Humean Research Articles

Exploring contrast generalisation in deep learning-based brain MRI-to-CT synthesis

Background:Synthetic computed tomography (sCT) has been proposed and increasingly clinically adopted to enable magnetic resonance imaging (MRI)-based radiotherapy. Deep learning (DL) has recently demonstrated the ability to generate accurate sCT from fixed MRI acquisitions. However, MRI protocols may change over time or differ between centres resulting in low-quality sCT due to poor model generalisation. Purpose:investigating domain randomisation (DR) to increase the generalisation of a DL model for brain sCT generation. Methods:CT and corresponding T1-weighted MRI with/without contrast, T2-weighted, and FLAIR MRI from 95 patients undergoing RT were collected, considering FLAIR the unseen sequence where to investigate generalisation.A “Baseline” generative adversarial network was trained with/without the FLAIR sequence to test how a model performs without DR. Image similarity and accuracy of sCT-based dose plans were assessed against CT to select the best-performing DR approach against the Baseline. Results:The Baseline model had the poorest performance on FLAIR, with mean absolute error (MAE) = 106 ± 20.7 HU (mean ±σ). Performance on FLAIR significantly improved for the DR model with MAE = 99.0 ± 14.9 HU, but still inferior to the performance of the Baseline+FLAIR model (MAE = 72.6 ± 10.1 HU). Similarly, an improvement in γ-pass rate was obtained for DR vs Baseline. Conclusion:DR improved image similarity and dose accuracy on the unseen sequence compared to training only on acquired MRI. DR makes the model more robust, reducing the need for re-training when applying a model on sequences unseen and unavailable for retraining.

Prediction of exercise-induced desaturation in COPD patients without resting hypoxemia: a retrospective study

BackgroundThere is no universally accepted criterion for assessing exercise-induced desaturation (EID). The purpose of this study is to compare the two methods regularly used for determining EID in COPD patients, as well as to explore the risk factors and predictors related to EID.MethodsThe 6MWT was performed with continuous SpO2 monitoring on patients with stable COPD. Using two methods (method A: “SpO2rest–SpO2min ≥ 4% and/or SpO2min < 90%”, method B: “SpO2rest–SpO2end ≥ 4% and/or SpO2end < 90%”) as EID determination criteria to assess the incidence of EID. The differences and consistency of the two methods are compared. Moreover, we collected data through the pulmonary function test, mMRC dyspnea score, COPD assessment test, BODE index and CT-defined emphysema. Univariate and multivariate logistic regression analyses were used to identify factors affecting the EID. For the parameters that predict EID in 6MWT, a receiver operating characteristic (ROC) curve analysis was employed.ResultsThe analysis included 124 patients. The overall incidence of EID was 62.1% by using method A as the criterion and 51.6% by method B. All of the EID patients found by method B were included in the EID patients identified by method A, as well as 13 new-EID patients. The difference in diagnostic outcomes between the two approaches was not statistically significant (P > 0.05), but they were in excellent agreement (Kappa = 0.807, P = 0.001). Logistic regression analyses found that DLCO SB% pred, DLCO/VA% pred, CAT score, mean density, PD15, emphysema volume and %LAA were significant determinants of the EID. For predicting EID, the ROC analysis produced AUC and cutoffs of 0.689 and 50.45% (DLCO SB% pred), 0.707 and 75.0% (DLCO/VA% pred), 0.727 and 15 points (CAT score), 0.691 and − 955.00HU (PD15), 0.671 and − 856.46HU (mean density), 0.668 and 338.14 ml (emphysema volume) and 0.656 and 7.63% (%LAA), respectively.ConclusionsTwo methods evaluating EID in this research are in a good agreement, method A can find more EID patients by focusing on SpO2min. When conditions are constrained, it is also sufficient to assess EID in COPD patients by method B. In terms of the predictors of EID, DLCO SB% pred, DLCO/VA% pred, CAT score and CT-defined emphysema are all statistically significant test variables to determine EID.

Comparative Role of MDCT and FDG-PET/CT in the Diagnostic Evaluation of Mediastinal Mass Lesions: An Institutional Experience.

