) are different between resting and exercised muscles and between muscles of healthy subjects and subjects with muscle pathology. However, studies specifically focusing on neck muscles are lacking. Furthermore, normative neck muscle T2 times are not well defined and methodology used to analyse T2 times in neck muscles is not robust. We analysed T2 times in key neck muscles and explored factors affecting variability between muscles. 20 healthy subjects were recruited. Two circular regions of interest (ROIs) were drawn in two mutually exclusive regions within neck muscles on T2 weighted images and values averaged. ROI measurements were performed by a co-investigator, supervised by a neuro-radiologist. For the first ten subjects, measurements were done from C1-T1. For the remaining subjects, ROIs were drawn at two pre-determined levels. Two MRIs were repeated at 31 degrees acquisition to evaluate the effect of muscle fibre orientation. ROI values were translated into T2 times. Results showed semispinalis capitis had the longest T2 times (range 46.88–51.42 ms), followed by splenius capitis (range 47.37–48.33 ms), trapezius (range 45.27–47.46 ms), levator scapulae (range 43.17–45.63 ms) and sternocleidomastoid (range 38.45–42.91 ms). T2 times did not vary along length of muscles and were unaffected by muscle fibre orientation (P > 0.05). T2 times of splenius capitis correlated significantly with age at C2/C3 and C5/C6 levels and trapezius at C7/T1 level. Gender did not influence relaxation times (P > 0.05). In conclusion, results of normative neck muscle T2 time values and factors influencing the T2 times could serve as a reference for future MR analysis of neck muscles. The methodology used may also be useful for related studies of neck muscles.

MP09-20 CONTRAST AGENTS FOR UROLOGIC FLUOROSCOPY: A COMPARATIVE PHANTOM STUDY

You have accessJournal of UrologyCME1 Apr 2023MP09-20 CONTRAST AGENTS FOR UROLOGIC FLUOROSCOPY: A COMPARATIVE PHANTOM STUDY Anahita Heshmat, Gene Austin, Manuel Arreola, and Benjamin Canales Anahita HeshmatAnahita Heshmat More articles by this author , Gene AustinGene Austin More articles by this author , Manuel ArreolaManuel Arreola More articles by this author , and Benjamin CanalesBenjamin Canales More articles by this author View All Author Informationhttps://doi.org/10.1097/JU.0000000000003224.20AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: The recent international shortage of iodinated contrast agents has driven many clinicians to search for alternative agents to image a variety of urologic conditions in order to maintain clinic workflow. We sought to standardize these agents by a quantitative comparison using a phantom model. METHODS: A donut shape gel phantom with 10-mm wide plastic rods filled with Omnipaque-300, Omnipaque-350, Visipaque-270, Isovue-370, Gadavist, Eovist, and liquid water, was used to represent a patient being imaged for an urethrogram study utilizing a mobile fluoroscopy. The phantom was imaged with various acquisitions (Auto tube voltage (kVp), tube current (mA), and various dose modes, frame rates, and magnification modes) to compare and asses image quality for different contrast agents. Quantitative evaluations were performed with extracting regions of interest (ROIs) of each contrast rod to gather mean pixel values and standard deviations. These ROI measurements were then utilized to calculate contrast-to-noise ratio (CNR) as the metric appropriate for image quality evaluation. RESULTS: The mean pixel and CNR values of Omnipaque-300, Omnipaque-350, Visipaque-270, Isovue-370, and Gadavist showed no difference in visibility among these five contrast agents while using different dose modes, frame rates, and magnification modes. However, that was not the case for Eovist as its mean pixel value and CNR value were close to those of liquid water. Furthermore, changing magnification from normal to magnification 1 and magnification 2 increased the mean pixel values and CNR values at normal dose and continuous frame rate. CONCLUSIONS: With the exception of Eovist, the iodinated and gadolinium-based contrast agents we tested can be interchanged without compromising visibility or image quality based upon this comparative phantom study. Source of Funding: None © 2023 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 209Issue Supplement 4April 2023Page: e113 Advertisement Copyright & Permissions© 2023 by American Urological Association Education and Research, Inc.MetricsAuthor Information Anahita Heshmat More articles by this author Gene Austin More articles by this author Manuel Arreola More articles by this author Benjamin Canales More articles by this author Expand All Advertisement PDF downloadLoading ...

Is Groin Hernia Associated with Pubic Edema Severity in Footballers?

