Metrology Standards for Quantitative Imaging Biomarkers

The Value of Quantitative Musculoskeletal Imaging

Musculoskeletal imaging is mainly based on the subjective and qualitative analysis of imaging examinations. However, integration of quantitative assessment of imaging data could increase the value of imaging in both research and clinical practice. Some imaging modalities, such as perfusion magnetic resonance imaging (MRI), diffusion MRI, or T2 mapping, are intrinsically quantitative. But conventional morphological imaging can also be analyzed through the quantification of various parameters. The quantitative data retrieved from imaging examinations can serve as biomarkers and be used to support diagnosis, determine patient prognosis, or monitor therapy.

We focus on the value, or clinical utility, of quantitative imaging in the musculoskeletal field. There is currently a trend to move from volume- to value-based payments. This review contains definitions and examines the role that quantitative imaging may play in the implementation of value-based health care. The influence of artificial intelligence on the value of quantitative musculoskeletal imaging is also discussed.

A multicenter study on radiomic features from T2 -weighted images of a customized MR pelvic phantom setting the basis for robust radiomic models in clinics

PURPOSE

To investigate the repeatability and reproducibility of radiomic features extracted from MR images and provide a workflow to identify robust features.

METHODS

T₂ -weighted images of a pelvic phantom were acquired on three scanners of two manufacturers and two magnetic field strengths. The repeatability and reproducibility of features were assessed by the intraclass correlation coefficient and the concordance correlation coefficient, respectively, and by the within-subject coefficient of variation, considering repeated acquisitions with and without phantom repositioning, and with different scanner and acquisition parameters. The features showing intraclass correlation coefficient or concordance correlation coefficient >0.9 were selected, and their dependence on shape information (Spearman's ρ > 0.8) analyzed. They were classified for their ability to distinguish textures, after shuffling voxel intensities of images.

RESULTS

From 944 two-dimensional features, 79.9% to 96.4% showed excellent repeatability in fixed position across all scanners. A much lower range (11.2% to 85.4%) was obtained after phantom repositioning. Three-dimensional extraction did not improve repeatability performance. Excellent reproducibility between scanners was observed in 4.6% to 15.6% of the features, at fixed imaging parameters. In addition, 82.4% to 94.9% of the features showed excellent agreement when extracted from images acquired with echo times 5 ms apart, but decreased with increasing echo-time intervals, and 90.7% of the features exhibited excellent reproducibility for changes in pulse repetition time. Of nonshape features, 2.0% was identified as providing only shape information.

CONCLUSION

We showed that radiomic features are affected by MRI protocols and propose a general workflow to identify repeatable, reproducible, and informative radiomic features to ensure robustness of clinical studies.

A systematic review of radiomics in osteosarcoma: utilizing radiomics quality score as a tool promoting clinical translation

OBJECTIVES

To assess the methodological quality and risk of bias in radiomics studies investigating diagnosis, therapy response, and survival of patients with osteosarcoma.

METHODS

In this systematic review, literatures on radiomics in osteosarcoma were included and assessed for methodological quality through the radiomics quality score (RQS). The risk of bias and concern of application was assessed using the Quality Assessment of Diagnostic Accuracy Studies tool. A meta-analysis of studies focusing on predicting osteosarcoma response to neoadjuvant chemotherapy was performed.

RESULTS

Twelve radiomics studies exploring osteosarcoma were identified, and five were included in meta-analysis. The RQS reached an average of 20.4% (6.92 of 36) with good inter-rater agreement (ICC 0.95, 95% CI 0.85-0.99). Four studies validated results with an internal dataset, none of which used external dataset; one study was prospectively designed, and another one shared part of the dataset. The risk of bias and concern of application were mainly related to index test aspect. The meta-analysis showed a diagnostic odds ratio of 43.68 (95%CI 13.5-141.31) for predicting response to neoadjuvant chemotherapy with high heterogeneity and low methodological quality.

CONCLUSIONS

The overall scientific quality of included studies is insufficient; however, radiomics remains a promising technology for predicting treatment response, which might guide therapeutic decision-making and related to prognosis. Improvements in study design, validation, and open science needs to be made to demonstrate the generalizability of findings and to achieve clinical applications. Widespread application of RQS, pre-trained RQS scoring procedure, and modification of RQS in response to clinical needs are necessary.

KEY POINTS

• Limited radiomics studies were established in osteosarcoma with mean RQS of 20.4%, commonly due to unvalidated results, retrospective study design, and absence of open science. • Meta-analysis of radiomics studies predicting osteosarcoma response to neoadjuvant chemotherapy showed high diagnostic odds ratio 43.68, while high heterogeneity and low methodological quality were the main concerns. • A previously trained data extraction instrument allowed reaching moderate inter-rater agreement in RQS applications, while RQS still needs improvement to become a wide adaptive tool in reviews of radiomics studies, in routine self-check before manuscript submitting and in study design.

Test-retest reliability of the human functional connectome over consecutive days: identifying highly reliable portions and assessing the impact of methodological choices

Countless studies have advanced our understanding of the human brain and its organization by using functional magnetic resonance imaging (fMRI) to derive network representations of human brain function. However, we do not know to what extent these "functional connectomes" are reliable over time. In a large public sample of healthy participants (N = 833) scanned on two consecutive days, we assessed the test-retest reliability of fMRI functional connectivity and the consequences on reliability of three common sources of variation in analysis workflows: atlas choice, global signal regression, and thresholding. By adopting the intraclass correlation coefficient as a metric, we demonstrate that only a small portion of the functional connectome is characterized by good (6-8%) to excellent (0.08-0.14%) reliability. Connectivity between prefrontal, parietal, and temporal areas is especially reliable, but also average connectivity within known networks has good reliability. In general, while unreliable edges are weak, reliable edges are not necessarily strong. Methodologically, reliability of edges varies between atlases, global signal regression decreases reliability for networks and most edges (but increases it for some), and thresholding based on connection strength reduces reliability. Focusing on the reliable portion of the connectome could help quantify brain trait-like features and investigate individual differences using functional neuroimaging.

Radiomics in medical imaging-"how-to" guide and critical reflection

Radiomics is a quantitative approach to medical imaging, which aims at enhancing the existing data available to clinicians by means of advanced mathematical analysis. Through mathematical extraction of the spatial distribution of signal intensities and pixel interrelationships, radiomics quantifies textural information by using analysis methods from the field of artificial intelligence. Various studies from different fields in imaging have been published so far, highlighting the potential of radiomics to enhance clinical decision-making. However, the field faces several important challenges, which are mainly caused by the various technical factors influencing the extracted radiomic features.The aim of the present review is twofold: first, we present the typical workflow of a radiomics analysis and deliver a practical "how-to" guide for a typical radiomics analysis. Second, we discuss the current limitations of radiomics, suggest potential improvements, and summarize relevant literature on the subject.

Clinical evaluation of automated quantitative MRI reports for assessment of hippocampal sclerosis

OBJECTIVES

Hippocampal sclerosis (HS) is a common cause of temporal lobe epilepsy. Neuroradiological practice relies on visual assessment, but quantification of HS imaging biomarkers-hippocampal volume loss and T2 elevation-could improve detection. We tested whether quantitative measures, contextualised with normative data, improve rater accuracy and confidence.

METHODS

Quantitative reports (QReports) were generated for 43 individuals with epilepsy (mean age ± SD 40.0 ± 14.8 years, 22 men; 15 histologically unilateral HS; 5 bilateral; 23 MR-negative). Normative data was generated from 111 healthy individuals (age 40.0 ± 12.8 years, 52 men). Nine raters with different experience (neuroradiologists, trainees, and image analysts) assessed subjects' imaging with and without QReports. Raters assigned imaging normal, right, left, or bilateral HS. Confidence was rated on a 5-point scale.

RESULTS

Correct designation (normal/abnormal) was high and showed further trend-level improvement with QReports, from 87.5 to 92.5% (p = 0.07, effect size d = 0.69). Largest magnitude improvement (84.5 to 93.8%) was for image analysts (d = 0.87). For bilateral HS, QReports significantly improved overall accuracy, from 74.4 to 91.1% (p = 0.042, d = 0.7). Agreement with the correct diagnosis (kappa) tended to increase from 0.74 ('fair') to 0.86 ('excellent') with the report (p = 0.06, d = 0.81). Confidence increased when correctly assessing scans with the QReport (p < 0.001, η²_p = 0.945).

