Tomography

This study characterizes the error that results when performing quantitative analysis of abbreviated dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) data of the breast with the Standard Kety–Tofts (SKT) model and its Patlak variant. More specifically, we used simulations and patient data to determine the accuracy with which abbreviated time course data could reproduce the pharmacokinetic parameters, K^trans (volume transfer constant) and v_e (extravascular/extracellular volume fraction), when compared to the full time course data. SKT analysis of simulated abbreviated time courses (ATCs) based on the imaging parameters from two available datasets (collected with a 3T MRI scanner) at a temporal resolution of 15 s (N = 15) and 7.23 s (N = 15) found a concordance correlation coefficient (CCC) greater than 0.80 for ATCs of length 3.0 and 2.5 min, respectively, for the K^trans parameter. Analysis of the experimental data found that at least 90% of patients met this CCC cut-off of 0.80 for the ATCs of the aforementioned lengths. Patlak analysis of experimental data found that 80% of patients from the 15 s resolution dataset and 90% of patients from the 7.27 s resolution dataset met the 0.80 CCC cut-off for ATC lengths of 1.25 and 1.09 min, respectively. This study provides evidence for both the feasibility and potential utility of performing a quantitative analysis of abbreviated breast DCE-MRI in conjunction with acquisition of current standard-of-care high resolution scans without significant loss of information in the community setting. Full article

We aimed to evaluate radiomic features’ stability across different region of interest (ROI) sizes in CT and MR images. We chose a phantom with a homogenous internal structure so no differences for a feature extracted from ROIs of different sizes would be expected. For this, we scanned a plastic cup filled with sodium chloride solution ten times in CT and per MR sequence (T1-weighted-gradient-echo and T2-weighted-turbo-inversion-recovery-magnitude). We placed sphere-shaped ROIs of different diameters (4, 8, and 16 mm, and 4, 8, and 16 pixels) into the phantom’s center. Features were extracted using PyRadiomics. We assessed feature stability across ROI sizes with overall concordance correlation coefficients (OCCCs). Differences were tested for significance with the Mann–Whitney U-test. Of 93 features, 87 T1w-derived, 87 TIRM-derived, and 70 CT-derived features were significantly different between ROI sizes. Among MR-derived features, OCCCs showed excellent (>0.90) agreement for mean, median, and root mean squared for ROI sizes between 4 and 16 mm and pixels. We further observed excellent agreement for 10th and 90th percentile in T1w and 10th percentile in T2w TIRM images. There was no excellent agreement among the OCCCs of CT-derived features. In summary, many features indicated significant differences and only few showed excellent agreement across varying ROI sizes, although we examined a homogenous phantom. Since we considered a small phantom in an experimental setting, further studies to investigate this size effect would be necessary for a generalization. Nevertheless, we believe knowledge about this effect is crucial in interpreting radiomics studies, as features that supposedly discriminate disease entities may only indicate a systematic difference in ROI size. Full article

Glioblastoma (GBM) is the most common and deadly primary brain tumor in adults. Some of the genetic variations identified thus far, such as IDH mutation and MGMT promotor methylation, have implications for survival and response to therapy. A recent analysis of long-term GBM survivors showed that concurrent gain of chromosomes 19 and 20 (19/20 co-gain) is a positive prognostic factor that is independent of IDH mutation status. In this study, we retrospectively identified 18 patients with 19/20 co-gain and compared their imaging features to a control cohort without 19/20 co-gain. Imaging features such as tumor location, size, pial invasion, and ependymal extension were examined manually. When compared without further genetic subclassification, both groups showed similar imaging features except for rates of pial invasion. When each group was subclassified by MGMT promotor methylation status however, the two groups showed different imaging features in a number of additional ways including tumor location, size, and ependymal extension. Our results indicate that different permutations of various genetic mutations that coexist in GBM may interact in unpredictable ways to affect imaging appearance, and that imaging prognostication may be better approached in the context of the global genomic profile rather than individual genetic alterations. Full article