Objective: The groin area, with its complex anatomical structure, can have different athletic problems simultaneously. In this study, it was aimed to investigate whether groin hernia had a relationship with pubic bone marrow edema (BME) severity in footballers. &#x0D; &#x0D; Method: Among the athletes with painful groin from different sports, only male footballers who had pubic BME in pelvic MRI report were included in the study. The included athletes underwent a second archive scan for groin hernia (hernia symptoms, previous hernia surgery, sonography report). Then, the MRI sections of hernia-related cases were re-evaluated. For this purpose, right and left pubic BME signal intensities (SI) were quantitatively measured using "region of interest (ROI)" program module on T2 fat-suppressed images.&#x0D; &#x0D; Results: A total of 93 footballers with pubic BME were found. Eleven of them (11.8%) had a concurrent hernia, while five (5.4%) had a repaired hernia prior to study. ROI measurements of these sixteen athletes revealed that the pubic bones on the side of concurrent and repaired hernia had a significantly greater mean BME intensity than opposite pubic bones (1049 SI versus 796 SI, p

Age-dependent changes in brain iron deposition and volume in deep gray matter nuclei using quantitative susceptibility mapping

BackgroundMicrostructural changes in deep gray matter (DGM) nuclei are related to physiological behavior, cognition, and memory. Therefore, it is critical to study age-dependent trajectories of biomarkers in DGM nuclei for understanding brain development and aging, as well as predicting cognitive or neurodegenerative diseases. ObjectivesWe aimed to (1) characterize age-dependent trajectories of mean susceptibility, adjusted volume, and total iron content simultaneously in DGM nuclei using quantitative susceptibility mapping (QSM); (2) examine potential contributions of sex related effects to the different age-dependence trajectories of volume and iron deposition; and (3) evaluate the ability of brain age prediction by combining mean magnetic susceptibility and volume of DGM nuclei. MethodsMagnetic susceptibilities and volumetric values of DGM nuclei were obtained from 220 healthy participants (aged 10–70 years) scanned on a 3T MRI system. Regions of interest (ROIs) were drawn manually on the QSM images. Univariate regression analysis between age and each of the MRI measurements in a single ROI was performed. Pearson correlation coefficients were calculated between magnetic susceptibility and adjusted volume in a single ROI. The statistical significance of sex differences in age-dependent trajectories of magnetic susceptibilities and adjusted volumes were determined using one-way ANCOVA. Multiple regression analysis was used to evaluate the ability to estimate brain age using a combination of the mean susceptibilities and adjusted volumes in multiple DGM nuclei. ResultsMean susceptibility and total iron content increased linearly, quadratically, or exponentially with age in all six DGM nuclei. Negative linear correlation was observed between adjusted volume and age in the head of the caudate nucleus (CN; R2 = 0.196, p < 0.001). Quadratic relationships were found between adjusted volume and age in the putamen (PUT; R2 = 0.335, p < 0.001), globus pallidus (GP; R2 = 0.062, p = 0.001), and dentate nucleus (DN; R2 = 0.077, p < 0.001). Males had higher mean magnetic susceptibility than females in the PUT (p = 0.001), red nucleus (RN; p = 0.002), and substantia nigra (SN; p < 0.001). Adjusted volumes of the CN (p < 0.001), PUT (p = 0.030), GP (p = 0.007), SN (p = 0.021), and DN (p < 0.001) were higher in females than those in males throughout the entire age range (10–70 years old). The total iron content of females was higher than that of males in the CN (p < 0.001), but lower than that of males in the PUT (p = 0.014) and RN (p = 0.043) throughout the entire age range (10–70 years old). Multiple regression analyses revealed that the combination of the mean susceptibility value of the PUT, and the volumes of the CN and PUT had the strongest associations with brain age (R2 = 0.586). ConclusionsQSM can be used to simultaneously investigate age- and sex- dependent changes in magnetic susceptibility and volume of DGM nuclei, thus enabling a comprehensive understanding of the developmental trajectories of iron accumulation and volume in DGM nuclei during brain development and aging.

Apparent Diffusion Coefficient in the Differentiation of Common Pediatric Brain Tumors in the Posterior Fossa: Different Region-of-Interest Selection Methods for Time Efficiency, Measurement Reproducibility, and Diagnostic Utility.