CONCLUSIONS

QReports of HS imaging biomarkers can improve rater accuracy and confidence, particularly in challenging bilateral cases. Improvements were seen across all raters, with large effect sizes, greatest for image analysts. These findings may have positive implications for clinical radiology services and justify further validation in larger groups.

KEY POINTS

• Quantification of imaging biomarkers for hippocampal sclerosis-volume loss and raised T2 signal-could improve clinical radiological detection in challenging cases. • Quantitative reports for individual patients, contextualised with normative reference data, improved diagnostic accuracy and confidence in a group of nine raters, in particular for bilateral HS cases. • We present a pre-use clinical validation of an automated imaging assessment tool to assist clinical radiology reporting of hippocampal sclerosis, which improves detection accuracy.

CT Phantom Evaluation of 67,392 American College of Radiology Accreditation Examinations: Implications for Opportunistic Screening of Osteoporosis Using CT

OBJECTIVE. The purpose of this study was to investigate whether systematic bias in attenuation measurements occurs among CT scanners made by four major manufacturers and the relevance of this bias regarding opportunistic screening for osteoporosis. MATERIALS AND METHODS. Data on attenuation measurement accuracy were acquired using the American College of Radiology (ACR) accreditation phantom and were evaluated in a blinded fashion for four CT manufacturers (8500 accreditation submissions for manufacturer A; 18,575 for manufacturer B; 8278 for manufacturer C; and 32,039 for manufacturer D). The attenuation value for water, acrylic (surrogate for trabecular bone), and Teflon (surrogate for cortical bone; Chemours) materials for an adult abdominal CT technique (120 kV, 240 mA, standard reconstruction algorithm) was used in the analysis. Differences in attenuation value across all manufacturers were assessed using the Kruskal-Wallis test followed by a post hoc test for pairwise comparisons. RESULTS. The mean attenuation value for water ranged from -0.3 to 2.7 HU, with highly significant differences among all manufacturers (p < 0.001). For the trabecular bone surrogate, differences in attenuation values across all manufacturers were also highly significant (p < 0.001), with mean values of 120.9 (SD, 3.5), 124.6 (3.3), 126.9 (4.4), and 123.9 (3.4) HU for manufacturers A, B, C, and D, respectively. For the cortical bone surrogate, differences in attenuation values across all manufacturers were also highly significant (p < 0.001), with mean values of 939.0 (14.2), 874.3 (13.3), 897.6 (11.3), and 912.7 (13.4) HU for manufacturers A, B, C, and D, respectively. CONCLUSION. CT scanners made by different manufacturers show systematic offsets in attenuation measurement when compared with each other. Knowledge of these off-sets is useful for optimizing the accuracy of opportunistic diagnosis of osteoporosis.

AI in Medical Imaging Informatics: Current Challenges and Future Directions

This paper reviews state-of-the-art research solutions across the spectrum of medical imaging informatics, discusses clinical translation, and provides future directions for advancing clinical practice. More specifically, it summarizes advances in medical imaging acquisition technologies for different modalities, highlighting the necessity for efficient medical data management strategies in the context of AI in big healthcare data analytics. It then provides a synopsis of contemporary and emerging algorithmic methods for disease classification and organ/ tissue segmentation, focusing on AI and deep learning architectures that have already become the de facto approach. The clinical benefits of in-silico modelling advances linked with evolving 3D reconstruction and visualization applications are further documented. Concluding, integrative analytics approaches driven by associate research branches highlighted in this study promise to revolutionize imaging informatics as known today across the healthcare continuum for both radiology and digital pathology applications. The latter, is projected to enable informed, more accurate diagnosis, timely prognosis, and effective treatment planning, underpinning precision medicine.

3-D-Printed Registration Phantom for Combined Ultrasound and Optical Imaging of Biological Tissues

Efforts to develop quantitative ultrasound biomarkers would benefit from comparisons between ultrasound data and higher-resolution images of the tissue microstructure, such as from optical microscopy. However, only a few studies have used these methods for multiscale imaging because it is difficult to register low-resolution (>100 μm) ultrasound images to high-resolution microscopy images. To address this need, we have designed a 3-D-printed registration phantom that is made of a hard fluorescent resin, fits into a glass-bottom dish and can be used to calculate a coordinate system transform between ultrasound and optical microscopy. We report the phantom design, a registration protocol and an example registration using 18.5-MHz ultrasound and second harmonic generation microscopy. We evaluate the registration precision, achieving standard deviations smaller than the ultrasound resolution across all axes, and illustrate on a mouse mammary gland that this method yields results superior to those of manual landmark registration.

Imaging Biomarkers and Pathobiological Profiling in a Rat Model of Drug-Induced Interstitial Lung Disease Induced by Bleomycin

A large number of systemically administered drugs have the potential to cause drug-induced interstitial lung disease (DIILD). We aim to characterize a model of DIILD in the rat and develop imaging biomarkers (IBs) for detection and quantification of DIILD. In this study, Sprague-Dawley rats received one single dose of intratracheal (i.t.) bleomycin and were longitudinally imaged at day 0, 3, 7, 14, 21, and 28 post dosing, applying the imaging techniques magnetic resonance imaging (MRI) and positron emission tomography (PET)/computed tomography (CT). Bronchoalveolar lavage fluid (BALF) was analyzed for total protein and inflammatory cells. Lungs were saved for further evaluation by gene analysis using quantitative-PCR and by histology. Lung sections were stained with Masson's-Trichrome staining and evaluated by modified Ashcroft score. Gene expression profiling of inflammatory and fibrotic markers was performed on lung tissue homogenates. Bleomycin induced significant increase in total protein concentration and total cell count in bronchoalveolar lavage (BAL), peaking at day 3 (p > 0.001) and day 7 (p > 0.001) compared to control, respectively. Lesions measured by MRI and PET signal in the lungs of bleomycin challenged rats were significantly increased during days 3-14, peaking at day 7. Two subgroups of animals were identified as low- and high-responders by their different change in total lung volume. Both groups showed signs of inflammation initially, while at later time points, the low-responder group recovered toward control, and the high-responder group showed sustained lung volume increase, and significant increase of lesion volume (p < 0.001) compared to control. Lastly, important inflammatory and pro-fibrotic markers were assessed from lung tissue, linking observed imaging pathological changes to gene expression patterns. In conclusion, bleomycin-induced lung injury is an adequate animal model for DIILD studies and for translational lung injury assessment by MRI and PET imaging. The scenario comprised disease responses, with different fractions of inflammation and fibrosis. Thereby, this study improved the understanding of imaging and biological biomarkers in DIILD and lung injury.

European Society of Radiology (ESR), Alberich-Bayarri A, Sourbron S, Golay X, deSouza N, Smits M, van der Lugt A, Boellard R. ESR Statement on the Validation of Imaging Biomarkers

Medical imaging capable of generating imaging biomarkers, specifically radiology and nuclear medicine image acquisition and analysis processes, differs from frequently used comparators like blood or urine biomarkers. This difference arises from the sample acquisition methodology. While different analysis methodologies and equipment provide slightly different results in any analytical domain, unlike blood or urine analysis where the samples are obtained by simple extraction or excretion, in radiology the acquisition of the sample is heterogeneous by design, since complex equipment from different vendors is used. Therefore, with this additional degree of freedom in medical imaging, there is still risk of persistent heterogeneity of image quality through time, due to different technological implementations across vendors and protocols used in different centres. Quantitative imaging biomarkers have yet to demonstrate an impact on clinical practice due to this lack of comprehensive standardisation in terms of technical aspects of image acquisition, analysis algorithms, processes and clinical validation.The aim is establishing a standard methodology based on metrology for the validation of image acquisition and analysis methods used in the extraction of biomarkers and radiomics data. The appropriate implementation of the guidelines herein proposed by radiology departments, research institutes and industry will allow for a significant reduction in inter-vendor & inter-centre variability in imaging biomarkers and determine the measurement error obtained, enabling them to be used in imaging-based criteria for diagnosis, prognosis or treatment response, ultimately improving clinical workflows and patient care. The validation of developed analytical methods must be based on a technical performance validation and clinical validation.

Optimizing data processing to improve the reproducibility of single-subject functional magnetic resonance imaging

INTRODUCTION

High reproducibility is critical for ensuring the confidence needed to use functional magnetic resonance imaging (fMRI) activation maps for presurgical planning.

