Publications

Research suggests electroconvulsive therapy (ECT) induces an acute neuroinflammatory response and changes in white matter (WM) structural connectivity. However, whether these processes are related, either to each other or to eventual treatment outcomes, has yet to be determined. We examined the relationship between levels of peripheral pro-inflammatory cytokines and diffusion imaging-indexed changes in WM microstructure in individuals with treatment-resistant depression (TRD) who underwent ECT. Forty-two patients were assessed at baseline, after their second ECT (T2), and after completion of ECT (T3). A Montgomery Åsberg Depression Rating Scale improvement of >50% post-ECT defined ECT-responders (n = 19) from non-responders (n = 23). Thirty-four controls were also examined. Tissue-specific fractional anisotropy (FAt) was estimated using diffusion imaging data and the Free-Water method in 17 WM tracts. Inflammatory panels were evaluated from peripheral blood. Cytokines were examined to characterize the association between potential ECT-induced changes in an inflammatory state and WM microstructure. Longitudinal trajectories of both measures were also examined separately for ECT-responders and non-responders. Patients exhibited elevated Interleukin-8 (IL-8) levels at baseline compared to controls. In patients, correlations between IL-8 and FAt changes from baseline to T2 were significant in the positive direction in the right superior longitudinal fasciculus (R-SLF) and right cingulum (R-CB) (psig = 0.003). In these tracts, linear mixed-effects models revealed that trajectories of IL-8 and FAt were significantly positively correlated across all time points in responders, but not non-responders (R-CB-p = .001; R-SLF-p = 0.008). Our results suggest that response to ECT in TRD may be mediated by IL-8 and WM microstructure.

PURPOSE: NousNav is a complete low-cost neuronavigation system that aims to democratize access to higher-quality healthcare in lower-resource settings. NousNav’s goal is to provide a model for local actors to be able to reproduce, build and operate a fully functional neuronavigation system at an affordable cost. METHODS: NousNav is entirely open source and relies on low-cost off-the-shelf components, which makes it easy to reproduce and deploy in any region. NousNav’s software is also specifically devised with the low-resource setting in mind. RESULTS: It offers means for intuitive intraoperative control. The designed interface is also clean and simple. This allows for easy intraoperative use by either the practicing clinician or a nurse. It thus alleviates the need for a dedicated technician for operation. CONCLUSION: A prototype implementation of the design was built. Hardware and algorithms were designed for robustness, ruggedness, modularity, to be standalone and data-agnostic. The built prototype demonstrates feasibility of the objectives.

Gambling Disorder (GD) is a condition constituting a public health concern, with a burden of harm which is much greater than that of drug addiction. Patients with GD are generally reluctant to pharmacologic treatment and seem to prefer nonpharmacological interventions. Therefore, this proof-of-concept study aimed to investigate the feasibility of continuous Theta Burst Stimulation (cTBS) on the pre-SMA in six patients (5 males, 1 female), aged 30-64 years, with a DSM-5 diagnosis of Gambling Disorder and no comorbid mood disorders. Participants received over 10 sessions of Continuous TBS (cTBS) over pre-SMA bilaterally and have been evaluated using rating scales, including the PG-YBOCS and the CGI, before treatment (T0), at day 10 of treatment (T1) and at day 30 after treatment (T2); cTBS intervention was safe and without side effects. Since the design of our study does not allow us to draw conclusions on the effectiveness of the intervention with respect to the improvement of the functioning of the subject with GD, a more in-depth study, including a sham condition, neurocognitive measures of disinhibition and decision making, and collecting follow-up data on the sustained effect of TBS over a longer period is ongoing.