This study aimed to explore the diagnostic ability of apparent diffusion coefficient (ADC) values obtained from different region of interest (ROI) measurements in tumor parenchyma for differentiating posterior fossa tumors (PFTs) and the correlations between ADC values and Ki-67. Seventy-three pediatric patients with PFTs who underwent conventional diffusion-weighted imaging were recruited in this study. Five different ROIs were manually drawn by 2 radiologists (ROI-polygon, ROI-3 sections, ROI-3-5 ovals, ROI-more ovals, and ROI-whole). The interreader/intrareader repeatability, time required, diagnostic ability, and Ki-67 correlation analysis of the ADC values based on these ROI strategies were calculated. Both interreader and intrareader reliabilities were excellent for ADC values among the different ROI strategies (intraclass correlation coefficient, 0.899-0.992). There were statistically significant differences in time consumption among the 5 ROI selection methods ( P < 0.001). The time required for the ROI-3-5 ovals was the shortest (32.23 ± 5.14 seconds), whereas the time required for the ROI-whole was the longest (204.52 ± 92.34 seconds). The diagnostic efficiency of the ADC values showed no significant differences among the different ROI measurements ( P > 0.05). The ADC value was negatively correlated with Ki-67 ( r = -0.745 to -0.798, all P < 0.0001). The ROI-3-5 ovals method has the best interobserver repeatability, the shortest amount of time spent, and the best diagnostic ability. Thus, it is considered an effective measurement to produce ADC values in the evaluation of pediatric PFTs.

Contrast thresholds for detection of various iodine concentrations in subtraction CT and dual-energy CT systems.

To estimate the minimum iodine concentrations detectable in simulated vessels of various diameters for both subtraction computed tomography (CT) and dual-energy CT systems. Fillable tubes (diameters: 1, 3, and 5mm) were filled with a variety of iodine concentrations (range: 0-20mg/ml), placed in the center of 28-mm cylindrical rods and surrounded with water. Rods with and without fillable tubes were placed in a 20-cm cylindrical solid-water phantom to simulate administration of iodine in blood vessels. The phantom was scanned with clinical subtraction CT (SCT) and dual-energy CT (DECT) head protocols to assess the detection of minimum iodine concentrations in both systems. The SCT and DECT images were evaluated quantitatively with a MATLAB script to extract regions of interest (ROIs) of each simulated vessel. ROI measurements were used to calculate the limit of detectability (LOD) and signal-to-noise ratio of Rose criteria for the assessment of the contrast thresholds. Both SNRRose and LOD methods agreed and determined the minimum detectable iodine concentration to be 0.4mg/ml in the 5-mm diameter vessel for SCT. However, the minimum detectable concentration in the 5-mm vessel with DECT was 1mg/ml. The 3-mm vessel had a minimum detectable concentration of 0.8mg/ml for SCT and 2mg/ml for DECT. Lastly, the minimum detectable iodine concentration for the 1-mm vessel was 10mg/ml for SCT and 10mg/ml for DECT. In this phantom study, SCT showed the capability to detect lower iodine concentrations compared to DECT. Contrast thresholds varied for vessels of different diameters and the smaller vessels required a higher iodine concentration for detection. Based on this knowledge, radiologists can modify their protocols to increase contrast enhancement.

Local ablation of hepatocellular carcinoma by interstitial brachytherapy: prediction of outcome by diffusion-weighted imaging.

Interstitial brachytherapy (iBT) has become a viable treatment option in the therapy of early and intermediate stage hepatocellular carcinoma (HCC). Prognostic imaging tools to predict patient outcome are missing. To assess the predictive value of baseline diffusion-weighted imaging in HCC before iBT with regard to local tumor control and overall survival (OS). We retrospectively identified 107 patients who underwent iBT for HCC from 2011 to 2018 from our database. Apparent diffusion coefficient (ADC) values for each treated lesion were analyzed in region of interest measurements. Additionally, explorative combined ratios adjusting total measured lesion area and mean measured lesion area per patient by ADC values were calculated. Measurements underwent a univariate and multivariate Cox regression analysis. The log rank test was then used to verify prognostic cutoff levels for median survival time. A total of 189 lesions in 81 patients were measured. Median survival of patients was 46.0 months. Neither ADC parameter was indicative of local tumor control. Lesion size >5 cm was associated with lower local tumor control (hazard ratio [HR]=4.292, 95% confidence interval [CI]=1.285-14.331; P = 0.018). Average measured lesion area divided by ADCmin (ADCarea mean, min) was identified to independently predict OS (HR=1.994, 95% CI=1.172-3.392; P = 0.011). A cutoff based on the variable's median (0.29 × 10-4 AU) identified patients with poor outcome (OS 36 vs. 61 months) for lower ADCarea mean, min values as verified by the log-rank test (P = 0.040). Pre-treatment ADCarea mean, min may serve as an independent predictor of OS in patients with HCC undergoing iBT.