METHODS

In this study, the comparison of different motion correction methods, spatial smoothing methods, regression methods, and thresholding methods was performed to see whether specific data processing methods can be employed to improve the reproducibility of single-subject fMRI activation. Three test-retest metrics were used: the percent difference in activation volume (PDAV), the difference in the center of mass (DCM), and the Dice Similarity Coefficient (DSC).

RESULTS

The PDAV was minimized when using little or no spatial smoothing and AMPLE thresholding. The DCM was minimized when using affine motion correction and little or no spatial smoothing. The DSC was improved when using affine motion correction and generous spatial smoothing. However, it is believed that the overlap metric may be unsuitable for testing fMRI reproducibility.

CONCLUSION

Processing methods to improve fMRI reproducibility were determined. Importantly, the processing methods needed to improve reproducibility were dependent on the fMRI activation metric of interest.

Radiomics of computed tomography and magnetic resonance imaging in renal cell carcinoma-a systematic review and meta-analysis

OBJECTIVES

(1) To assess the methodological quality of radiomics studies investigating histological subtypes, therapy response, and survival in patients with renal cell carcinoma (RCC) and (2) to determine the risk of bias in these radiomics studies.

METHODS

In this systematic review, literature published since 2000 on radiomics in RCC was included and assessed for methodological quality using the Radiomics Quality Score. The risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies tool and a meta-analysis of radiomics studies focusing on differentiating between angiomyolipoma without visible fat and RCC was performed.

RESULTS

Fifty-seven studies investigating the use of radiomics in renal cancer were identified, including 4590 patients in total. The average Radiomics Quality Score was 3.41 (9.4% of total) with good inter-rater agreement (ICC 0.96, 95% CI 0.93-0.98). Three studies validated results with an independent dataset, one used a publically available validation dataset. None of the studies shared the code, images, or regions of interest. The meta-analysis showed moderate heterogeneity among the included studies and an odds ratio of 6.24 (95% CI 4.27-9.12; p < 0.001) for the differentiation of angiomyolipoma without visible fat from RCC.

CONCLUSIONS

Radiomics algorithms show promise for answering clinical questions where subjective interpretation is challenging or not established. However, the generalizability of findings to prospective cohorts needs to be demonstrated in future trials for progression towards clinical translation. Improved sharing of methods including code and images could facilitate independent validation of radiomics signatures.

KEY POINTS

• Studies achieved an average Radiomics Quality Score of 10.8%. Common reasons for low Radiomics Quality Scores were unvalidated results, retrospective study design, absence of open science, and insufficient control for multiple comparisons. • A previous training phase allowed reaching almost perfect inter-rater agreement in the application of the Radiomics Quality Score. • Meta-analysis of radiomics studies distinguishing angiomyolipoma without visible fat from renal cell carcinoma show moderate diagnostic odds ratios of 6.24 and moderate methodological diversity.

Generative Adversarial Networks Improve the Reproducibility and Discriminative Power of Radiomic Features

Purpose

To assess the contribution of a generative adversarial network (GAN) to improve intermanufacturer reproducibility of radiomic features (RFs).

Materials and Methods

The authors retrospectively developed a cycle-GAN to translate texture information from chest radiographs acquired using one manufacturer (Siemens) to chest radiographs acquired using another (Philips), producing fake chest radiographs with different textures. The authors prospectively evaluated the ability of this texture-translation cycle-GAN to reduce the intermanufacturer variability of RFs extracted from the lung parenchyma. This study assessed the cycle-GAN's ability to fool several machine learning (ML) classifiers tasked with recognizing the manufacturer on the basis of chest radiography inputs. The authors also evaluated the cycle-GAN's ability to mislead radiologists who were asked to perform the same recognition task. Finally, the authors tested whether the cycle-GAN had an impact on radiomic diagnostic accuracy for chest radiography in patients with congestive heart failure (CHF).

Results

RFs, extracted from chest radiographs after the cycle-GAN's texture translation (fake chest radiographs), showed decreased intermanufacturer RF variability. Using cycle-GAN-generated chest radiographs as inputs, ML classifiers categorized the fake chest radiographs as belonging to the target manufacturer rather than to a native one. Moreover, cycle-GAN fooled two experienced radiologists who identified fake chest radiographs as belonging to a target manufacturer class. Finally, reducing intermanufacturer RF variability with cycle-GAN improved the discriminative power of RFs for patients without CHF versus patients with CHF (from 55% to 73.5%, P < .001).

Conclusion

Both ML classifiers and radiologists had difficulty recognizing the chest radiographs' manufacturer. The cycle-GAN improved RF intermanufacturer reproducibility and discriminative power for identifying patients with CHF. This deep learning approach may help counteract the sensitivity of RFs to differences in acquisition.Supplemental material is available for this article.© RSNA, 2020See also the commentary by Alderson in this issue.

T1 mapping performance and measurement repeatability: results from the multi-national T1 mapping standardization phantom program (T1MES)

BACKGROUND

The T1 Mapping and Extracellular volume (ECV) Standardization (T1MES) program explored T1 mapping quality assurance using a purpose-developed phantom with Food and Drug Administration (FDA) and Conformité Européenne (CE) regulatory clearance. We report T1 measurement repeatability across centers describing sequence, magnet, and vendor performance.

METHODS

Phantoms batch-manufactured in August 2015 underwent 2 years of structural imaging, B0 and B1, and "reference" slow T1 testing. Temperature dependency was evaluated by the United States National Institute of Standards and Technology and by the German Physikalisch-Technische Bundesanstalt. Center-specific T1 mapping repeatability (maximum one scan per week to minimum one per quarter year) was assessed over mean 358 (maximum 1161) days on 34 1.5 T and 22 3 T magnets using multiple T1 mapping sequences. Image and temperature data were analyzed semi-automatically. Repeatability of serial T1 was evaluated in terms of coefficient of variation (CoV), and linear mixed models were constructed to study the interplay of some of the known sources of T1 variation.

RESULTS

Over 2 years, phantom gel integrity remained intact (no rips/tears), B0 and B1 homogenous, and "reference" T1 stable compared to baseline (% change at 1.5 T, 1.95 ± 1.39%; 3 T, 2.22 ± 1.44%). Per degrees Celsius, 1.5 T, T1 (MOLLI 5s(3s)3s) increased by 11.4 ms in long native blood tubes and decreased by 1.2 ms in short post-contrast myocardium tubes. Agreement of estimated T1 times with "reference" T1 was similar across Siemens and Philips CMR systems at both field strengths (adjusted R2 ranges for both field strengths, 0.99-1.00). Over 1 year, many 1.5 T and 3 T sequences/magnets were repeatable with mean CoVs < 1 and 2% respectively. Repeatability was narrower for 1.5 T over 3 T. Within T1MES repeatability for native T1 was narrow for several sequences, for example, at 1.5 T, Siemens MOLLI 5s(3s)3s prototype number 448B (mean CoV = 0.27%) and Philips modified Look-Locker inversion recovery (MOLLI) 3s(3s)5s (CoV 0.54%), and at 3 T, Philips MOLLI 3b(3s)5b (CoV 0.33%) and Siemens shortened MOLLI (ShMOLLI) prototype 780C (CoV 0.69%). After adjusting for temperature and field strength, it was found that the T1 mapping sequence and scanner software version (both P < 0.001 at 1.5 T and 3 T), and to a lesser extent the scanner model (P = 0.011, 1.5 T only), had the greatest influence on T1 across multiple centers.

CONCLUSION

The T1MES CE/FDA approved phantom is a robust quality assurance device. In a multi-center setting, T1 mapping had performance differences between field strengths, sequences, scanner software versions, and manufacturers. However, several specific combinations of field strength, sequence, and scanner are highly repeatable, and thus, have potential to provide standardized assessment of T1 times for clinical use, although temperature correction is required for native T1 tubes at least.