Background Mediastinal mass lesions span a wide histopathological and radiological spectrum. Partition of the mediastinum into specific compartments aids in differential diagnosis of mass lesions, assistance in biopsies, and other surgical procedures. Multidetector row computed tomography (MDCT) is a promising three-dimensional imaging tool allowing substantial anatomical volumes to be routinely covered with isotropic submillimeter spatial resolution to precisely localize lesions and biopsy needles for both benign and malignant disease lesions of the mediastinum. Objective The aim of this study was to categorize mass lesions according to the mediastinal compartments to study their MDCT characteristics and to provide a comparative role of fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) in the diagnostic evaluation of mediastinal mass lesions. Materials and Methods Patients with clinical or radiological suspicion of mediastinal lesions on the basis of an abnormal chest radiograph were referred to the department of radiodiagnosis at a tertiary care center between April 2015 and December 2019 for MDCT evaluation. A total of 80 cases were correlated with the histopathological diagnosis excluding aneurysms. Size, CT density (Hounsfield unit [HU] mean), and maximum standardized uptake value (SUV max ) of mediastinal and chest wall lesions were determined on FDG-PET/CT. Results This study included a total of 102 cases, 72 males and 29 females. Mediastinal mass lesions were most common in the age group 46 to 60 years. Anterior mediastinum ( n = 43, 42.2%) is the most commonly involved compartment followed by posterior mediastinum ( n = 37, 35.9%) and middle mediastinum ( n = 22, 21.8%). Transcompartmental involvement is more commonly seen involving the anterior and middle mediastinum. The SUV max , HU mean, and size were higher in malignant cases ( p = 0.001, p = 0.003, and p = 0.004, respectively). The current study found a cutoff value of 4.61 for SUV max to discriminate benign lesions from malignant ones with a sensitivity and specificity of 73.7 and 75.9%, respectively (area under the curve: 0.841, 95% confidence interval: 0.793–0.965, p = 0.0001). The values of SUV max and HU mean were higher in solid benign lesions than those of cystic benign lesions ( p = 0.007 and p = 0.003, respectively). Conclusion In the current study, MDCT has high diagnostic accuracy of ∼94% overall as compared with histopathology, and 97 and 92% for benign and malignant lesions, respectively, in the evaluation of mediastinal mass lesions. FDG-PET/CT may be complementary to conventional imaging methods for the evaluation of mediastinal and chest wall mass lesions. However, confirmatory tissue sampling is required to confirm PET positive findings for the definite diagnosis.

Associations between pericarotid fat density and image-based risk characteristics of carotid plaque

ObjectivesIn this study, we aimed to evaluate the associations between pericarotid fat density (PFD) and various risk characteristics of carotid plaque. MethodsWe retrospectively evaluated consecutive patients who were subjected to both high-resolution MRI and carotid artery CT angiography CTA at our institution between January 2016 and April 2021. The section of the carotid artery with the most severe lumen stenosis was selected from each patient for analysis. Two separated regions of interest (ROI) (each with an area of 2.5 mm2 and located at least 1 mm from the outer margin of the carotid artery wall) were defined in the perivascular fat tissue. The mean value of PFD (mean HU) was measured on the plaque side and the same axial non-plaque side. Then, the bilateral difference (D-value HU) was calculated (plaque side mean HU minus non-plaque side mean HU). According to carotid plaque risk characteristics (American Heart Association VI type [AHA VI], intraplaque hemorrhage [IPH], thinning and/or rupture of the fibrous cap [TRFC], lipid-rich necrotic core [LRNC], and calcification [CA]), the associations between PFD and five different risk characteristic subgroups were analyzed. The Student’s t-test, Mann-Whitney U test, and Chi-square test were used to compare differences between different risk subgroups. Receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive efficacy of PFD for carotid plaque risk characteristics. P < 0.05 was considered statistically significant. ResultsA total of 71 eligible patients (mean age 61.25 ± 10.35 years, 57 male) were examined in this study. For the plaque side and the non-plaque side, the mean PFD values were −36.25 ± 20.65 HU and −66.87 ± 15.00 HU, respectively. In the non-AHA VI and AHA VI subgroups, the values for the mean HU of the plaque side were −49.50 ± 20.53 and –33.55 ± 19.78, respectively (P = 0.014). The D-value HU was higher for the AHA VI group compared to the non-AHA VI group (33.61 ± 16.72 vs. 15.91 ± 14.52, respectively; P = 0.001). Compared to the non-IPH subgroup, the IPH subgroup had a higher mean HU value for the plaque side (-47.68 ± 18.26 vs. −29.63 ± 19.16, respectively; P < 0.001) and a higher D-value HU (17.80 ± 13.27 vs. 38.03 ± 15.46, respectively; P < 0.001). Compared to the low risk non-TRFC subgroup, the TRFC subgroup had a higher D-value HU (24.51 ± 16.16 vs. 33.55 ± 17.65, respectively; P = 0.042). The D-value of PFD was found to be a significant predictor of both AHA VI classification (AUC: 0.79; SE: 64.41%; SP: 83.33%; P = 0.0001) and IPH (AUC: 0.83; SE: 88.89%; SP: 65.38%; P < 0.0001). ConclusionOur study found that PFD was significantly associated with high risk AHA VI plaque characterization, IPH, and TRFC. Therefore, PFD has the potential to be used as an indirect clinical marker of plaque instability.