Simplification of 2D shear wave elastography by enlarged SWE box and multiple regions of interest in one acquisition

Two-dimensional (2D) Shear Wave Elastography (SWE) is an easy to perform technique to evaluate the liver stiffness. To simplify the procedure and reduce the acquisition time we enlarged the size of the SWE-box and set ten regions of interest (ROI) in one acquisition. We compare the accuracy of this method to ten separate acquisitions in a small box each with a single ROI measurement. Sixty-nine volunteers with diffuse chronic liver disease were studied with 2D-SWE using a Canon Aplio i800 ultrasound system. The shear-wave-speed was measured in the right lobe in ten separate acquisitions and compared to one acquisition with increased size of the SWE-box and ten different ROI measurements. A Bland-Altmann plot was drawn and the interclass correlation coefficient (ICC) was calculated to compare both methods. Finally, 2D-SWE was successfully performed thru both methods in sixty-six participants. Between both methods the ICC is 0.82. The results of this study show a good reliability between ten separate measures and one grouped measure with ten ROI if the mean is below 1.6m/s (7.7kPa). For higher degrees of fibrosis (≥F2) further investigations are needed.

Gingival shape analysis using surface curvature estimation of the intraoral scans

BackgroundDespite many advances in dentistry, no objective and quantitative method is available to evaluate gingival shape. The surface curvature of the optical scans represents an unexploited possibility. The present study aimed to test surface curvature estimation of intraoral scans for objective evaluation of gingival shape.MethodsThe method consists of four main steps, i.e., optical scanning, surface curvature estimation, region of interest (ROI) definition, and gingival shape analysis. Six different curvature measures and three different diameters were tested for surface curvature estimation on central (n = 78) and interdental ROI (n = 88) of patients with advanced periodontitis to quantify gingiva with a novel gingival shape parameter (GS). The reproducibility was evaluated by repeating the method on two consecutive intraoral scans obtained with a scan-rescan process of the same patient at the same time point (n = 8).ResultsMinimum and mean curvature measures computed at 2 mm diameter seem optimal GS to quantify shape at central and interdental ROI, respectively. The mean (and standard deviation) of the GS was 0.33 ± 0.07 and 0.19 ± 0.09 for central ROI using minimum, and interdental ROI using mean curvature measure, respectively, computed at a diameter of 2 mm. The method’s reproducibility evaluated on scan-rescan models for the above-mentioned ROI and curvature measures was 0.02 and 0.01, respectively.ConclusionsSurface curvature estimation of the intraoral optical scans presents a precise and highly reproducible method for the objective gingival shape quantification enabling the detection of subtle changes. A careful selection of parameters for surface curvature estimation and curvature measures is required.

Contrast Media Reduction in Computed Tomography With Deep Learning Using a Generative Adversarial Network in an Experimental Animal Study.

This feasibility study aimed to use optimized virtual contrast enhancement through generative adversarial networks (GAN) to reduce the dose of iodine-based contrast medium (CM) during abdominal computed tomography (CT) in a large animal model. Multiphasic abdominal low-kilovolt CTs (90 kV) with low (low CM, 105 mgl/kg) and normal contrast media doses (normal CM, 350 mgl/kg) were performed with 20 healthy Göttingen minipigs on 3 separate occasions for a total of 120 examinations. These included an early arterial, late arterial, portal venous, and venous contrast phase. One animal had to be excluded because of incomplete examinations. Three of the 19 animals were randomly selected and withheld for validation (18 studies). Subsequently, the GAN was trained for image-to-image conversion from low CM to normal CM (virtual CM) with the remaining 16 animals (96 examinations). For validation, region of interest measurements were performed in the abdominal aorta, inferior vena cava, portal vein, liver parenchyma, and autochthonous back muscles, and the contrast-to-noise ratio (CNR) was calculated. In addition, the normal CM and virtual CM data were presented in a visual Turing test to 3 radiology consultants. On the one hand, they had to decide which images were derived from the normal CM examination. On the other hand, they had to evaluate whether both images are pathological consistent. Average vascular CNR (low CM 6.9 ± 7.0 vs virtual CM 28.7 ± 23.8, P < 0.0001) and parenchymal (low CM 1.5 ± 0.7 vs virtual CM 3.8 ± 2.0, P < 0.0001) CNR increased significantly by GAN-based contrast enhancement in all contrast phases and was not significantly different from normal CM examinations (vascular: virtual CM 28.7 ± 23.8 vs normal CM 34.2 ± 28.8; parenchymal: virtual CM 3.8 ± 2.0 vs normal CM 3.7 ± 2.6). During the visual Turing testing, the radiology consultants reported that images from normal CM and virtual CM were pathologically consistent in median in 96.5% of the examinations. Furthermore, it was possible for the examiners to identify the normal CM data as such in median in 91% of the cases. In this feasibility study, it could be demonstrated in an experimental setting with healthy Göttingen minipigs that the amount of CM for abdominal CT can be reduced by approximately 70% by GAN-based contrast enhancement with satisfactory image quality.

