Myo-regressor Deep Informed Neural NetwOrk (Myo-DINO) for fast MR parameters mapping in neuromuscular disorders.

Magnetic Resonance (MR) parameters mapping in muscle Magnetic Resonance Imaging (mMRI) is predominantly performed using pattern recognition-based algorithms, which are characterised by high computational costs and scalability issues in the context of multi-parametric mapping. Deep Learning (DL) has been demonstrated to be a robust and efficient method for rapid MR parameters mapping. However, its application in mMRI domain to investigate Neuromuscular Disorders (NMDs) has not yet been explored. In addition, data-driven DL models suffered in interpretation and explainability of the learning process. We developed a Physics Informed Neural Network called Myo-Regressor Deep Informed Neural NetwOrk (Myo-DINO) for efficient and explainable Fat Fraction (FF), water-T2 (wT2) and B1 mapping from a cohort of NMDs.A total of 2165 slices (232 subjects) from Multi-Echo Spin Echo (MESE) images were selected as the input dataset for which FF, wT2,B1 ground truth maps were computed using the MyoQMRI toolbox. This toolbox exploits the Extended Phase Graph (EPG) theory with a two-component model (water and fat signal) and slice profile to simulate the signal evolution in the MESE framework. A customized U-Net architecture was implemented as the Myo-DINO architecture. The squared L2 norm loss was complemented by two distinct physics models to define two 'Physics-Informed' loss functions: Cycling Loss 1 embedded a mono-exponential model to describe the relaxation of water protons, while Cycling Loss 2 incorporated the EPG theory with slice profile to model the magnetization dephasing under the effect of gradients and RF pulses. The Myo-DINO was trained with the hyperparameter value of the 'Physics-Informed' component held constant, i.e. λmodel = 1, while different hyperparameter values (λcnn) were applied to the squared L2 norm component in both the cycling loss. In particular, hard (λcnn=10), normal (λcnn=1) and self-supervised (λcnn=0) constraints were applied to gradually decrease the impact of the squared L2 norm component on the 'Physics Informed' term during the Myo-DINO training process. Myo-DINO achieved higher performance with Cycling Loss 2 for FF, wT2 and B1 prediction. In particular, high reconstruction similarity and quality (Structural Similarity Index > 0.92, Peak Signal to Noise ratio > 30.0 db) and small reconstruction error (Normalized Root Mean Squared Error < 0.038) to the reference maps were shown with self-supervised weighting of the Cycling Loss 2. In addition muscle-wise FF, wT2 and B1 predicted values showed good agreement with the reference values. The Myo-DINO has been demonstrated to be a robust and efficient workflow for MR parameters mapping in the context of mMRI. This provides preliminary evidence that it can be an effective alternative to the reference post-processing algorithm. In addition, our results demonstrate that Cycling Loss 2, which incorporates the Extended Phase Graph (EPG) model, provides the most robust and relevant physical constraints for Myo-DINO in this multi-parameter regression task. The use of Cycling Loss 2 with self-supervised constraint improved the explainability of the learning process because the network acquired domain knowledge solely in accordance with the assumptions of the EPG model.

CT and MRI radiomic features of lung cancer (NSCLC): comparison and software consistency.

BACKGROUND: Radiomics is a quantitative approach that allows the extraction of mineable data from medical images. Despite the growing clinical interest, radiomics studies are affected by variability stemming from analysis choices. We aimed to investigate the agreement between two open-source radiomics software for both contrast-enhanced computed tomography (CT) and contrast-enhanced magnetic resonance imaging (MRI) of lung cancers and to preliminarily evaluate the existence of radiomic features stable for both techniques. METHODS: Contrast-enhanced CT and MRI images of 35 patients affected with non-small cell lung cancer (NSCLC) were manually segmented and preprocessed using three different methods. Sixty-six Image Biomarker Standardisation Initiative-compliant features common to the considered platforms, PyRadiomics and LIFEx, were extracted. The correlation among features with the same mathematical definition was analyzed by comparing PyRadiomics and LIFEx (at fixed imaging technique), and MRI with CT results (for the same software). RESULTS: When assessing the agreement between LIFEx and PyRadiomics across the considered resampling, the maximum statistically significant correlations were observed to be 94% for CT features and 95% for MRI ones. When examining the correlation between features extracted from contrast-enhanced CT and MRI using the same software, higher significant correspondences were identified in 11% of features for both software. CONCLUSIONS: Considering NSCLC, (i) for both imaging techniques, LIFEx and PyRadiomics agreed on average for 90% of features, with MRI being more affected by resampling and (ii) CT and MRI contained mostly non-redundant information, but there are shape features and, more importantly, texture features that can be singled out by both techniques. RELEVANCE STATEMENT: Identifying and selecting features that are stable cross-modalities may be one of the strategies to pave the way for radiomics clinical translation. KEY POINTS: • More than 90% of LIFEx and PyRadiomics features contain the same information. • Ten percent of features (shape, texture) are stable among contrast-enhanced CT and MRI. • Software compliance and cross-modalities stability features are impacted by the resampling method.