The retromolar canal is an anatomical variation that occurs in the mandibular bone. The retromolar canal typically originates in the mandibular canal on the distal side of the third molar and extends forward and upward to the retromolar foramen (RMF), which contains the neurovascular bundle. Accidentally damaging the neurovascular bundle in the retromolar canal during the extraction of the third molar, dental implant surgery, or maxillofacial orthognathic surgery may lead to subsequent complications such as incomplete local anesthesia, paresthesia, and bleeding during operation. The objective of this study was to investigate the prevalence of the RMF in the Taiwanese population in a medical center by using dental cone-beam computed tomography (CBCT) and to identify the position of the RMF in the mandibular bone. The dental CBCT images for the mandibular bone of 68 hemi-mandible were uploaded to the medical imaging software Mimics 15.1 to determine the prevalence of the RMF in the Taiwanese population and the three positional parameters of the RMF in the mandibular bone: (1) The diameter of the RMF, (2) the horizontal distance from the midpoint of the RMF to the distal cementoenamel junction of the second molar, and (3) the vertical distance from the midpoint of the RMF to the upper border of the mandibular canal. Seven RMFs were observed in the 68 hemi-mandibles. Thus, the RMF prevalence was 10.3%. In addition, the diameter of the RMF was 1.41 ± 0.30 mm (mean ± standard deviation), the horizontal distance from the midpoint of the RMF to the distal cementoenamel junction of the the second molar was 12.93 ± 2.87 mm, and the vertical distance from the midpoint of the RMF to the upper border of the mandibular canal below second molar was 13.62 ± 1.3487 mm. This study determined the prevalence of the RMF in the Taiwanese population in a medical center and its relative position in the mandibular bone. This information can provide clinicians with a reference for posterior mandible anesthesia and surgery to ensure medical safety. Full article

The purpose of this study was to compare the scan time, image quality and radiation dose of CT urograms (CTU) using a split vs. single bolus contrast media injection technique. A total of 241 consecutive CTUs performed between August 2019-February 2020 were retrospectively reviewed. There were three study groups: Group 1, <50 years old, 50/80 cc split-bolus administered at 0 and 700 s post initiation of injection, with combined nephrographic and excretory phases; group 2, ≥50 years old, same split-bolus protocol; and group 3, ≥50 years old, 130 cc single bolus injection, with nephrographic and excretory phases acquired at 100 s and 460 s post injection initiation. The recorded data elements were scan time, number of excretory phases, imaging quality based on opacification of the urinary collecting system (<50%, 50–75%, 75–100%), and dose-length product (DLP). Associations between group and categorical variables were assessed (Chi-square); mean scan time and DLP were compared (one-way ANOVA). Following analysis, proportionally fewer CTUs required a repeat excretory phase in group 3 (32/112, 28.6%) than in groups 1 (25/48, 52.1%) and 2 (37/80, 46.3%) (p = 0.006). Mean scan time was significantly lower in group 3 (678 s) than in groups 1 (1046 s) and 2 (978 s) (p < 0.0001). There was no association between groups and image quality (p = 0.13). DLP was higher in group 3 (1422 ± 837 mGy·cm) than in groups 1 (1041 ± 531 mGy·cm) and 2 (1137 ± 646 mGy·cm) (p = 0.003). In conclusion, single bolus CTU resulted in significantly fewer repeat phases and faster scan time at the expense of a slightly higher radiation dose. Full article