Background The clinical impact of interstitial lung abnormalities (ILAs) on poor prognosis has been reported in many studies, but risk stratification in ILA will contribute to clinical practice. Purpose To investigate the association of traction bronchiectasis/bronchiolectasis index (TBI) with mortality and clinical outcomes in individuals with ILA by using the COPDGene cohort. Materials and Methods This study was a secondary analysis of prospectively collected data. Chest CT scans of participants with ILA for traction bronchiectasis/bronchiolectasis were evaluated and outcomes were compared with participants without ILA from the COPDGene study (January 2008 to June 2011). TBI was classified as follows: TBI-0, ILA without traction bronchiectasis/bronchiolectasis; TBI-1, ILA with bronchiolectasis but without bronchiectasis or architectural distortion; TBI-2, ILA with mild to moderate traction bronchiectasis; and TBI-3, ILA with severe traction bronchiectasis and/or honeycombing. Clinical outcomes and overall survival were compared among the TBI groups and the non-ILA group by using multivariable linear regression model and Cox proportional hazards model, respectively. Results Overall, 5295 participants (median age, 59 years; IQR, 52-66 years; 2779 men) were included, and 582 participants with ILA and 4713 participants without ILA were identified. TBI groups were associated with poorer clinical outcomes such as quality of life scores in the multivariable linear regression model (TBI-0: coefficient, 3.2 [95% CI: 0.6, 5.7; P = .01]; TBI-1: coefficient, 3.3 [95% CI: 1.1, 5.6; P = .003]; TBI-2: coefficient, 7.6 [95% CI: 4.0, 11; P < .001]; TBI-3: coefficient, 32 [95% CI: 17, 48; P < .001]). The multivariable Cox model demonstrated that ILA without traction bronchiectasis (TBI-0-1) and with traction bronchiectasis (TBI-2-3) were associated with shorter overall survival (TBI-0-1: hazard ratio [HR], 1.4 [95% CI: 1.0, 1.9; P = .049]; TBI-2-3: HR, 3.8 [95% CI: 2.6, 5.6; P < .001]). Conclusion Traction bronchiectasis/bronchiolectasis was associated with poorer clinical outcomes compared with the group without interstitial lung abnormalities; TBI-2 and 3 were associated with shorter survival. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Lee and Im in this issue.

INTRODUCTION: A plasma-derived factor VIII product (pdFVIII, Factane 100 or 200 IU/mL) and a plasma-derived von Willebrand factor product (pdVWF, Wilfactin 100 IU/mL) are approved for replacement therapy by intravenous bolus injections in haemophilia A (HA) and von Willebrand disease (VWD), respectively. However, in situations requiring intensive treatment, continuous infusion (CI) may be desirable to better control target plasma factor levels. AIM: To evaluate the perioperative haemostatic efficacy and safety of these concentrates administered by CI. METHODS: Three phase III trials were conducted. Adults with HA (FVIII:C <1%) (Studies 1 and 2) or VWD (VWF:RCo <20%) (Study 3) received a preoperative bolus followed by CI of undiluted concentrate for at least 6 days. Bolus doses and CI rates were based on individual recovery and clearance, respectively. Initial infusion rate had to be higher for 48 hours for HA and 24 hours for VWD patients to anticipate potential fluctuations of factor concentrations during major surgery. Target levels of FVIII:C in HA and VWF:RCo in VWD were 80 and 70 IU/dL, respectively. Efficacy was assessed using a global haemostatic efficacy score. RESULTS: Studies 1, 2, and 3 included 12, 4 and 6 patients, respectively. Efficacy outcomes were excellent/good in all 22 major surgeries including 18 orthopaedic procedures. Most daily measured FVIII and VWF levels (92%) were on target. No safety concerns, thrombotic events or inhibitors were identified. CONCLUSION: pdFVIII and pdVWF administered by CI represent an effective and safe alternative to bolus injections in patients with severe HA or VWD undergoing surgery.