Role of fronto-limbic circuit in neuropsychiatric symptoms of dementia: clinical evidence from an exploratory study.

Background: Neuropsychiatric symptoms (NPSs) are a distressful aspect of dementia and the knowledge of structural correlates of NPSs is limited. We aimed to identify associations of fronto-limbic circuit with specific NPSs in patients with various types of cognitive impairment. Methods: Of 84 participants, 27 were diagnosed with mild cognitive impairment (MCI), 41 with Alzheimer's disease (AD) dementia and 16 with non-AD dementia. In all patients we assessed regional brain morphometry using a region of interest (ROI)-based analysis. The mean cortical thickness (CT) of 20 cortical regions and the volume (V) of 4 subcortical areas of the fronto-limbic system were extracted. NPSs were rated with the Neuropsychiatric Inventory (NPI). We used multiple linear regression models adjusted for age and disease duration to identify significant associations between scores of NPI sub-domains and MRI measures of brain morphometry. Results: All significant associations found were negative, except those between irritability and the fronto-opercular regions in MCI patients (corresponding to a 40-50% increase in CT) and between delusions and hippocampus and anterior cingulate gyrus (with a 40-60% increase). Apathy showed predominant involvement of the inferior frontal regions in AD group (a 30% decrease in CT) and of the cingulate cortex in non-AD group (a 50-60% decrease in CT). Anxiety correlated in MCI patients with the cingulate gyrus and caudate, with a CT and V decrease of about 40%, while hallucinations were associated with left enthorinal gyrus and right amygdala and temporal pole. Agitation showed associations in the AD group with the frontal regions and the temporal pole, corresponding to a 30-40% decrease in CT. Euphoria, disinhibition and eating abnormalities were associated in the MCI group with the entorhinal, para-hippocampal and fusiform gyri, the temporal pole and the amygdala (with a 40-70% decrease in CT and V). Finally, aberrant motor behavior reported a significant association with frontal and cingulate regions with a 50% decrease in CT. Conclusion: Our findings indicate that specific NPSs are associated with the structural involvement of the fronto-limbic circuit across different types of neurocognitive disorders. Factors, such as age and disease duration, can partly account for the variability of the associations observed.

Fast deep learning reconstruction techniques for preclinical magnetic resonance fingerprinting.

We propose a deep learning (DL) model and a hyperparameter optimization strategy to reconstruct T1 and T2 maps acquired with the magnetic resonance fingerprinting (MRF) methodology. We applied two different MRF sequence routines to acquire images of ex vivo rat brain phantoms using a 7-T preclinical scanner. Subsequently, the DL model was trained using experimental data, completely excluding the use of any theoretical MRI signal simulator. The best combination of the DL parameters was implemented by an automatic hyperparameter optimization strategy, whose key aspect is to include all the parameters to the fit, allowing the simultaneous optimization of the neural network architecture, the structure of the DL model, and the supervised learning algorithm. By comparing the reconstruction performances of the DL technique with those achieved from the traditional dictionary-based method on an independent dataset, the DL approach was shown to reduce the mean percentage relative error by a factor of 3 for T1 and by a factor of 2 for T2 , and to improve the computational time by at least a factor of 37. Furthermore, the proposed DL method enables maintaining comparable reconstruction performance, even with a lower number of MRF images and a reduced k-space sampling percentage, with respect to the dictionary-based method. Our results suggest that the proposed DL methodology may offer an improvement in reconstruction accuracy, as well as speeding up MRF for preclinical, and in prospective clinical, investigations.

Quantification of pulmonary involvement in COVID-19 pneumonia: an upgrade of the LungQuant software for lung CT segmentation.