Objective: To ascertain non-cardiac abnormality (NCA) incidence in patients undergoing clinical cardiovascular magnetic resonance imaging (CMR) and determine such patients’ clinical importance. Methods: Consecutive patients undertaking CMR study from January 2012 to June 2017 for various cardiovascular diseases were enrolled. To assess NCA’s therapeutic importance, all incidental findings that were not expected from the patient’s history were analyzed. A careful review of medical history determines the information on mortality. Results: Three hundred and eighty-two consecutive patients (mean age 58 ± 11 years) who underwent CMR for different clinical indications were enrolled in the present study. Potentially significant results have been identified as abnormalities that require further clinical or radiological follow-up or therapy. On CMR, 118 NCA (30.9%) were found. In 25 patients, potential clinically significant NCAs, such as aortic aneurysm (n = 3), aortic dissection (n = 2), pulmonary thromboembolism (n = 2), and malignancy (n = 18), were identified (6.54%). In terms of one-year mortality data, in a patient without NCA, we observed a significantly higher survival rate than those with NCA (p = 0.0085) and a higher mortality rate in a patient with clinically significant NCA than a patient with irrelevant NCA (p = 0.02). Survival, as assessed via Kaplan‒Meier analysis, disclosed significantly higher mortality in the patients with clinically significant NCA than patients with irrelevant NCA (HR = 11.20, CI = 4.71–26.60, p < 0.001). Conclusions: We concluded that it is vital for the CMR study to determine the relevance of NCA, especially in the cholangiocarcinoma endemic region such as northeastern Thailand. Eventually, to reorganize the patients according to appropriate management, clinical correlation and prognosis must be summarily established. Full article

Abdominal aortic aneurysm (AAA) is a complex disease that requires regular imaging surveillance to monitor for aneurysm stability. Current imaging surveillance techniques use maximum diameter, often assessed by computed tomography angiography (CTA), to assess risk of rupture and determine candidacy for operative repair. However, maximum diameter measurements can be variable, do not reliably predict rupture risk and future AAA growth, and may be an oversimplification of complex AAA anatomy. Vascular deformation mapping (VDM) is a recently described technique that uses deformable image registration to quantify three-dimensional changes in aortic wall geometry, which has been previously used to quantify three-dimensional (3D) growth in thoracic aortic aneurysms, but the feasibility of the VDM technique for measuring 3D growth in AAA has not yet been studied. Seven patients with infra-renal AAAs were identified and VDM was used to identify three-dimensional maps of AAA growth. In the present study, we demonstrate that VDM is able to successfully identify and quantify 3D growth (and the lack thereof) in AAAs that is not apparent from maximum diameter. Furthermore, VDM can be used to quantify growth of the excluded aneurysm sac after endovascular aneurysm repair (EVAR). VDM may be a useful adjunct for surgical planning and appears to be a sensitive modality for detecting regional growth of AAAs. Full article

Objective: Cardiovascular magnetic resonance imaging (CMR) late gadolinium enhancement technique (LGE) detects thrombus rather than anatomical presence based on tissue properties and is theoretically highly accurate. The present study’s goal was to compare the diagnostic accuracy obtained with various CMR techniques and transthoracic echocardiography to diagnose left ventricular thrombus and evaluate the prevalence and perspectives of left ventricular (LV) thrombus among patients with impaired systolic left ventricular function. Methods: In a single academic referral center, a retrospective database review of all CMR assessments of the established left ventricular thrombus was carried out in 206 consecutive patients with reduced systolic function for five years. To assess thrombus risk factors, clinical and imaging parameters were analyzed. Sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), echocardiography, and cine-CMR sequence accuracy have been identified. LV structural parameters were quantified to detect markers for thrombus and predictors of the additive usefulness of contrast-enhanced thrombus imaging. Comparisons against LGE-CMR were made, which was used as the standard. Results: A 7.8 percent prevalence of left ventricular thrombus was identified by LGE-CMR. Cine-CMR increased the diagnostic efficiency for echocardiographic thrombus identification in this group, with sensitivity increasing from 50 percent by echocardiography to 75 percent by cine-CMR (p = 0.008). Dark blood CMR (DB-CMR) has better sensitivity and accuracy than echocardiography (p < 0.001), comparable to cine-CMR. The transmural infarct size was an independent marker for thrombus after correction for the LVEF and LV volume while considering only CMR parameters. There were significantly higher embolic events (HR = 71.33; CI 8.31–616.06, p < 0.0001) in LV thrombus patients detected by LGE-CMR. Conclusion: CMR imaging was more sensitive to left ventricular thrombi identification compared with transthoracic echocardiography. An additional parameter available from LGE-CMR and shown as an independent risk factor for left ventricular thrombus is the myocardial scar. Full article