Juvenile Idiopathic Arthritis: Diffusion-weighted MRI in the Assessment of Arthritis in the Knee

Background Diffusion-weighted imaging (DWI) can depict the inflamed synovial membrane in arthritis. Purpose To study the diagnostic accuracy of DWI for the detection of arthritis compared with the clinical reference standard and to compare DWI to contrast material-enhanced MRI for the detection of synovial inflammation. Materials and Methods In this institutional review board-approved prospective study, 45 participants with juvenile idiopathic arthritis (JIA) or suspected of having JIA (seven boys, 38 girls; median age, 14 years [interquartile range, 12-16 years]) were included between December 2015 and December 2018. Study participants underwent pre- and postcontrast 3.0-T MRI of the knee with an additional DWI sequence. For the clinical reference standard, a multidisciplinary team determined the presence or absence of arthritis on the basis of clinical, laboratory, and imaging findings (excluding DWI). Two data sets were scored by two radiologists blinded to all clinical data; data set 1 contained pre- and postcontrast sequences (contrast-enhanced MRI), and data set 2 contained precontrast and DWI sequences (DWI). Diagnostic accuracy was determined by comparing the scores of the DWI data set to those of the clinical reference standard. Second, DWI was compared with contrast-enhanced MRI regarding detection of synovial inflammation. Results Sensitivity for detection of arthritis for DWI was 93% (13 of the 14 participants with arthritis were correctly classified with DWI; 95% confidence interval [CI]: 64%, 100%) and specificity was 81% (25 of 31 participants without arthritis were correctly classified with DWI; 95% CI: 62%, 92%). Scores for synovial inflammation at DWI and contrast-enhanced MRI agreed in 37 of 45 participants (82%), resulting in a sensitivity of 92% (12 of 13 participants; 95% CI: 62%, 100%) and specificity of 78% (25 of 32 participants; 95% CI: 60%, 90%) with DWI when contrast-enhanced MRI was considered the reference standard. Conclusion Diffusion-weighted imaging (DWI) was accurate in detecting arthritis in pediatric participants with juvenile idiopathic arthritis (JIA) or suspected of having JIA and showed agreement with contrast-enhanced MRI. The results indicate that DWI could replace contrast-enhanced MRI for imaging of synovial inflammation in this patient group. © RSNA, 2020 Online supplemental material is available for this article.

Acceleration of 2D-MR fingerprinting by reducing the number of echoes with increased in-plane resolution: a volunteer study

Objective

To compare the absolute values and repeatability of magnetic resonance fingerprinting (MRF) with 3000 and 1500 echoes/slice acquired in 41 s and 20 s (MRF3k and MRF1.5k, respectively).

Materials and methods

MRF3k and MRF1.5k scans based on fast imaging with steady precession (FISP) were conducted using a 3 T scanner. Inter-scan agreement and intra-scan repeatability were investigated in 41 and 28 subjects, respectively. Region-of-interest (ROI) analysis was conducted on T1 values of MRF3k by two raters, and their agreement was evaluated using intraclass correlation coefficients (ICCs). Between MRF3k and MRF1.5k, differences in T1 and T2 values and inter-measurement correlation coefficients (CCs) were investigated. Intra-measurement repeatability was evaluated using coefficients of variation (CVs). A p value < 0.05 was considered statistically significant.

Results

The ICCs of ROI measurements were 0.77–0.96. Differences were observed between the two MRF scans, but the CCs of the overall ROIs were 0.99 and 0.97 for the T1 and T2 values, respectively. The mean and median CVs of repeatability were equal to or less than 1.58% and 3.13% in each of the ROIs for T1 and T2, respectively; there were some significant differences between MRF3k and MRF1.5k, but they were small, measuring less than 1%.

Discussion

Both MRF3k and MRF1.5k had high repeatability, and a strong to very strong correlation was observed, with a trend toward slightly higher values in MRF1.5k.

Electronic supplementary material

The online version of this article (10.1007/s10334-020-00842-8) contains supplementary material, which is available to authorized users.

Data transformations for statistical assessment of quantitative imaging biomarkers: Application to lung nodule volumetry

Variance stabilization is an important step in the statistical assessment of quantitative imaging biomarkers. The objective of this study is to compare the Log and the Box-Cox transformations for variance stabilization in the context of assessing the performance of a particular quantitative imaging biomarker, the estimation of lung nodule volume from computed tomography images. First, a model is developed to generate and characterize repeated measurements typically observed in computed tomography lung nodule volume estimation. Given this model, we derive the parameter of the Box-Cox transformation that stabilizes the variance of the measurements across lung nodule volumes. Second, simulated, phantom, and clinical datasets are used to compare the Log and the Box-Cox transformations. Two metrics are used for quantifying the stability of the measurements across the transformed lung nodule volumes: the coefficient of variation for the standard deviation and the repeatability coefficient. The results for simulated, phantom, and clinical datasets show that the Box-Cox transformation generally had better variance stabilization performance compared to the Log transformation for lung nodule volume estimates from computed tomography scans.

The QIBA Profile for FDG PET/CT as an Imaging Biomarker Measuring Response to Cancer Therapy

The Quantitative Imaging Biomarkers Alliance (QIBA) Profile for fluorodeoxyglucose (FDG) PET/CT imaging was created by QIBA to both characterize and reduce the variability of standardized uptake values (SUVs). The Profile provides two complementary claims on the precision of SUV measurements. First, tumor glycolytic activity as reflected by the maximum SUV (SUVmax) is measurable from FDG PET/CT with a within-subject coefficient of variation of 10%-12%. Second, a measured increase in SUVmax of 39% or more, or a decrease of 28% or more, indicates that a true change has occurred with 95% confidence. Two applicable use cases are clinical trials and following individual patients in clinical practice. Other components of the Profile address the protocols and conformance standards considered necessary to achieve the performance claim. The Profile is intended for use by a broad audience; applications can range from discovery science through clinical trials to clinical practice. The goal of this report is to provide a rationale and overview of the FDG PET/CT Profile claims as well as its context, and to outline future needs and potential developments.

Advanced Cancer Imaging Applied in the Comparative Setting

The potential for companion (pet) species with spontaneously arising tumors to act as surrogates for preclinical development of advanced cancer imaging technologies has become more apparent in the last decade. The utility of the companion model specifically centers around issues related to body size (including spatial target/normal anatomic characteristics), physical size and spatial distribution of metastasis, tumor heterogeneity, the presence of an intact syngeneic immune system and a syngeneic tumor microenvironment shaped by the natural evolution of the cancer. Companion species size allows the use of similar equipment, hardware setup, software, and scan protocols which provide the opportunity for standardization and harmonization of imaging operating procedures and quality assurance across imaging protocols, imaging hardware, and the imaged species. Murine models generally do not replicate the size and spatial distribution of human metastatic cancer and these factors strongly influence image resolution and dosimetry. The following review will discuss several aspects of comparative cancer imaging in more detail while providing several illustrative examples of investigational approaches performed or currently under exploration at our institutions. Topics addressed include a discussion on interested consortia; image quality assurance and harmonization; image-based biomarker development and validation; contrast agent and radionuclide tracer development; advanced imaging to assess and predict response to cytotoxic and immunomodulatory anticancer agents; imaging of the tumor microenvironment; development of novel theranostic approaches; cell trafficking assessment via non-invasive imaging; and intraoperative imaging to inform surgical oncology decision making. Taken in totality, these comparative opportunities predict that safety, diagnostic and efficacy data generated in companion species with naturally developing and progressing cancers would better recapitulate the human cancer condition than that of artificial models in small rodent systems and ultimately accelerate the integration of novel imaging technologies into clinical practice. It is our hope that the examples presented should serve to provide those involved in cancer investigations who are unfamiliar with available comparative methodologies an understanding of the potential utility of this approach.

[A primer on radiomics]

CLINICAL ISSUE

The reproducible and exhaustive extraction of information from radiological images is a central task in the practice of radiology. Dynamic developments in the fields of artificial intelligence (AI) and machine learning are introducing new methods for this task. Radiomics is one such method and offers new opportunities and challenges for the future of radiology.

METHODOLOGICAL INNOVATIONS

Radiomics describes the quantitative evaluation, interpretation, and clinical assessment of imaging markers in radiological data. Components of a radiomics analysis are data acquisition, data preprocessing, data management, segmentation of regions of interest, computation and selection of imaging markers, as well as the development of a radiomics model used for diagnosis and prognosis. This article explains these components and aims at providing an introduction to the field of radiomics while highlighting existing limitations.

MATERIALS AND METHODS

This article is based on a selective literature search with the PubMed search engine.

ASSESSMENT

Even though radiomics applications have yet to arrive in routine clinical practice, the quantification of radiological data in terms of radiomics is underway and will increase in the future. This holds the potential for lasting change in the discipline of radiology. Through the successful extraction and interpretation of all the information encoded in radiological images the next step in the direction of a more personalized, future-oriented form of medicine can be taken.