Deep learning-based synthetic CT generation for paediatric brain MR-only photon and proton radiotherapy

Background and PurposeTo enable accurate magnetic resonance imaging (MRI)-based dose calculations, synthetic computed tomography (sCT) images need to be generated. We aim at assessing the feasibility of dose calculations from MRI acquired with a heterogeneous set of imaging protocol for paediatric patients affected by brain tumours. Materials and methodsSixty paediatric patients undergoing brain radiotherapy were included. MR imaging protocols varied among patients, and data heterogeneity was maintained in train/validation/test sets. Three 2D conditional generative adversarial networks (cGANs) were trained to generate sCT from T1-weighted MRI, considering the three orthogonal planes and its combination (multi-plane sCT). For each patient, median and standard deviation (σ) of the three views were calculated, obtaining a combined sCT and a proxy for uncertainty map, respectively. The sCTs were evaluated against the planning CT in terms of image similarity and accuracy for photon and proton dose calculations. ResultsA mean absolute error of 61 ± 14 HU (mean±1σ) was obtained in the intersection of the body contours between CT and sCT. The combined multi-plane sCTs performed better than sCTs from any single plane. Uncertainty maps highlighted that multi-plane sCTs differed at the body contours and air cavities. A dose difference of −0.1 ± 0.3% and 0.1 ± 0.4% was obtained on the D > 90% of the prescribed dose and mean γ2%,2mm pass-rate of 99.5 ± 0.8% and 99.2 ± 1.1% for photon and proton planning, respectively. ConclusionAccurate MR-based dose calculation using a combination of three orthogonal planes for sCT generation is feasible for paediatric brain cancer patients, even when training on a heterogeneous dataset.

Carbon fiber\u2013reinforced PEEK versus titanium implants: an in vitro comparison of susceptibility artifacts in CT and MR imaging

Artifacts in computed tomography (CT) and magnetic resonance imaging (MRI) due to titanium implants in spine surgery are known to cause difficulties in follow-up imaging, radiation planning, and precise dose delivery in patients with spinal tumors. Carbon fiber–reinforced polyetheretherketon (CFRP) implants aim to reduce these artifacts. Our aim was to analyze susceptibility artifacts of these implants using a standardized in vitro model. Titanium and CFRP screw-rod phantoms were embedded in 3% agarose gel. Phantoms were scanned with Siemens Somatom AS Open and 3.0-T Siemens Skyra scanners. Regions of interest (ROIs) were plotted and analyzed for CT and MRI at clinically relevant localizations. CT voxel–based imaging analysis showed a significant difference of artifact intensity and central overlay between titanium and CFRP phantoms. For the virtual regions of the spinal canal, titanium implants (ti) presented − 30.7 HU vs. 33.4 HU mean for CFRP (p < 0.001), at the posterior margin of the vertebral body 68.9 HU (ti) vs. 59.8 HU (CFRP) (p < 0.001) and at the anterior part of the vertebral body 201.2 HU (ti) vs. 70.4 HU (CFRP) (p < 0.001), respectively. MRI data was only visually interpreted due to the low sample size and lack of an objective measuring system as Hounsfield units in CT. CT imaging of the phantom with typical implant configuration for thoracic stabilization could demonstrate a significant artifact reduction in CFRP implants compared with titanium implants for evaluation of index structures. Radiolucency with less artifacts provides a better interpretation of follow-up imaging, radiation planning, and more precise dose delivery.