(1) Background: This work characterizes the sensitivity of magnetic resonance-based Relaxivity Contrast Imaging (RCI) to Amyotrophic Lateral Sclerosis (ALS)-induced changes in myofiber microstructure. Transverse Relaxivity at Tracer Equilibrium (TRATE), an RCI-based parameter, was evaluated in the lower extremities of ALS patients and healthy subjects. (2) Methods: In this IRB-approved study, 23 subjects (12 ALS patients and 11 healthy controls) were scanned at 3T (Philips, The Netherlands). RCI data were obtained during injection of a gadolinium-based contrast agent. TRATE, fat fraction and T₂ measures, were compared in five muscle groups of the calf muscle, between ALS and control populations. TRATE was also evaluated longitudinally (baseline and 6 months) and was compared to clinical measures, namely ALS Functional Rating Scale (ALSFRS-R) and Hand-Held Dynamometry (HHD), in a subset of the ALS population. (3) Results: TRATE was significantly lower (p < 0.001) in ALS-affected muscle than in healthy muscle in all muscle groups. Fat fraction differences between ALS and healthy muscle were statistically significant for the tibialis anterior (p = 0.01), tibialis posterior (p = 0.004), and peroneus longus (p = 0.02) muscle groups but were not statistically significant for the medial (p = 0.07) and lateral gastrocnemius (p = 0.06) muscles. T₂ differences between ALS and healthy muscle were statistically significant for the tibialis anterior (p = 0.004), peroneus longus (p = 0.004) and lateral gastrocnemius (p = 0.03) muscle groups but were not statistically significant for the tibialis posterior (p = 0.06) and medial gastrocnemius (p = 0.07) muscles. Longitudinally, TRATE, averaged over all patients, decreased by 28 ± 16% in the tibialis anterior, 47 ± 18% in the peroneus longus, 25 ± 19% in the tibialis posterior, 29 ± 14% in the medial gastrocnemius and 35 ± 18% in the lateral gastrocnemius muscles between two timepoints. ALSFRS-R scores were stable in two of four ALS patients. HHD scores decreased in three of four ALS patients. (4) Conclusion: RCI-based TRATE was shown to consistently differentiate ALS-affected muscle from healthy muscle and also provide a quantitative measure of longitudinal muscle degeneration. Full article

Lung cancer causes more deaths globally than any other type of cancer. To determine the best treatment, detecting EGFR and KRAS mutations is of interest. However, non-invasive ways to obtain this information are not available. Furthermore, many times there is a lack of big enough relevant public datasets, so the performance of single classifiers is not outstanding. In this paper, an ensemble approach is applied to increase the performance of EGFR and KRAS mutation prediction using a small dataset. A new voting scheme, Selective Class Average Voting (SCAV), is proposed and its performance is assessed both for machine learning models and CNNs. For the EGFR mutation, in the machine learning approach, there was an increase in the sensitivity from 0.66 to 0.75, and an increase in AUC from 0.68 to 0.70. With the deep learning approach, an AUC of 0.846 was obtained, and with SCAV, the accuracy of the model was increased from 0.80 to 0.857. For the KRAS mutation, both in the machine learning models (0.65 to 0.71 AUC) and the deep learning models (0.739 to 0.778 AUC), a significant increase in performance was found. The results obtained in this work show how to effectively learn from small image datasets to predict EGFR and KRAS mutations, and that using ensembles with SCAV increases the performance of machine learning classifiers and CNNs. The results provide confidence that as large datasets become available, tools to augment clinical capabilities can be fielded. Full article