Computed tomography (CT) scans are used to evaluate the severity of lung involvement in patients affected by COVID-19 pneumonia. Here, we present an improved version of the LungQuant automatic segmentation software (LungQuant v2), which implements a cascade of three deep neural networks (DNNs) to segment the lungs and the lung lesions associated with COVID-19 pneumonia. The first network (BB-net) defines a bounding box enclosing the lungs, the second one (U-net 1 ) outputs the mask of the lungs, and the final one (U-net 2 ) generates the mask of the COVID-19 lesions. With respect to the previous version (LungQuant v1), three main improvements are introduced: the BB-net, a new term in the loss function in the U-net for lesion segmentation and a post-processing procedure to separate the right and left lungs. The three DNNs were optimized, trained and tested on publicly available CT scans. We evaluated the system segmentation capability on an independent test set consisting of ten fully annotated CT scans, the COVID-19-CT-Seg benchmark dataset. The test performances are reported by means of the volumetric dice similarity coefficient (vDSC) and the surface dice similarity coefficient (sDSC) between the reference and the segmented objects. LungQuant v2 achieves a vDSC (sDSC) equal to 0.96 ± 0.01 (0.97 ± 0.01) and 0.69 ± 0.08 (0.83 ± 0.07) for the lung and lesion segmentations, respectively. The output of the segmentation software was then used to assess the percentage of infected lungs, obtaining a Mean Absolute Error (MAE) equal to 2%.

A multicenter evaluation of a deep learning software (LungQuant) for lung parenchyma characterization in COVID-19 pneumonia.

BACKGROUND: The role of computed tomography (CT) in the diagnosis and characterization of coronavirus disease 2019 (COVID-19) pneumonia has been widely recognized. We evaluated the performance of a software for quantitative analysis of chest CT, the LungQuant system, by comparing its results with independent visual evaluations by a group of 14 clinical experts. The aim of this work is to evaluate the ability of the automated tool to extract quantitative information from lung CT, relevant for the design of a diagnosis support model. METHODS: LungQuant segments both the lungs and lesions associated with COVID-19 pneumonia (ground-glass opacities and consolidations) and computes derived quantities corresponding to qualitative characteristics used to clinically assess COVID-19 lesions. The comparison was carried out on 120 publicly available CT scans of patients affected by COVID-19 pneumonia. Scans were scored for four qualitative metrics: percentage of lung involvement, type of lesion, and two disease distribution scores. We evaluated the agreement between the LungQuant output and the visual assessments through receiver operating characteristics area under the curve (AUC) analysis and by fitting a nonlinear regression model. RESULTS: Despite the rather large heterogeneity in the qualitative labels assigned by the clinical experts for each metric, we found good agreement on the metrics compared to the LungQuant output. The AUC values obtained for the four qualitative metrics were 0.98, 0.85, 0.90, and 0.81. CONCLUSIONS: Visual clinical evaluation could be complemented and supported by computer-aided quantification, whose values match the average evaluation of several independent clinical experts. KEY POINTS: We conducted a multicenter evaluation of the deep learning-based LungQuant automated software. We translated qualitative assessments into quantifiable metrics to characterize coronavirus disease 2019 (COVID-19) pneumonia lesions. Comparing the software output to the clinical evaluations, results were satisfactory despite heterogeneity of the clinical evaluations. An automatic quantification tool may contribute to improve the clinical workflow of COVID-19 pneumonia.

Preliminary report on harmonization of features extraction process using the ComBat tool in the multi-center "Blue Sky Radiomics" study on stage III unresectable NSCLC.