PURPOSE: Although both relaxation and diffusion imaging are sensitive to tissue microstructure, studies have reported limited sensitivity and robustness of using relaxation or conventional diffusion alone to characterize tissue microstructure. Recently, it has been shown that tensor-valued diffusion encoding and joint relaxation-diffusion quantification enable more reliable quantification of compartment-specific microstructural properties. However, scan times to acquire such data can be prohibitive. Here, we aim to simultaneously quantify relaxation and diffusion using MR fingerprinting (MRF) and b-tensor encoding in a clinically feasible time. METHODS: We developed multidimensional MRF scans (mdMRF) with linear and spherical b-tensor encoding (LTE and STE) to simultaneously quantify T1, T2, and ADC maps from a single scan. The image quality, accuracy, and scan efficiency were compared between the mdMRF using LTE and STE. Moreover, we investigated the robustness of different sequence designs to signal errors and their impact on the maps. RESULTS: T1 and T2 maps derived from the mdMRF scans have consistently high image quality, while ADC maps are sensitive to different sequence designs. Notably, the fast imaging steady state precession (FISP)-based mdMRF scan with peripheral pulse gating provides the best ADC maps that are free of image distortion and shading artifacts. CONCLUSION: We demonstrated the feasibility of quantifying T1, T2, and ADC maps simultaneously from a single mdMRF scan in around 24 s/slice. The map quality and quantitative values are consistent with the reference scans.

The main objective of this work is to develop and evaluate an artificial intelligence system based on deep learning capable of automatically identifying, quantifying, and characterizing COVID-19 pneumonia patterns in order to assess disease severity and predict clinical outcomes, and to compare the prediction performance with respect to human reader severity assessment and whole lung radiomics. We propose a deep learning based scheme to automatically segment the different lesion subtypes in nonenhanced CT scans. The automatic lesion quantification was used to predict clinical outcomes. The proposed technique has been independently tested in a multicentric cohort of 103 patients, retrospectively collected between March and July of 2020. Segmentation of lesion subtypes was evaluated using both overlapping (Dice) and distance-based (Hausdorff and average surface) metrics, while the proposed system to predict clinically relevant outcomes was assessed using the area under the curve (AUC). Additionally, other metrics including sensitivity, specificity, positive predictive value and negative predictive value were estimated. 95% confidence intervals were properly calculated. The agreement between the automatic estimate of parenchymal damage (%) and the radiologists’ severity scoring was strong, with a Spearman correlation coefficient (R) of 0.83. The automatic quantification of lesion subtypes was able to predict patient mortality, admission to the Intensive Care Units (ICU) and need for mechanical ventilation with an AUC of 0.87, 0.73 and 0.68 respectively. The proposed artificial intelligence system enabled a better prediction of those clinically relevant outcomes when compared to the radiologists’ interpretation and to whole lung radiomics. In conclusion, deep learning lesion subtyping in COVID-19 pneumonia from noncontrast chest CT enables quantitative assessment of disease severity and better prediction of clinical outcomes with respect to whole lung radiomics or radiologists’ severity score.

Quantification methods based on the acquisition of diffusion magnetic resonance imaging (dMRI) with multiple diffusion weightings (e.g., multi-shell) are becoming increasingly applied to study the in-vivo brain. Compared to single-shell data for diffusion tensor imaging (DTI), multi-shell data allows to apply more complex models such as diffusion kurtosis imaging (DKI), which attempts to capture both diffusion hindrance and restriction effects, or biophysical models such as NODDI, which attempt to increase specificity by separating biophysical components. Because of the strong dependence of the dMRI signal on the measurement hardware, DKI and NODDI metrics show scanner and site differences, much like other dMRI metrics. These effects limit the implementation of multi-shell approaches in multicenter studies, which are needed to collect large sample sizes for robust analyses. Recently, a post-processing technique based on rotation invariant spherical harmonics (RISH) features was introduced to mitigate cross-scanner differences in DTI metrics. Unlike statistical harmonization methods, which require repeated application to every dMRI metric of choice, RISH harmonization is applied once on the raw data, and can be followed by any analysis. RISH features harmonization has been tested on DTI features but not its generalizability to harmonize multi-shell dMRI. In this work, we investigated whether performing the RISH features harmonization of multi-shell dMRI data removes cross-site differences in DKI and NODDI metrics while retaining longitudinal effects. To this end, 46 subjects underwent a longitudinal (up to 3 time points) two-shell dMRI protocol at 3 imaging sites. DKI and NODDI metrics were derived before and after harmonization and compared both at the whole brain level and at the voxel level. Then, the harmonization effects on cross-sectional and on longitudinal group differences were evaluated. RISH features averaged for each of the 3 sites exhibited prominent between-site differences in the frontal and posterior part of the brain. Statistically significant differences in fractional anisotropy, mean diffusivity and mean kurtosis were observed both at the whole brain and voxel level between all the acquisition sites before harmonization, but not after. The RISH method also proved effective to harmonize NODDI metrics, particularly in white matter. The RISH based harmonization maintained the magnitude and variance of longitudinal changes as compared to the non-harmonized data of all considered metrics. In conclusion, the application of RISH feature based harmonization to multi-shell dMRI data can be used to remove cross-site differences in DKI metrics and NODDI analyses, while retaining inherent relations between longitudinal acquisitions.