ACRIN 6687, a multi-center clinical trial evaluating differential response of bone metastases to dasatinib in men with metastatic castration-resistant prostate cancer (mCRPC), used [¹⁸F]-fluoride (NaF) PET imaging. We extend previous ACRIN 6687 dynamic imaging results by examining NaF whole-body (WB) static SUV PET scans acquired after dynamic scanning. Eighteen patients underwent WB NaF imaging prior to and 12 weeks into dasatinib treatment. Regional VOI analysis of the most NaF avid bone metastases and an automated whole-body method using Quantitative Total Bone Imaging software (QTBI; AIQ Solutions, Inc., Madison, WI, USA) were used. We assessed differences in tumor and normal bone, between pre- and on-treatment dasatinib, and evaluated parameters in association with PFS and OS. Significant decrease in average SUV_max and average SUV_peak occurred in response to dasatinib. Univariate and multivariate analysis showed NaF uptake had significant association with PFS. Pharmacodynamic changes with dasatinib in tumor bone can be identified by WB NaF PET in men with mCRPC. WB PET has the benefit of examining the entire body and is less complicated than single FOV dynamic imaging. Full article

Objective: Cardiac T2* magnetic resonance imaging (MRI) has recently attracted considerable attention as a non-invasive method for detecting iron overload in various organs in thalassemia major patients. This study aimed to identify the prevalence of cardiac siderosis in thalassemia major patients and evaluate cardiac T2* MRI for monitoring cardiac siderosis before and after patients receive iron chelation therapy and its relation to serum ferritin, left ventricular ejection fraction, and liver iron concentration. The information gathered would be used for the direct monitoring, detection, and treatment of complications early on. Methods: A total of 119 thalassemia major patients were recruited in the present study. The cardiac T2* MRI was compared to serum ferritin levels, liver iron concentration (LIC), and left ventricular ejection fraction. All patients were classified into four groups based on their cardiac siderosis as having normal, marginal, mild to moderate, or severe cardiac iron overload. At the follow-up at years one, three, and five, the cardiac T2* MRI, LIC, serum ferritin, and left ventricular ejection fraction (LVEF) were determined. Results: The prevalence of cardiac siderosis with cardiac T2* MRI ≤ 25 ms was 17.6% (n = 21). There was no correlation between cardiac T2* MRI and serum ferritin, liver iron concentration, and LVEF (p = 0.39, 0.54, and 0.09, respectively). During one year to five years’ follow-up periods, cardiac T2* MRI (ms) in patients with severe cardiac siderosis had significantly improved from 8.5 ± 1.49 at baseline to 33.9 ± 1.9 at five years (p < 0.0001). Patients with severe, mild-moderate, marginal, and no cardiac siderosis had median LIC (mg/g dw) of 23.9 ± 6.5, 21.6 ± 13.3, 25.3 ± 7.7, and 19.9 ± 5.5 at baseline, respectively. Conclusions: This study supports the use of cardiac T2* MRI to monitor cardiac iron overload in patients who have had multiple blood transfusions. Early diagnosis and treatment of patients at risk of cardiac siderosis is a reasonable method of reducing the substantial cardiac mortality burden associated with myocardial siderosis. Cardiac T2* MRI is the best test that can identify at-risk patients who can be managed with optimization of their chelation therapy. Full article

Accurate measurement of object volumes using computed tomography is often important but can be challenging, especially for finely convoluted objects with severe marginal blurring from volume averaging. We aimed to test the accuracy of a simple method for volumetry by constructing, scanning and analyzing a phantom object with these characteristics which consisted of a cluster of small lucite beads embedded in petroleum jelly. Our method involves drawing simple regions of interest containing the entirety of the object and a portion of the surrounding material and using its density, along with the densities of pure lucite and petroleum jelly and the slice thickness to calculate the volume of the object in each slice. Comparison of our results with the object’s true volume showed the technique to be highly accurate, irrespective of slice thickness, image noise, reconstruction planes, spatial resolution and variations in regions of interest. We conclude that the method can be easily used for accurate volumetry in clinical and research scans without the need for specialized volumetry computer programs. Full article