Quantitative factors of unenhanced CT for predicting fragmenting efficacy of extracorporeal shock wave lithotripsy on pancreatic duct stones

To find potential predictors at unenhanced computed tomography (CT) to evaluate the stone clearance rate (SCR) of pancreatic duct stones (PDS) by extracorporeal shock wave lithotripsy (ESWL). A total of 106 consecutive patients with multiple stones of maximal transverse length (MTL) 0.3-3.6 cm in pancreatic head region were enrolled. Unenhanced CT was performed both before and after ESWL therapy. All patients underwent ESWL therapy using an electromagnetic lithotripter, and were divided into three groups with threshold values of >90%, 50-90% and <50% of SCR. Factors including sex, age, MTL, mean stone density (MSD), standard deviation of stone density (SDSD), variation coefficient of stone density (VCSD), attenuation density (AD), pancreatic duct diameter (PDD), skin-to-stone distance (SSD), and number of ESWL sessions were analysed, and their impact on SCR was evaluated by univariate and multivariate analyses. The diagnostic threshold of the independent predicting index was further calculated, and its relationship with SCR and ESWL was analysed by drawing a trend line. The only significant predictor of SCR by ESWL was MSD (p<0.05). More courses of ESWL therapy are required in PDS patients with an MSD higher than the cut-off value of 375.4 HU (mean, 2.6 versus 1.8), but with a relatively lower SCR (mean, 43.96 versus 88.68%) . MSD on pre-treatment unenhanced CT is an independent predictor for SCR of PDS patients treated with ESWL. Using a cut-off value of 375.4 HU, patients with low MSD are good candidates for ESWL therapy.

Sinhababu, Neil. Humean Nature: How Desire Explains Action, Thought, and Feeling

Sinhababu, Neil. Humean Nature: How Desire Explains Action, Thought, and Feeling

Humean Nature: How Desire Explains Action, Thought, and Feeling

Humean Nature: How Desire Explains Action, Thought, and Feeling

The Humean Way to China: Beyond the Stereotype

David Hume seems to receive several stereotypes and commonplace sentiments about China regarding its religion, national character, government, practices and economy, that he goes on to dismantle. Doing so, he allows the eighteenth-century reader to look at China from a different perspective. This perspective can still be useful especially today, when the “immense distance” between China and Europe has been reduced and, as Hume would say, almost everything we use is Chinese. In the name of an ambivalent European tradition, we are often inclined to revive these commonplace sentiments (for example, the uniformity of Chinese character) and neglect that part of our own tradition that tries to understand what is behind them and that also offers us the tools to go beyond them. This study endeavors to assemble an array of Hume’s scattered remarks, consider them in their context, and explore their possible sources in order to obtain not only a more Humean China, but also a more Chinese Hume.

Précis of "Humean Nature. How Desire Explains Action, Thought, and Feeling"

Précis of "Humean Nature. How Desire Explains Action, Thought, and Feeling"

Humean Nature: How Desire Explains Action, Thought, and Feeling, by Neil Sinhababu

Humean Nature: How Desire Explains Action, Thought, and Feeling, by Neil Sinhababu

The Humean Notion of Sympathy

The Humean Notion of Sympathy

Cloud storage for endosomes.

Coordinated regulation of vesicle trafficking is critical for the proper functioning of a cell. The bulk of cellular transport vesicles are sequestered in a “perinuclear cloud”, with only a small fraction released to the cell periphery. Jongsma et al (2016) found that the ER‐associated E3 ubiquitin ligase RNF26 is responsible for establishing and maintaining the architecture of the perinuclear cloud and that this spatiotemporal positioning is critical for effective regulation of the endocytic and exocytic systems.

The influence of cardiac function on coronary arterial enhancement at coronary computed tomography angiography: A cross-sectional study.