Following the success of deep learning in a wide range of applications, neural network-based machine-learning techniques have received significant interest for accelerating magnetic resonance imaging (MRI) acquisition and reconstruction strategies. A number of ideas inspired by deep learning techniques for computer vision and image processing have been successfully applied to nonlinear image reconstruction in the spirit of compressed sensing for accelerated MRI. Given the rapidly growing nature of the field, it is imperative to consolidate and summarize the large number of deep learning methods that have been reported in the literature, to obtain a better understanding of the field in general. This article provides an overview of the recent developments in neural-network based approaches that have been proposed specifically for improving parallel imaging. A general background and introduction to parallel MRI is also given from a classical view of k-space based reconstruction methods. Image domain based techniques that introduce improved regularizers are covered along with k-space based methods which focus on better interpolation strategies using neural networks. While the field is rapidly evolving with plenty of papers published each year, in this review, we attempt to cover broad categories of methods that have shown good performance on publicly available data sets. Limitations and open problems are also discussed and recent efforts for producing open data sets and benchmarks for the community are examined.

Background Stratifying high-risk histopathologic features in endometrial carcinoma is important for treatment planning. Radiomics analysis at preoperative MRI holds potential to identify high-risk phenotypes. Purpose To evaluate the performance of multiparametric MRI three-dimensional radiomics-based machine learning models for differentiating low- from high-risk histopathologic markers-deep myometrial invasion (MI), lymphovascular space invasion (LVSI), and high-grade status-and advanced-stage endometrial carcinoma. Materials and Methods This dual-center retrospective study included women with histologically proven endometrial carcinoma who underwent 1.5-T MRI before hysterectomy between January 2011 and July 2015. Exclusion criteria were tumor diameter less than 1 cm, missing MRI sequences or histopathology reports, neoadjuvant therapy, and malignant neoplasms other than endometrial carcinoma. Three-dimensional radiomics features were extracted after tumor segmentation at MRI (T2-weighted, diffusion-weighted, and dynamic contrast-enhanced MRI). Predictive features were selected in the training set with use of random forest (RF) models for each end point, and trained RF models were applied to the external test set. Five board-certified radiologists conducted MRI-based staging and deep MI assessment in the training set. Areas under the receiver operating characteristic curve (AUCs) were reported with balanced accuracies, and radiologists’ readings were compared with radiomics with use of McNemar tests. Results In total, 157 women were included: 94 at the first institution (training set; mean age, 66 years ± 11 [SD]) and 63 at the second institution (test set; 67 years ± 12). RF models dichotomizing deep MI, LVSI, high grade, and International Federation of Gynecology and Obstetrics (FIGO) stage led to AUCs of 0.81 (95% CI: 0.68, 0.88), 0.80 (95% CI: 0.67, 0.93), 0.74 (95% CI: 0.61, 0.86), and 0.84 (95% CI: 0.72, 0.92), respectively, in the test set. In the training set, radiomics provided increased performance compared with radiologists’ readings for identifying deep MI (balanced accuracy, 86% vs 79%; P = .03), while no evidence of a difference was observed in performance for advanced FIGO stage (80% vs 78%; P = .27). Conclusion Three-dimensional radiomics can stratify patients by using preoperative MRI according to high-risk histopathologic end points in endometrial carcinoma and provide nonsignificantly different or higher performance than radiologists in identifying advanced stage and deep myometrial invasion, respectively. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Kido and Nishio in this issue.