Magician's Corner: 4. Image Segmentation with U-Net

A popular deep learning framework (Keras) is applied to the problem of image segmentation using a U-Net.

Test-retest reproducibility of a deep learning-based automatic detection algorithm for the chest radiograph

OBJECTIVES

To perform test-retest reproducibility analyses for deep learning-based automatic detection algorithm (DLAD) using two stationary chest radiographs (CRs) with short-term intervals, to analyze influential factors on test-retest variations, and to investigate the robustness of DLAD to simulated post-processing and positional changes.

METHODS

This retrospective study included patients with pulmonary nodules resected in 2017. Preoperative CRs without interval changes were used. Test-retest reproducibility was analyzed in terms of median differences of abnormality scores, intraclass correlation coefficients (ICC), and 95% limits of agreement (LoA). Factors associated with test-retest variation were investigated using univariable and multivariable analyses. Shifts in classification between the two CRs were analyzed using pre-determined cutoffs. Radiograph post-processing (blurring and sharpening) and positional changes (translations in x- and y-axes, rotation, and shearing) were simulated and agreement of abnormality scores between the original and simulated CRs was investigated.

RESULTS

Our study analyzed 169 patients (median age, 65 years; 91 men). The median difference of abnormality scores was 1-2% and ICC ranged from 0.83 to 0.90. The 95% LoA was approximately ± 30%. Test-retest variation was negatively associated with solid portion size (β, - 0.50; p = 0.008) and good nodule conspicuity (β, - 0.94; p < 0.001). A small fraction (15/169) showed discordant classifications when the high-specificity cutoff (46%) was applied to the model outputs (p = 0.04). DLAD was robust to the simulated positional change (ICC, 0.984, 0.996), but relatively less robust to post-processing (ICC, 0.872, 0.968).

CONCLUSIONS

DLAD was robust to the test-retest variation. However, inconspicuous nodules may cause fluctuations of the model output and subsequent misclassifications.

KEY POINTS

• The deep learning-based automatic detection algorithm was robust to the test-retest variation of the chest radiographs in general. • The test-retest variation was negatively associated with solid portion size and good nodule conspicuity. • High-specificity cutoff (46%) resulted in discordant classifications of 8.9% (15/169; p = 0.04) between the test-retest radiographs.

Machine learning and glioma imaging biomarkers

AIM

To review how machine learning (ML) is applied to imaging biomarkers in neuro-oncology, in particular for diagnosis, prognosis, and treatment response monitoring.

MATERIALS AND METHODS

The PubMed and MEDLINE databases were searched for articles published before September 2018 using relevant search terms. The search strategy focused on articles applying ML to high-grade glioma biomarkers for treatment response monitoring, prognosis, and prediction.

RESULTS

Magnetic resonance imaging (MRI) is typically used throughout the patient pathway because routine structural imaging provides detailed anatomical and pathological information and advanced techniques provide additional physiological detail. Using carefully chosen image features, ML is frequently used to allow accurate classification in a variety of scenarios. Rather than being chosen by human selection, ML also enables image features to be identified by an algorithm. Much research is applied to determining molecular profiles, histological tumour grade, and prognosis using MRI images acquired at the time that patients first present with a brain tumour. Differentiating a treatment response from a post-treatment-related effect using imaging is clinically important and also an area of active study (described here in one of two Special Issue publications dedicated to the application of ML in glioma imaging).

CONCLUSION

Although pioneering, most of the evidence is of a low level, having been obtained retrospectively and in single centres. Studies applying ML to build neuro-oncology monitoring biomarker models have yet to show an overall advantage over those using traditional statistical methods. Development and validation of ML models applied to neuro-oncology require large, well-annotated datasets, and therefore multidisciplinary and multi-centre collaborations are necessary.

Reproducibility of hepatic MR elastography across field strengths, pulse sequences, scan intervals, and readers

PURPOSE

To evaluate the reproducibility of hepatic MRE under various combinations of settings of field strength, pulse sequence, scan interval, and reader in non-alcoholic fatty liver disease (NAFLD) patients.

METHODS

Adult NAFLD patients were prospectively enrolled for serial hepatic MRE with 1.5 T using 2D GRE sequence and 3.0 T using 2D SE-EPI sequence on the same day and after 2 weeks, resulting a total of four MRE examinations per patient. Three readers with various levels of background knowledge in MRE technique and liver anatomy measured liver stiffness after a training session. Linear regression, Bland-Altman analysis, within-subject coefficient of variation, and reproducibility coefficient (RDC) were used to determine reproducibility of hepatic MRE measurement.

RESULTS

Twenty patients completed the MRE sessions. Liver stiffness through MRE showed pooled RDC of 26% (upper 95% CI 30.6%) and corresponding limits of agreement (LOA) within 0.55 kPa across field strengths, MRE sequences, and 2-week interscan interval in three readers. Small mean biases and narrow LOA were observed among readers (0.05-0.19 kPa ± 0.53).

CONCLUSION

The magnitude of change across combinations of scan parameters is within acceptable clinical range, rendering liver stiffness through MRE a reproducible quantitative imaging biomarker. A lower reproducibility was observed for measurements under different field strengths/MRE sequences at a longer (2 weeks) interscan interval. Operators should be trained to acquire region of interest consistently in repeat examinations.

Evaluation of Quantitative Imaging Biomarkers for Early-phase Clinical Trials of Steatohepatitis in Adolescents

OBJECTIVES

Early-phase pediatric nonalcoholic fatty liver disease (NAFLD) clinical trials are designed with noninvasive parameters to assess potential efficacy. Increasingly, these parameters include magnetic resonance imaging (MRI)-derived proton density fat fraction (PDFF) and MR elastography (MRE)-derived shear stiffness as biomarkers of hepatic steatosis and fibrosis, respectively. Understanding fluctuations in these measures is essential for calculating trial sample sizes, interpreting results, and planning clinical drug trials in children with NAFLD. Lack of such data in children constitutes a critical knowledge gap. Therefore, the primary aim of this study was to assess whole-liver MRI-PDFF change in adolescents with nonalcoholic steatohepatitis (NASH) over 12 weeks.

METHODS

Adolescents 12 to 19 years with biopsy-proven NASH undergoing standard-of-care treatment were enrolled. Baseline and week-12 assessments of anthropometrics, transaminases, MRI-PDFF, and MRE stiffness were obtained.

RESULTS

Fifteen adolescents were included (mean age 15.7 [SD 2.9] years). Hepatic MRI-PDFF was stable over 12 weeks (mean absolute change -0.8%, P = 0.24). Correlation between baseline and week-12 values of MRI-PDFF was high (ICC = 0.97, 95% CI 0.90-0.99). MRE stiffness was stable (mean percentage change 2.7%, P = 0.44); correlation between baseline and week-12 values was moderate (ICC = 0.47; 95% CI 0-0.79). Changes in weight, BMI, and aminotransferases were not statistically significant.

CONCLUSION

In adolescents with NASH, fluctuations in hepatic MRI-PDFF and MRE stiffness over 12 weeks of standard-of-care were small. These data on the natural fluctuations in quantitative imaging biomarkers can serve as a reference for interventional trials in pediatric NASH and inform the interpretation and planning of clinical trials.

3D-printed biological cell phantom for testing 3D quantitative phase imaging systems

As the 3D quantitative phase imaging (QPI) methods mature, their further development calls for reliable tools and methods to characterize and compare their metrological parameters. We use refractive index engineering during two-photon laser photolithography to fabricate a life-scale phantom of a biological cell with internal structures that mimic optical and structural properties of mammalian cells. After verification with a number of reference techniques, the phantom is used to characterize the performance of a limited-angle holographic tomography microscope.