Background:The purpose of this study was to evaluate the influence of ejection fraction (EF) on peak aortic time (PAT) and peak aortic enhancement (PAE) during coronary computed tomography angiography (CTA).Materials and Methods:One-hundred and twenty patients (64 men, 56 women) underwent measurement of coronary CTA with a measurement of EF within 3 months of coronary CTA. Pearson's correlation coefficient analysis was used to investigate the relationships between EF, PAT and PAE, and peak attenuation of all coronary arteries.Results:The range of EF was (25%–70%) (mean: 55 ± 7.7). The range of PAT and PAE of ascending aorta on bolus test was 13–31 s (mean: 19.3 ± 2) and 153–435 HU (mean: 235 ± 40.6), respectively. Mean peak attenuation of ascending aorta, right coronary artery, left coronary artery, left circumflex artery, and left anterior descending were (561 ± 119), (476 ± 109), (505 ± 108), (467 ± 113), and (473 ± 104), respectively. There was a negative correlation between EF and PAT (r = −0.266, P = 0.003); however, there was no significant correlation between EF and PAE (r = −0.027, P = 0.767). In addition, there was no significant correlation between EF and the peak attenuation of coronary arteries.Conclusion:PAT was related to EF, but there was no relationship between PAE and EF. One of the explanation is that the left ventricular EF used for our study was assessed with echocardiography which is used roughly estimation of EF with interval of 5%–10% and may cause confounding results.

Characterization of intraventricular and intracerebral hematomas in non-contrast CT.

Characterization of hematomas is essential in scan reading, manual delineation, and designing automatic segmentation algorithms. Our purpose is to characterize the distribution of intraventricular (IVH) and intracerebral hematomas (ICH) in NCCT scans, study their relationship to gray matter (GM), and to introduce a new tool for quantitative hematoma delineation. We used 289 serial retrospective scans of 51 patients. Hematomas were manually delineated in a two-stage process. Hematoma contours generated in the first stage were quantified and enhanced in the second stage. Delineation was based on new quantitative rules and hematoma profiling, and assisted by a dedicated tool superimposing quantitative information on scans with 3D hematoma display. The tool provides: density maps (40-85HU), contrast maps (8/15HU), mean horizontal/vertical contrasts for hematoma contours, and hematoma contours below a specified mean contrast (8HU). White matter (WM) and GM were segmented automatically. IVH/ICH on serial NCCT is characterized by 59.0HU mean, 60.0HU median, 11.6HU standard deviation, 23.9HU mean contrast, -0.99HU/day slope, and -0.24 skewness (changing over time from negative to positive). Its 0.1(st)-99.9(th) percentile range corresponds to 25-88HU range. WM and GM are highly correlated (R (2)=0.88; p<10(-10)) whereas the GM-GS correlation is weak (R (2)=0.14; p<10(-10)). The intersection point of mean GM-hematoma density distributions is at 55.6±5.8HU with the corresponding GM/hematoma percentiles of 88(th)/40(th). Objective characterization of IVH/ICH and stating the rules quantitatively will aid raters to delineate hematomas more robustly and facilitate designing algorithms for automatic hematoma segmentation. Our two-stage process is general and potentially applicable to delineate other pathologies on various modalities more robustly and quantitatively.

SU‐F‐18C‐07: Automated CT QC Program with Analytics, Archival, and Notification Capabilities

Purpose:Tracking metrics over time is a well‐established means of establishing a quality control program. The number of metrics followed and testing frequency is limited by available resources. Automating the image analysis and data archival of a QC program enables objective and efficient tracking of performance metrics. The purpose of this study was to develop such a QC method and to assess its utility at a large clinical facility.Methods:The QC program at our institution is based on the acquisition of daily water‐phantom scans, and biweekly ACR‐phantom scans for each CT system. We developed a QC program to analyze these data. The QC software operates on the images sent directly to our server. The relevant information from DICOM headers was extracted, data analyzed, and a database was populated. The measurements performed on the waterphantom included water CT‐number, uniformity, noise, and artifact. The measurements performed on the ACR‐phantom included the MTF, NPS, detectability, artifact, uniformity, CNR, and the CT‐numbers for water, polyethylene, bone, air, and acrylic. Email notifications and criteria limits were directly based upon ACR accreditation requirements and developing threshold values.Results:Across ten clinical CT scanners, the daily water CT‐number was −0.2+/−1.4 HU(mean+/‐standard deviation). The corresponding numbers for 10%MTF, uniformity, CNR squared normalized to CTDI, and detectability squared normalized to CTDI were 0.69+/−0.01 (1/mm), 0.93+/−0.29, 0.06+/−0.02 (1/mGy), and 3.3+/−0.7 (1/mGy), respectively. For all ACR‐phantom inserts, the largest standard deviation for any individual scanner was 1.9 HU. Artifact analysis triggers successfully identified problematic images.Conclusion:Automating image analysis allows one to frequently track meaningful metrics that would be impractical to follow otherwise. System inconsistencies are more likely to be identified and corrected earlier. Much tighter system specific criteria limits are possible.