Repeatability of quantitative MRI in patients with rheumatoid arthritis

IntroductionRheumatoid arthritis (RA) affects 1% of the population and is principally associated with joint inflammation. It is suggested however that muscle involvement may be one of the earliest clinical features of RA. It is therefore important that techniques exist to accurately assess muscle health in those with RA to enable successful treatment. This study assesses the inter-rater and intra-rater repeatability of Diffusion Tensor MRI (DTI), 2-Point Dixon fat fraction, and T2 relaxation of the thigh muscle in patients with RA using manual regions of interest (ROI). MethodsNineteen patients (10/19 males; mean age 59; range 18–85) diagnosed with RA had an MRI scan of their hamstrings and quadriceps muscles to obtain fat fraction (FF), mean diffusivity (MD), fractional anisotropy (FA), and T2 quantitative measurements. Two raters DB & MF independently contoured ROIs for each patient. DB repeated the ROI for the same 19 patients after a 6-month hiatus to assess intra-rater repeatability. Inter-rater and intra-rater repeatability for the ROI measurements were compared using Inter Class Correlation (ICC) and Bland–Altman plots. ResultsThere was excellent agreement for both inter-rater and intra-rater repeatability. ICC results ranged from 0.900 to 0.998 (P < 0.001), and intra-rater ICC results ranged from 0.977 to 0.999 (P < 0.001). Bland–Altman plots also showed excellent agreement. ConclusionsICC measurements and Bland–Altman plots showed excellent repeatability and agreement with no statistically significant differences when assessing the inter-rater and intra-rater repeatability of FF, MD, FA, and T2 relaxation of the thigh muscle using manual regions of interest in patients with RA. Implications for practiceManual ROI drawing does not introduce significant errors obtaining FF, MD, FA, and T2 MRI measurements in an RA population.

Measurement of neurodegeneration using a multivariate early frame amyloid PET classifier.

IntroductionAmyloid measurement provides important confirmation of pathology for Alzheimer's disease (AD) clinical trials. However, many amyloid positive (Am+) early‐stage subjects do not worsen clinically during a clinical trial, and a neurodegenerative measure predictive of decline could provide critical information. Studies have shown correspondence between perfusion measured by early amyloid frames post‐tracer injection and fluorodeoxyglucose (FDG) positron emission tomography (PET), but with limitations in sensitivity. Multivariate machine learning approaches may offer a more sensitive means for detection of disease related changes as we have demonstrated with FDG.MethodsUsing summed dynamic florbetapir image frames acquired during the first 6 minutes post‐injection for 107 Alzheimer's Disease Neuroimaging Initiative subjects, we applied optimized machine learning to develop and test image classifiers aimed at measuring AD progression. Early frame amyloid (EFA) classification was compared to that of an independently developed FDG PET AD progression classifier by scoring the FDG scans of the same subjects at the same time point. Score distributions and correlation with clinical endpoints were compared to those obtained from FDG. Region of interest measures were compared between EFA and FDG to further understand discrimination performance.ResultsThe EFA classifier produced a primary pattern similar to that of the FDG classifier whose expression correlated highly with the FDG pattern (R‐squared 0.71), discriminated cognitively normal (NL) amyloid negative (Am–) subjects from all Am+ groups, and that correlated in Am+ subjects with Mini‐Mental State Examination, Clinical Dementia Rating Sum of Boxes, and Alzheimer's Disease Assessment Scale–13‐item Cognitive subscale (R = 0.59, 0.63, 0.73) and with subsequent 24‐month changes in these measures (R = 0.67, 0.73, 0.50).DiscussionOur results support the ability to use EFA with a multivariate machine learning–derived classifier to obtain a sensitive measure of AD‐related loss in neuronal function that correlates with FDG PET in preclinical and early prodromal stages as well as in late mild cognitive impairment and dementia.Highlights The summed initial post‐injection minutes of florbetapir positron emission tomography correlate with fluorodeoxyglucose.A machine learning classifier enabled sensitive detection of early prodromal Alzheimer's disease.Early frame amyloid (EFA) classifier scores correlate with subsequent change in Mini‐Mental State Examination, Clinical Dementia Rating Sum of Boxes, and Alzheimer's Disease Assessment Scale–13‐item Cognitive subscale.EFA classifier effect sizes and clinical prediction outperformed region of interest standardized uptake value ratio.EFA classification may aid in stratifying patients to assess treatment effect.