Quantitative Test-Retest Measurement of 68Ga-PSMA-HBED-CC in Tumor and Normal Tissue

The PET radiotracer ⁶⁸Ga-PSMA (prostate-specific membrane antigen)-HBED-CC (N,N'-bis [2-hydroxy-5-(carboxyethyl)benzyl]ethylenediamine-N,N'-diacetic acid) shows potential as an imaging biomarker for recurrent and metastatic prostate cancer. The purpose of this study was to determine the repeatability of ⁶⁸Ga-PSMA-HBED-CC in a test-retest trial in subjects with metastatic prostate adenocarcinoma. Methods: Subjects with metastatic prostate cancer underwent 2 PET/CT scans with ⁶⁸Ga-PSMA-HBED-CC within 14 d (mean, 6 ± 4 d). Lesions in bone, nodes, prostate/bed, and visceral organs, as well as representative normal tissues (salivary glands and spleen), were segmented separately by 2 readers. Absolute and percentage differences in SUV_max and SUV_mean were calculated for all test-retest regions. Repeatability was assessed using percentage difference, within-subject coefficient of variation (wCV), repeatability coefficient (RC), and Bland-Altman analysis. Results: Eighteen subjects were evaluated, 16 of whom demonstrated local or metastatic disease on ⁶⁸Ga-PSMA-HBED-CC PET/CT. In total, 136 lesions were segmented in bone (n = 99), nodes (n = 27), prostate/bed (n = 7), and viscera (n = 3). The wCV for SUV_max was 11.7% for bone lesions and 13.7% for nodes. The RC was ±32.5% SUV_max for bone lesions and ±37.9% SUV_max for nodal lesions, meaning 95% of the normal variability between 2 measurements will be within these numbers, so larger differences are likely attributable to true biologic changes in tumor rather than normal physiologic or measurement variability. wCV in the salivary glands and spleen was 8.9% and 10.7% SUV_mean, respectively. Conclusion: Repeatability measurements for PET/CT test-retests with ⁶⁸Ga-PSMA-HBED-CC showed a wCV of 12%-14% SUV_max and an RC of ±33%-38% SUV_max in bone and nodal lesions. These estimates are an important aspect of ⁶⁸Ga-PSMA-HBED-CC as a quantitative imaging biomarker. These estimates are similar to those reported for ¹⁸F-FDG, suggesting that ⁶⁸Ga-PSMA-HBED-CC PET/CT may be useful in monitoring response to therapy.

In Vivo Measurement of Brain Tissue Response After Irradiation: Comparison of T2 Relaxation, Apparent Diffusion Coefficient, and Electrical Conductivity

Radiation therapy (RT) has been widely used as a powerful treatment tool to address cancerous tissues because of its ability to control cell growth. Its ionizing radiation damages the DNA of cancerous tissues, leading to cell death. Medical imaging, however, still has limitations regarding the reliability of its assessment of tissue response and in predicting the treatment effect because of its inability to provide contrast information on the gradual, minute tissue changes after RT. A recently developed magnetic resonance (MR)-based conductivity imaging method may provide direct, highly sensitive information on this tissue response because its contrast mechanism is based on the concentration and mobility of ions in intracellular and extracellular spaces. In this feasibility study, we applied T2-weighted, diffusion-weighted, and electrical conductivity imaging to mouse brain, thus, using the MR imaging to map the tissue response after radiation exposure. To evaluate the degree of response, we measured the T2 relaxation, apparent diffusion coefficient (ADC), and electrical conductivity of brain tissues before and after irradiation. The conductivity images, which showed significantly higher sensitivity than other MR imaging methods, indicated that the contrast is distinguishable in different ways at different areas of the brain. Future studies will focus on verifying these results and the long-term evaluation of conductivity changes using various irradiation methods for clinical applications.

Harmonizing brain magnetic resonance imaging methods for vascular contributions to neurodegeneration

INTRODUCTION

Many consequences of cerebrovascular disease are identifiable by magnetic resonance imaging (MRI), but variation in methods limits multicenter studies and pooling of data. The European Union Joint Program on Neurodegenerative Diseases (EU JPND) funded the HARmoNizing Brain Imaging MEthodS for VaScular Contributions to Neurodegeneration (HARNESS) initiative, with a focus on cerebral small vessel disease.

METHODS

Surveys, teleconferences, and an in-person workshop were used to identify gaps in knowledge and to develop tools for harmonizing imaging and analysis.

RESULTS

A framework for neuroimaging biomarker development was developed based on validating repeatability and reproducibility, biological principles, and feasibility of implementation. The status of current MRI biomarkers was reviewed. A website was created at www.harness-neuroimaging.org with acquisition protocols, a software database, rating scales and case report forms, and a deidentified MRI repository.

CONCLUSIONS

The HARNESS initiative provides resources to reduce variability in measurement in MRI studies of cerebral small vessel disease.

REVIEW: MR elastography of brain tumors

MR elastography allows non-invasive quantification of the shear modulus of tissue, i.e. tissue stiffness and viscosity, information that offers the potential to guide presurgical planning for brain tumor resection. Here, we review brain tumor MRE studies with particular attention to clinical applications. Studies that investigated MRE in patients with intracranial tumors, both malignant and benign as well as primary and metastatic, were queried from the Pubmed/Medline database in August 2018. Reported tumor and normal appearing white matter stiffness values were extracted and compared as a function of tumor histopathological diagnosis and MRE vibration frequencies. Because different studies used different elastography hardware, pulse sequences, reconstruction inversion algorithms, and different symmetry assumptions about the mechanical properties of tissue, effort was directed to ensure that similar quantities were used when making inter-study comparisons. In addition, because different methodologies and processing pipelines will necessarily bias the results, when pooling data from different studies, whenever possible, tumor values were compared with the same subject's contralateral normal appearing white matter to minimize any study-dependent bias. The literature search yielded 10 studies with a total of 184 primary and metastatic brain tumor patients. The group mean tumor stiffness, as measured with MRE, correlated with intra-operatively assessed stiffness of meningiomas and pituitary adenomas. Pooled data analysis showed significant overlap between shear modulus values across brain tumor types. When adjusting for the same patient normal appearing white matter shear modulus values, meningiomas were the stiffest tumor-type. MRE is increasingly being examined for potential in brain tumor imaging and might have value for surgical planning. However, significant overlap of shear modulus values between a number of different tumor types limits applicability of MRE for diagnostic purposes. Thus, further rigorous studies are needed to determine specific clinical applications of MRE for surgical planning, disease monitoring and molecular stratification of brain tumors.

Quantitative imaging biomarkers in nuclear medicine: from SUV to image mining studies. Highlights from annals of nuclear medicine 2018

OBJECTIVE

Quantification in medical imaging is one of the main goals in research and clinical practice since it allows immediate understanding, objective communication, and comparison. Our aim was to summarize relevant investigations on quantification in nuclear medicine studies published in the volume 32 of Annals of Nuclear Medicine.

METHODS

In this article, we summarized the data of 14 selected papers from international research groups that were published between January and December 2018. This is a descriptive review with an inherently subjective selection of articles.

RESULTS

We discussed the role of parameters ranging from standardized uptake value to ratios, to flow within a region of interest, to volumetric parameters and to texture indices in different clinical scenarios in oncology, cardiology, and neurology.

CONCLUSIONS

In all the medical disciplines in which nuclear medicine examinations play a role, quantification is essential both in research and in clinical practice. Standardization and high-quality protocols are crucial for the success and reliability of imaging biomarkers.

Future Directions in Coronary CT Angiography: CT-Fractional Flow Reserve, Plaque Vulnerability, and Quantitative Plaque Assessment

Coronary computed tomography angiography (CCTA) is a well-validated and noninvasive imaging modality for the assessment of coronary artery disease (CAD) in patients with stable ischemic heart disease and acute coronary syndromes (ACSs). CCTA not only delineates the anatomy of the heart and coronary arteries in detail, but also allows for intra- and extraluminal imaging of coronary arteries. Emerging technologies have promoted new CCTA applications, resulting in a comprehensive assessment of coronary plaques and their clinical significance. The application of computational fluid dynamics to CCTA resulted in a robust tool for noninvasive assessment of coronary blood flow hemodynamics and determination of hemodynamically significant stenosis. Detailed evaluation of plaque morphology and identification of high-risk plaque features by CCTA have been confirmed as predictors of future outcomes, identifying patients at risk for ACSs. With quantitative coronary plaque assessment, the progression of the CAD or the response to therapy could be monitored by CCTA. The aim of this article is to review the future directions of emerging applications in CCTA, such as computed tomography (CT)-fractional flow reserve, imaging of vulnerable plaque features, and quantitative plaque imaging. We will also briefly discuss novel methods appearing in the coronary imaging scenario, such as machine learning, radiomics, and spectral CT.