A Humean Insight into the Epistemic Normativity of the Belief in the Self

A Humean Insight into the Epistemic Normativity of the Belief in the Self

Very low-dose coronary artery calcium scanning with high-pitch spiral acquisition mode: Comparison between 120-kV and 100-kV tube voltage protocols

BackgroundEffective radiation dose from a single coronary artery calcification CT scan can range from 0.8 to 10.5 mSv, depending on the protocol. Reducing the effective radiation dose to reasonable levels without affecting diagnostic image quality can result in substantial dose reduction in CT. ObjectivesWe prospectively compared tube voltages of 120 and 100 kV in a low-dose CT acquisition protocol for measuring coronary artery calcified plaque with prospectively electrocardiogram (ECG)–triggered high-pitch spiral acquisition. MethodsIn 150 consecutive patients, measurement of coronary artery calcified plaque was performed with prospectively ECG-triggered high-pitch spiral acquisition. Imaging was first done with tube voltage of 120 kV voltage and subsequently repeated with 100 kV and otherwise unchanged parameters. CT was performed with a dual-source CT system with 280 milliseconds of rotation time, 2 × 128 slices, pitch of 3.4, triggered at 60% of the R–R interval. Tube current for both protocols was set at 80 mAs. With the use of a medium sharp reconstruction kernel (Siemens B35f), cross-sectional images were reconstructed with 3.0-mm slice thickness and 1.5-mm increment. Agatston scores were determined per patient for both scan settings by 2 independent readers with the use of a standard threshold of 130 HU for calcium detection. In addition, the Agatston score was calculated with a previously proposed threshold of 147 HU for 100-kV acquisitions. ResultsMean image noise was 20 ± 5 and 27 ± 7 for 120 and 100 kV, respectively (P < 0.0001). Mean dose length product was 24 ± 6 cm · cGy for the 120-kV protocol and 14 ± 4 cm · cGy for the 100-kV protocol, corresponding to average estimated effective doses of 0.3 and 0.2 mSv (P < 0.0001). Five patients were excluded from the analysis. In the remaining 145 patients, using the standard tube voltage of 120 kV, any coronary calcium was detected in 76 identical patients by both observers. In 75 of these patients, calcium was also identified by both observers in 100-kV data sets, whereas 1 patient was scored negative by 1 reader and was assigned an Agatston score of 0.7 (threshold, 130 HU) and 0.2 (threshold, 147 HU) by the other. Interobserver disagreement for assigning a patient a zero Agatston score was the same for both scan settings (each 4 patients). The mean Agatston scores for 120-kV and 100-kV (threshold, 147 HU) scans were 105 ± 245 (range, 0–1865) and 116 ± 261 (range, 0–1917), respectively (P < 0.0001). Bland-Altman analysis indicated a systematic overestimation of the Agatston score with tube voltage of 100 kV and threshold of 147 HU (mean difference, 11; 95% limits of agreement, 62 to -40). Similar results were observed for coronary calcium volume scores. ConclusionHigh-pitch spiral acquisition allows coronary calcium scoring with effective doses below 0.5 mSv. The use of 100-kV tube voltage further reduces effective radiation dose compared with the standard of 120 kV; however, it leads to significant overestimation of the Agatston score when the standard threshold of 130 HU is used. Adjusting the threshold to 147 HU leads to a better agreement compared with standard 120 kV protocols yet with a remaining systematic bias toward overestimation of the Agatston score. For high-pitch spiral acquisition mode, effective radiation dose reduction when using a 100-kV setting is minimal compared with the standard 120-kV setting and may be considered nonsignificant in a clinical setting.

096 A comparative study of standard filtered back projection with novel iterative reconstruction techniques in cardiac CT

BackgroundIterative reconstruction (IR) is a novel but significant development in CT image acquisition. There have been a number of studies that have reported on the potential of IR in cardiac...