NigraNet: An automatic framework to assess nigral neuromelanin content in early Parkinson’s disease using convolutional neural network

BackgroundParkinson’s disease (PD) demonstrates neurodegenerative changes in the substantia nigra pars compacta (SNc) using neuromelanin-sensitive (NM)-MRI. As SNc manual segmentation is prone to substantial inter-individual variability across raters, development of a robust automatic segmentation framework is necessary to facilitate nigral neuromelanin quantification. Artificial intelligence (AI) is gaining traction in the neuroimaging community for automated brain region segmentation tasks using MRI. ObjectiveDeveloping and validating AI-based NigraNet, a fully automatic SNc segmentation framework allowing nigral neuromelanin quantification in patients with PD using NM-MRI. MethodsWe prospectively included 199 participants comprising 144 early-stage idiopathic PD patients (disease duration = 1.5 ± 1.0 years) and 55 healthy volunteers (HV) scanned using a 3 Tesla MRI including whole brain T1-weighted anatomical imaging and NM-MRI. The regions of interest (ROI) were delineated in all participants automatically using NigraNet, a modified U-net, and compared to manual segmentations performed by two experienced raters. The SNc volumes (Vol), volumes corrected by total intracranial volume (Cvol), normalized signal intensity (NSI) and contrast-to-noise ratio (CNR) were computed. One-way GLM-ANCOVA was performed while adjusting for age and sex as covariates. Diagnostic performance measurement was assessed using the receiver operating characteristic (ROC) analysis. Inter and intra-observer variability were estimated using Dice similarity coefficient (DSC). The agreements between methods were tested using intraclass correlation coefficient (ICC) based on a mean-rating, two-way, mixed-effects model estimates for absolute agreement. Cronbach’s alpha and Bland-Altman plots were estimated to assess inter-method consistency. ResultsUsing both methods, Vol, Cvol, NSI and CNR measurements differed between PD and HV with an effect of sex for Cvol and CNR. ICC values between the methods demonstrated optimal agreement for Cvol and CNR (ICC > 0.9) and high reproducibility (DSC: 0.80) was also obtained. The SNc measurements also showed good to excellent consistency values (Cronbach's alpha > 0.87). Bland-Altman plots of agreement demonstrated no association of SNc ROI measurement differences between the methods and ROI average measurements while confirming that 95 % of the data points were ranging between the limits of mean difference (d ± 1.96xSD). Percentage changes between PD and HV were −27.4 % and −17.7 % for Vol, −30.0 % and –22.2 % for Cvol, −15.8 % and −14.4 % for NSI, −17.1 % and −16.0 % for CNR for automatic and manual measurements respectively. Using automatic method, in the entire dataset, we obtained the areas under the ROC curve (AUC) of 0.83 for Vol, 0.85 for Cvol, 0.79 for NSI and 0.77 for CNR whereas in the training dataset of 0.96 for Vol, 0.95 for Cvol, 0.85 for NSI and 0.85 for CNR. Disease duration correlated negatively with NSI of the patients for both the automatic and manual measurements. ConclusionsWe presented an AI-based NigraNet framework that utilizes a small MRI training dataset to fully automatize the SNc segmentation procedure with an increased precision and more reproducible results. Considering the consistency, accuracy and speed of our approach, this study could be a crucial step towards the implementation of a time-saving non-rater dependent fully automatic method for studying neuromelanin changes in clinical settings and large-scale neuroimaging studies.