Translating preclinical MRI methods to clinical oncology

The complexity of modern in vivo magnetic resonance imaging (MRI) methods in oncology has dramatically changed in the last 10 years. The field has long since moved passed its (unparalleled) ability to form images with exquisite soft-tissue contrast and morphology, allowing for the enhanced identification of primary tumors and metastatic disease. Currently, it is not uncommon to acquire images related to blood flow, cellularity, and macromolecular content in the clinical setting. The acquisition of images related to metabolism, hypoxia, pH, and tissue stiffness are also becoming common. All of these techniques have had some component of their invention, development, refinement, validation, and initial applications in the preclinical setting using in vivo animal models of cancer. In this review, we discuss the genesis of quantitative MRI methods that have been successfully translated from preclinical research and developed into clinical applications. These include methods that interrogate perfusion, diffusion, pH, hypoxia, macromolecular content, and tissue mechanical properties for improving detection, staging, and response monitoring of cancer. For each of these techniques, we summarize the 1) underlying biological mechanism(s); 2) preclinical applications; 3) available repeatability and reproducibility data; 4) clinical applications; and 5) limitations of the technique. We conclude with a discussion of lessons learned from translating MRI methods from the preclinical to clinical setting, and a presentation of four fundamental problems in cancer imaging that, if solved, would result in a profound improvement in the lives of oncology patients. Level of Evidence: 5 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2019;50:1377-1392.

Diffusion-weighted MRI in Advanced Epithelial Ovarian Cancer: Apparent Diffusion Coefficient as a Response Marker

Background Treatment of advanced epithelial ovarian cancer results in a relapse rate of 75%. Early markers of response would enable optimization of management and improved outcome in both primary and recurrent disease. Purpose To assess the apparent diffusion coefficient (ADC), derived from diffusion-weighted MRI, as an indicator of response, progression-free survival (PFS), and overall survival. Materials and Methods This prospective multicenter trial (from 2012-2016) recruited participants with stage III or IV ovarian, primary peritoneal, or fallopian tube cancer (newly diagnosed, cohort one; relapsed, cohort two) scheduled for platinum-based chemotherapy, with interval debulking surgery in cohort one. Cohort one underwent two baseline MRI examinations separated by 0-7 days to assess ADC repeatability; an additional MRI was performed after three treatment cycles. Cohort two underwent imaging at baseline and after one and three treatment cycles. ADC changes in responders and nonresponders were compared (Wilcoxon rank sum tests). PFS and overall survival were assessed by using a multivariable Cox model. Results A total of 125 participants (median age, 63.3 years [interquartile range, 57.0-70.7 years]; 125 women; cohort one, n = 47; cohort two, n = 78) were included. Baseline ADC (range, 77-258 × 10-5mm2s-1) was repeatable (upper and lower 95% limits of agreement of 12 × 10-5mm2s-1 [95% confidence interval {CI}: 6 × 10-5mm2s-1 to 18 × 10-5mm2s-1] and -15 × 10-5mm2s-1 [95% CI: -21 × 10-5mm2s-1 to -9 × 10-5mm2s-1]). ADC increased in 47% of cohort two after one treatment cycle, and in 58% and 53% of cohorts one and two, respectively, after three cycles. Percentage change from baseline differed between responders and nonresponders after three cycles (16.6% vs 3.9%; P = .02 [biochemical response definition]; 19.0% vs 6.2%; P = .04 [radiologic definition]). ADC increase after one cycle was associated with longer PFS in cohort two (adjusted hazard ratio, 0.86; 95% CI: 0.75, 0.98; P = .03). ADC change was not indicative of overall survival for either cohort. Conclusion After three cycles of platinum-based chemotherapy, apparent diffusion coefficient (ADC) changes are indicative of response. After one treatment cycle, increased ADC is indicative of improved progression-free survival in relapsed disease. Published under a CC BY 4.0 license. Online supplemental material is available for this article.

Automated measurement of liver attenuation to identify moderate-to-severe hepatic steatosis from chest CT scans

PURPOSE

Develop and validate an automated method for measuring liver attenuation in non-contrast low-dose chest CT (LDCT) scans and compare it to the standard manual method for identifying moderate-to-severe hepatic steatosis (HS).

METHOD

The automated method identifies a region below the right lung within the liver and uses statistical sampling techniques to exclude non-liver parenchyma. The method was used to assess moderate-to-severe HS on two IRB-approved cohorts: 1) 24 patients with liver disease examined between 1/2013-1/2017 with non-contrast chest CT and abdominal MRI scans obtained within three months of liver biopsy, and 2) 319 lung screening participants with baseline LDCT performed between 8/2011-1/2017. Agreement between the manual and automated CT methods, the manual MRI method, and pathology for determining moderate-to-severe HS was assessed using Cohen's Kappa by applying a 40 HU threshold to the CT method and 17.4% fat fraction to MRI. Agreement between the manual and automated CT methods was assessed using the intraclass correlation coefficient (ICC). Variability was assessed using Bland-Altman limits of agreement (LoA).

RESULTS

In the first cohort, the manual and automated CT methods had almost perfect agreement (ICC = 0.97, κ = 1.00) with LoA of -7.6 to 4.7 HU. Both manual and automated CT methods had almost perfect agreement with MRI (κ = 0.90) and substantial agreement with pathology (κ = 0.77). In the second cohort, the manual and automated CT methods had almost perfect agreement (ICC = 0.94, κ = 0.87). LoA were -10.6 to 5.2 HU.

CONCLUSION

Automated measurements of liver attenuation from LDCT scans can be used to identify moderate-to-severe HS on LDCT.

Emerging quantitative MR imaging biomarkers in inflammatory arthritides

PURPOSE

To review quantitative magnetic resonance imaging (qMRI) methods for imaging inflammation in connective tissues and the skeleton in inflammatory arthritis. This review is designed for a broad audience including radiologists, imaging technologists, rheumatologists and other healthcare professionals.

METHODS

We discuss the use of qMRI for imaging skeletal inflammation from both technical and clinical perspectives. We consider how qMRI can be targeted to specific aspects of the pathological process in synovium, cartilage, bone, tendons and entheses. Evidence for the various techniques from studies of both adults and children with inflammatory arthritis is reviewed and critically appraised.

RESULTS

qMRI has the potential to objectively identify, characterize and quantify inflammation of the connective tissues and skeleton in both adult and pediatric patients. Measurements of tissue properties derived using qMRI methods can serve as imaging biomarkers, which are potentially more reproducible and informative than conventional MRI methods. Several qMRI methods are nearing transition into clinical practice and may inform diagnosis and treatment decisions, with the potential to improve patient outcomes.

CONCLUSIONS

qMRI enables specific assessment of inflammation in synovium, cartilage, bone, tendons and entheses, and can facilitate a more consistent, personalized approach to diagnosis, characterisation and monitoring of disease.

Rapid automated liver quantitative susceptibility mapping

BACKGROUND

Accurate measurement of the liver iron concentration (LIC) is needed to guide iron-chelating therapy for patients with transfusional iron overload. In this work, we investigate the feasibility of automated quantitative susceptibility mapping (QSM) to measure the LIC.

PURPOSE

To develop a rapid, robust, and automated liver QSM for clinical practice.

STUDY TYPE

Prospective.

POPULATION

13 healthy subjects and 22 patients.

FIELD STRENGTH/SEQUENCES

1.5 T and 3 T/3D multiecho gradient-recalled echo (GRE) sequence.

ASSESSMENT

Data were acquired using a 3D GRE sequence with an out-of-phase echo spacing with respect to each other. All odd echoes that were in-phase (IP) were used to initialize the fat-water separation and field estimation (T2 *-IDEAL) before performing QSM. Liver QSM was generated through an automated pipeline without manual intervention. This IP echo-based initialization method was compared with an existing graph cuts initialization method (simultaneous phase unwrapping and removal of chemical shift, SPURS) in healthy subjects (n = 5). Reproducibility was assessed over four scanners at two field strengths from two manufacturers using healthy subjects (n = 8). Clinical feasibility was evaluated in patients (n = 22).

STATISTICAL TESTS

IP and SPURS initialization methods in both healthy subjects and patients were compared using paired t-test and linear regression analysis to assess processing time and region of interest (ROI) measurements. Reproducibility of QSM, R2 *, and proton density fat fraction (PDFF) among the four different scanners was assessed using linear regression, Bland-Altman analysis, and the intraclass correlation coefficient (ICC).