The presence of a swallow-tail sign in the nigrosome-1 with hyperintense signal shown on the susceptibility-weighted imaging (SWI) has been shown to be sensitive in detecting the abnormal iron deposits in this area. A systematic evaluation in healthy subjects is required before this tool can be recommended in a widespread application. We evaluated a simple and practical SWI approach using a multiecho gradient-echo sequence with an improved contrast-to-noise ratio (CNR). We also evaluated the association of the neuromelanin imaging contrast behavior with the susceptibility and relaxation measurements. Twenty-five older and 23 young healthy adults were evaluated. The CNRs of the nigrosome-1 were compared along with method and group. Correlations of the nigrosome-1 neuromelanin signal in the neuromelanin-sensitive imaging with CNRs in the susceptibility, T1 and T2 mappings were examined. Two different coils were used to confirm the reproducibility. Compared with the single-echo, multiecho SWI can improve the CNR of the swallow-tail sign. We found significant correlations between neuromelanin signal and CNRs in the susceptibility and T2 mappings, and T1 value. The older subjects exhibited increased CNRs compared with the young adults. No significant difference was observed in the measurements between 20 and 64 channels. The multiecho technique allows the high-quality nigrosome-1 images in SWI and allows for a joint analysis of T2* and quantitative-susceptibility mapping at high spatial resolution. The correlations of neuromelanin-sensitive imaging with susceptibility and T2 imply that the iron content in the nigrosome-1 may have significant influences on the hyperintensity of neuromelanin in the magnetization transfer-related contrast. Full article

[¹²³I]FP-CIT SPECT has been valuable for distinguishing Parkinson disease (PD) from essential tremor. However, its performance for quantitative assessment of motor dysfunction has not been established. A virtual reality (VR) application was developed and compared with [¹²³I]FP-CIT SPECT/CT for detection of severity of motor dysfunction. Forty-four patients (21 males, 23 females, age 64.5 ± 12.4) with abnormal [¹²³I]FP-CIT SPECT/CT underwent assessment of bradykinesia, activities of daily living, and tremor with VR. Support vector machines (SVM) machine learning models were applied to VR and SPECT data. Receiver operating characteristic (ROC) analysis demonstrated greater area under the curve (AUC) for VR (0.8418, 95% CI 0.6071–0.9617) compared with brain SPECT (0.5357, 95% CI 0.3373–0.7357, p = 0.029) for detection of motor dysfunction. Logistic regression identified VR as an independent predictor of motor dysfunction (Odds Ratio 326.4, SE 2.17, p = 0.008). SVM for prediction of the Unified Parkinson’s Disease Rating Scale Part III (UPDRS-III) demonstrated greater R-squared of 0.713 (p = 0.008) for VR, compared with 0.0764 (p = 0.361) for brain SPECT. This study demonstrates that VR can be safely used in patients prior to [¹²³I]FP-CIT SPECT imaging and may improve prediction of motor dysfunction. This test has the potential to provide a simple, objective, quantitative analysis of motor symptoms in PD patients. Full article

Upper urinary tract obstructions (UTOs) are blockages that inhibit the flow of urine through its normal course, leading to impaired kidney function. Imaging plays a significant role in the initial diagnosis of UTO, with anatomic imaging (primarily ultrasound (US) and non-contrast computed tomography (CT)) serving as screening tools for the detection of the dilation of the urinary collecting systems (i.e., hydronephrosis). Whether hydronephrosis represents UTO or a non-obstructive process is determined by functional imaging (typically nuclear medicine renal scintigraphy). If these exams reveal evidence of UTO but no discernable source, multiphase contrast enhanced CT urography and/or dynamic contrast enhanced MR urography (DCE-MRU) may be performed to delineate a cause. These are often performed in conjunction with direct ureteroscopic evaluation. While contrast-enhanced CT currently predominates, it can induce renal injury due to contrast induced nephropathy (CIN), subject patients to ionizing radiation and is limited in quantifying renal function (traditionally assessed by renal scintigraphy) and establishing the extent to which hydronephrosis is due to functional obstruction. Traditional MRI is similarly limited in its ability to quantify function. DCE-MRU presents concerns regarding nephrogenic systemic fibrosis (NSF), although decreased with newer gadolinium-based contrast agents, and regarding cumulative gadolinium deposition in the basal ganglia. DCE-MR CEST urography is a promising alternative, employing new MRI contrast agents and imaging schemes and allowing for concurrent assessment of renal anatomy and functional parameters. In this review we highlight clinical challenges in the diagnosis and management of UTO, identify key advances in imaging agents and techniques for DCE-MR CEST urography and provide perspective on how this technique may evolve in clinical importance. Full article