BACKGROUND AND PURPOSE: In the retrospective-prospective multi-center "Blue Sky Radiomics" study (NCT04364776), we plan to test a pre-defined radiomic signature in a series of stage III unresectable NSCLC patients undergoing chemoradiotherapy and maintenance immunotherapy. As a necessary preliminary step, we explore the influence of different image-acquisition parameters on radiomic features' reproducibility and apply methods for harmonization. MATERIAL AND METHODS: We identified the primary lung tumor on two computed tomography (CT) series for each patient, acquired before and after chemoradiation with i.v. contrast medium and with different scanners. Tumor segmentation was performed by two oncological imaging specialists (thoracic radiologist and radio-oncologist) using the Oncentra Masterplan® software. We extracted 42 radiomic features from the specific ROIs (LIFEx). To assess the impact of different acquisition parameters on features extraction, we used the Combat tool with nonparametric adjustment and the longitudinal version (LongComBat). RESULTS: We defined 14 CT acquisition protocols for the harmonization process. Before harmonization, 76% of the features were significantly influenced by these protocols. After, all extracted features resulted in being independent of the acquisition parameters. In contrast, 5% of the LongComBat harmonized features still depended on acquisition protocols. CONCLUSIONS: We reduced the impact of different CT acquisition protocols on radiomic features extraction in a group of patients enrolled in a radiomic study on stage III NSCLC. The harmonization process appears essential for the quality of radiomic data and for their reproducibility. ClinicalTrials.gov Identifier: NCT04364776, First Posted:April 28, 2020, Actual Study Start Date: April 15, 2020, https://clinicaltrials.gov/ct2/show/NCT04364776 .

Quantification of pulmonary involvement in COVID-19 pneumonia by means of a cascade of two U-nets: training and assessment on multiple datasets using different annotation criteria.

PURPOSE: This study aims at exploiting artificial intelligence (AI) for the identification, segmentation and quantification of COVID-19 pulmonary lesions. The limited data availability and the annotation quality are relevant factors in training AI-methods. We investigated the effects of using multiple datasets, heterogeneously populated and annotated according to different criteria. METHODS: We developed an automated analysis pipeline, the LungQuant system, based on a cascade of two U-nets. The first one (U-net[Formula: see text]) is devoted to the identification of the lung parenchyma; the second one (U-net[Formula: see text]) acts on a bounding box enclosing the segmented lungs to identify the areas affected by COVID-19 lesions. Different public datasets were used to train the U-nets and to evaluate their segmentation performances, which have been quantified in terms of the Dice Similarity Coefficients. The accuracy in predicting the CT-Severity Score (CT-SS) of the LungQuant system has been also evaluated. RESULTS: Both the volumetric DSC (vDSC) and the accuracy showed a dependency on the annotation quality of the released data samples. On an independent dataset (COVID-19-CT-Seg), both the vDSC and the surface DSC (sDSC) were measured between the masks predicted by LungQuant system and the reference ones. The vDSC (sDSC) values of 0.95±0.01 and 0.66±0.13 (0.95±0.02 and 0.76±0.18, with 5 mm tolerance) were obtained for the segmentation of lungs and COVID-19 lesions, respectively. The system achieved an accuracy of 90% in CT-SS identification on this benchmark dataset. CONCLUSION: We analysed the impact of using data samples with different annotation criteria in training an AI-based quantification system for pulmonary involvement in COVID-19 pneumonia. In terms of vDSC measures, the U-net segmentation strongly depends on the quality of the lesion annotations. Nevertheless, the CT-SS can be accurately predicted on independent test sets, demonstrating the satisfactory generalization ability of the LungQuant.

Diagnostic Yield and Cost-Effectiveness of "Dynamic" Exome Analysis in Epilepsy with Neurodevelopmental Disorders: A Tertiary-Center Experience in Northern Italy.

BACKGROUND: The advent of next-generation sequencing (NGS) techniques in clinical practice led to a significant advance in gene discovery. We aimed to describe diagnostic yields of a "dynamic" exome-based approach in a cohort of patients with epilepsy associated with neurodevelopmental disorders. METHODS: We conducted a retrospective, observational study on 72 probands. All patients underwent a first diagnostic level of a 135 gene panel, a second of 297 genes for inconclusive cases, and finally, a whole-exome sequencing for negative cases. Diagnostic yields at each step and cost-effectiveness were the objects of statistical analysis. RESULTS: Overall diagnostic yield in our cohort was 37.5%: 29% of diagnoses derived from the first step analysis, 5.5% from the second step, and 3% from the third. A significant difference emerged between the three diagnostic steps (p < 0.01), between the first and second (p = 0.001), and the first and third (p << 0.001). The cost-effectiveness plane indicated that our exome-based "dynamic" approach was better in terms of cost savings and higher diagnostic rate. CONCLUSIONS: Our findings suggested that "dynamic" NGS techniques applied to well-phenotyped individuals can save both time and resources. In patients with unexplained epilepsy comorbid with NDDs, our approach might maximize the number of diagnoses achieved.