RESULTS

Liver QSM using the IP method was found to be ~5.5 times faster than SPURS (P < 0.05) in initializing T2 *-IDEAL with similar outputs. Liver QSM using the IP method were reproducibly generated in all four scanners (average coefficient of determination 0.95, average slope 0.90, average bias 0.002 ppm, 95% limits of agreement between -0.06 to 0.07 ppm, ICC 0.97).

DATA CONCLUSION

Use of IP echo-based initialization enables robust water/fat separation and field estimation for automated, rapid, and reproducible liver QSM for clinical applications.

LEVEL OF EVIDENCE

1 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:725-732.

A Phantom-Based Assessment of Repeatability and Reproducibility of Transvaginal Quantitative Ultrasound

This article evaluated the repeatability and reproducibility (R&R) of quantitative ultrasound (QUS) biomarkers attenuation coefficient (AC) and backscatter coefficient (BSC) in transvaginal QUS reference phantoms for obstetric applications. Five phantoms were scanned by three sonographers according to the scanning protocol. Each sonographer scanned each phantom with four transvaginal transducers of the same model (MC9-4) and three probe cover types (latex cover, nonlatex cover, and no cover). The AC and BSC were estimated by using a reference phantom method. The R&R analysis was performed for the frequency-averaged AC and logBSC (= 10log₁₀BSC) (5.4-5.8 MHz) by using three-factor random effects Analysis of Variance with interaction. The total R&R variabilities for AC and logBSC are small (AC: 0.042-0.065 dB/cm-MHz; logBSC: 0.50-0.68 dB), indicating high measurement precision. These values are small compared to the ranges of AC (0.28-0.99 dB/cm-MHz) and logBSC (-33.16 to -20.35 dB) of the five phantoms. The AC and logBSC biomarkers measured on transvaginal QUS phantoms using the reference phantom method are repeatable, and reproducible between sonographers, transducers, and probe covers.

Validated imaging biomarkers as decision-making tools in clinical trials and routine practice: current status and recommendations from the EIBALL* subcommittee of the European Society of Radiology (ESR)

Observer-driven pattern recognition is the standard for interpretation of medical images. To achieve global parity in interpretation, semi-quantitative scoring systems have been developed based on observer assessments; these are widely used in scoring coronary artery disease, the arthritides and neurological conditions and for indicating the likelihood of malignancy. However, in an era of machine learning and artificial intelligence, it is increasingly desirable that we extract quantitative biomarkers from medical images that inform on disease detection, characterisation, monitoring and assessment of response to treatment. Quantitation has the potential to provide objective decision-support tools in the management pathway of patients. Despite this, the quantitative potential of imaging remains under-exploited because of variability of the measurement, lack of harmonised systems for data acquisition and analysis, and crucially, a paucity of evidence on how such quantitation potentially affects clinical decision-making and patient outcome. This article reviews the current evidence for the use of semi-quantitative and quantitative biomarkers in clinical settings at various stages of the disease pathway including diagnosis, staging and prognosis, as well as predicting and detecting treatment response. It critically appraises current practice and sets out recommendations for using imaging objectively to drive patient management decisions.

The quantitative neuroradiology initiative framework: application to dementia

There are numerous challenges to identifying, developing and implementing quantitative techniques for use in clinical radiology, suggesting the need for a common translational pathway. We developed the quantitative neuroradiology initiative (QNI), as a model framework for the technical and clinical validation necessary to embed automated segmentation and other image quantification software into the clinical neuroradiology workflow. We hypothesize that quantification will support reporters with clinically relevant measures contextualized with normative data, increase the precision of longitudinal comparisons, and generate more consistent reporting across levels of radiologists' experience. The QNI framework comprises the following steps: (1) establishing an area of clinical need and identifying the appropriate proven imaging biomarker(s) for the disease in question; (2) developing a method for automated analysis of these biomarkers, by designing an algorithm and compiling reference data; (3) communicating the results via an intuitive and accessible quantitative report; (4) technically and clinically validating the proposed tool pre-use; (5) integrating the developed analysis pipeline into the clinical reporting workflow; and (6) performing in-use evaluation. We will use current radiology practice in dementia as an example, where radiologists have established visual rating scales to describe the degree and pattern of atrophy they detect. These can be helpful, but are somewhat subjective and coarse classifiers, suffering from floor and ceiling limitations. Meanwhile, several imaging biomarkers relevant to dementia diagnosis and management have been proposed in the literature; some clinically approved radiology software tools exist but in general, these have not undergone rigorous clinical validation in high volume or in tertiary dementia centres. The QNI framework aims to address this need. Quantitative image analysis is developing apace within the research domain. Translating quantitative techniques into the clinical setting presents significant challenges, which must be addressed to meet the increasing demand for accurate, timely and impactful clinical imaging services.

British Thyroid Association 2014 classification ultrasound scoring of thyroid nodules in predicting malignancy: Diagnostic performance and inter-observer agreement

Objectives

To assess the inter-observer agreement amongst five observers of differing levels of expertise in applying the British Thyroid Association (2014) guidelines for ultrasound scoring of thyroid nodules (BTA-U score) in the management of thyroid cancer, and to assess the U-score diagnostic performance in predicting malignancy.

Method

A total of 73 consecutive patients were included over a two-year period (July 2012 to July 2014), after referral to a tertiary head and neck oncology centre for ultrasound plus fine needle aspiration and cytology. Our five observers retrospectively and independently reviewed static ultrasound images on PACS and scored the thyroid nodules according to BTA-U classification. The observers were blinded to each other's scoring, cytology and histology results. Either the Kappa-statistic or intra-class correlation was used to assess the level of inter-observer agreement, plus agreement between the radiological and cytological diagnoses. The diagnostic performance of U-scoring for predicting final histological diagnosis was assessed with sensitivity, specificity, positive and negative predictive values.

Results

A Kappa-value of 0.73 (95% CI: 0.68-0.77) confirmed substantial inter-observer agreement amongst the five observers. All 17 histology confirmed malignant nodules were correctly classified as potentially malignant by all observers. The sensitivity and negative predictive value of BTA-U score in detecting and predicting malignancy were 100%, whereas the specificity and positive predictive values were 34% and 32%, respectively.

Conclusions

There is good inter-observer agreement in using the BTA-U score amongst different observers at differing levels of expertise. Adhering to BTA-U scoring can potentially achieve 100% sensitivity in selecting malignant nodules for sampling.

Measurement reproducibility of slice-interleaved T1 and T2 mapping sequences over 20 months: A single center study

BACKGROUND

Quantifying reproducibility of native T1 and T2 mapping over a long period (> 1 year) is necessary to assess whether changes in T1 and T2 over repeated sessions in a longitudinal study are associated with variability due to underlying tissue composition or technical confounders.

OBJECTIVES

To carry out a single-center phantom study to 1) investigate measurement reproducibility of slice-interleaved T1 (STONE) and T2 mapping over 20 months, 2) quantify sources of variability, and 3) compare reproducibility and measurements against reference spin-echo measurements.

METHODS

MR imaging was performed on a 1.5 Tesla Philips Achieva scanner every 2-3 weeks over 20 months using the T1MES phantom. In each session, slice-interleaved T1 and T2 mapping was repeated 3 times for 5 slices, and maps were reconstructed using both 2-parameter and 3-parameter fit models. Reproducibility between sessions, and repeatability between repetitions and slices were evaluated using coefficients of variation (CV). Different sources of variability were quantified using variance decomposition analysis. The slice-interleaved measurement was compared to the spin-echo reference and MOLLI.

RESULTS

Slice-interleaved T1 had excellent reproducibility and repeatability with a CV < 2%. The main sources of T1 variability were temperature in 2-parameter maps, and slice in 3-parameter maps. Superior between-session reproducibility to the spin-echo T1 was shown in 2-parameter maps, and similar reproducibility in 3-parameter maps. Superior reproducibility to MOLLI T1 was also shown. Similar measurements to the spin-echo T1 were observed with linear regression slopes of 0.94-0.99, but slight underestimation. Slice-interleaved T2 showed good reproducibility and repeatability with a CV < 7%. The main source of T2 variability was slice location/orientation. Between-session reproducibility was lower than the spin-echo T2 reference and showed good measurement agreement with linear regression slopes of 0.78-1.06.

CONCLUSIONS

Slice-interleaved T1 and T2 mapping sequences yield excellent long-term reproducibility over 20 months.