Respiratory motion and increased susceptibility effects at high magnetic fields pose challenges for quantitative diffusion-weighted MRI (DWI) of a mouse abdomen on preclinical MRI systems. We demonstrate the first application of radial k-space-sampled (RAD) DWI of a mouse abdomen using a genetically engineered mouse model of pancreatic ductal adenocarcinoma (PDAC) on a 4.7 T preclinical scanner equipped with moderate gradient capability. RAD DWI was compared with the echo-planar imaging (EPI)-based DWI method with similar voxel volumes and acquisition times over a wide range of b-values (0.64, 535, 1071, 1478, and 2141 mm²/s). The repeatability metrics are assessed in a rigorous test–retest study (n = 10 for each DWI protocol). The four-shot EPI DWI protocol leads to higher signal-to-noise ratio (SNR) in diffusion-weighted images with persisting ghosting artifacts, whereas the RAD DWI protocol produces relatively artifact-free images over all b-values examined. Despite different degrees of motion mitigation, both RAD DWI and EPI DWI allow parametric maps of apparent diffusion coefficients (ADC) to be produced, and the ADC of the PDAC tumor estimated by the two methods are 1.3 ± 0.24 and 1.5 ± 0.28 × 10⁻³ mm²/s, respectively (p = 0.075, n = 10), and those of a water phantom are 3.2 ± 0.29 and 2.8 ± 0.15 × 10⁻³ mm²/s, respectively (p = 0.001, n = 10). Bland-Altman plots and probability density function reveal good repeatability for both protocols, whose repeatability metrics do not differ significantly. In conclusion, RAD DWI enables a more effective respiratory motion mitigation but lower SNR, while the performance of EPI DWI is expected to improve with more advanced gradient hardware. Full article

We propose a novel framework for determining radiomics feature robustness by considering the effects of both biological and noise signals. This framework is preliminarily tested in a study predicting the epidermal growth factor receptor (EGFR) mutation status in non-small cell lung cancer (NSCLC) patients. Pairs of CT images (baseline, 3-week post therapy) of 46 NSCLC patients with known EGFR mutation status were collected and a FDA-customized anthropomorphic thoracic phantom was scanned on two vendors’ scanners at four different tube currents. Delta radiomics features were extracted from the NSCLC patient CTs and reproducible, non-redundant, and informative features were identified. The feature value differences between EGFR mutant and EGFR wildtype patients were quantitatively measured as the biological signal. Similarly, radiomics features were extracted from the phantom CTs. A pairwise comparison between settings resulted in a feature value difference that was quantitatively measured as the noise signal. Biological signals were compared to noise signals at each setting to determine if the distributions were significantly different by two-sample t-test, and thus robust. Four optimal features were selected to predict EGFR mutation status, Tumor-Mass, Sigmoid-Offset-Mean, Gabor-Energy and DWT-Energy, which quantified tumor mass, tumor-parenchyma density transition at boundary, line-like pattern inside tumor and intratumoral heterogeneity, respectively. The first three variables showed robustness across the majority of studied CT acquisition parameters. The textual feature DWT-Energy was less robust. The proposed framework was able to determine robustness of radiomics features at specific settings by comparing biological signal to noise signal. Identification of robust radiomics features may improve the generalizability of radiomics models in future studies. Full article