Evolving Characteristics of Gadolinium-Based Contrast Agents for MR Imaging: A Systematic Review of the Importance of Relaxivity.

Gadolinium-based contrast agents (GBCAs) are widely and routinely used to enhance the diagnostic performance of magnetic resonance imaging and magnetic resonance angiography examinations. T1 relaxivity (r1) is the measure of their ability to increase signal intensity in tissues and blood on T1-weighted images at a given dose. Pharmaceutical companies have invested in the design and development of GBCAs with higher and higher T1 relaxivity values, and "high relaxivity" is a claim frequently used to promote GBCAs, with no clear definition of what "high relaxivity" means, or general concurrence about its clinical benefit. To understand whether higher relaxivity values translate into a material clinical benefit, well-designed, and properly powered clinical studies are necessary, while mere in vitro measurements may be misleading. This systematic review of relevant peer-reviewed literature provides high-quality clinical evidence showing that a difference in relaxivity of at least 40% between two GBCAs results in superior diagnostic efficacy for the higher-relaxivity agent when this is used at the same equimolar gadolinium dose as the lower-relaxivity agent, or similar imaging performance when used at a lower dose. Either outcome clearly implies a relevant clinical benefit. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 3.

Development and validation of an image biomarker to identify metabolic dysfunction associated steatohepatitis: MR-MASH score.

BACKGROUND AND AIMS: Diagnosis of metabolic dysfunction-associated steatohepatitis (MASH) requires histology. In this study, a magnetic resonance imaging (MRI) score was developed and validated to identify MASH in patients with metabolic dysfunction-associated steatotic liver disease (MASLD). Secondarily, a screening strategy for MASH diagnosis was investigated. METHODS: This prospective multicentre study included 317 patients with biopsy-proven MASLD and contemporaneous MRI. The discovery cohort (Spain, Portugal) included 194 patients. NAFLD activity score (NAS) and fibrosis were assessed with the NASH-CRN histologic system. MASH was defined by the presence of steatosis, lobular inflammation, and ballooning, with NAS ≥4 with or without fibrosis. An MRI-based composite biomarker of Proton Density Fat Fraction and waist circumference (MR-MASH score) was developed. Findings were afterwards validated in an independent cohort (United States, Spain) with different MRI protocols. RESULTS: In the derivation cohort, 51% (n = 99) had MASH. The MR-MASH score identified MASH with an AUC = .88 (95% CI .83-.93) and strongly correlated with NAS (r = .69). The MRI score lower cut-off corresponded to 88% sensitivity with 86% NPV, while the upper cut-off corresponded to 92% specificity with 87% PPV. MR-MASH was validated with an AUC = .86 (95% CI .77-.92), 91% sensitivity (lower cut-off) and 87% specificity (upper cut-off). A two-step screening strategy with sequential MR-MASH examination performed in patients with indeterminate-high FIB-4 or transient elastography showed an 83-84% PPV to identify MASH. The AUC of MR-MASH was significantly higher than that of the FAST score (p < .001). CONCLUSIONS: The MR-MASH score has clinical utility in the identification and management of patients with MASH at risk of progression.

Development and evaluation of two open-source nnU-Net models for automatic segmentation of lung tumors on PET and CT images with and without respiratory motion compensation.

OBJECTIVES: In lung cancer, one of the main limitations for the optimal integration of the biological and anatomical information derived from Positron Emission Tomography (PET) and Computed Tomography (CT) is the time and expertise required for the evaluation of the different respiratory phases. In this study, we present two open-source models able to automatically segment lung tumors on PET and CT, with and without motion compensation. MATERIALS AND METHODS: This study involved time-bin gated (4D) and non-gated (3D) PET/CT images from two prospective lung cancer cohorts (Trials 108237 and 108472) and one retrospective. For model construction, the ground truth (GT) was defined by consensus of two experts, and the nnU-Net with 5-fold cross-validation was applied to 560 4D-images for PET and 100 3D-images for CT. The test sets included 270 4D- images and 19 3D-images for PET and 80 4D-images and 27 3D-images for CT, recruited at 10 different centres. RESULTS: In the performance evaluation with the multicentre test sets, the Dice Similarity Coefficients (DSC) obtained for our PET model were DSC(4D-PET) = 0.74 ± 0.06, improving 19% relative to the DSC between experts and DSC(3D-PET) = 0.82 ± 0.11. The performance for CT was DSC(4D-CT) = 0.61 ± 0.28 and DSC(3D-CT) = 0.63 ± 0.34, improving 4% and 15% relative to DSC between experts. CONCLUSIONS: Performance evaluation demonstrated that the automatic segmentation models have the potential to achieve accuracy comparable to manual segmentation and thus hold promise for clinical application. The resulting models can be freely downloaded and employed to support the integration of 3D- or 4D- PET/CT and to facilitate the evaluation of its impact on lung cancer clinical practice. CLINICAL RELEVANCE STATEMENT: We provide two open-source nnU-Net models for the automatic segmentation of lung tumors on PET/CT to facilitate the optimal integration of biological and anatomical information in clinical practice. The models have superior performance compared to the variability observed in manual segmentations by the different experts for images with and without motion compensation, allowing to take advantage in the clinical practice of the more accurate and robust 4D-quantification. KEY POINTS: Lung tumor segmentation on PET/CT imaging is limited by respiratory motion and manual delineation is time consuming and suffer from inter- and intra-variability. Our segmentation models had superior performance compared to the manual segmentations by different experts. Automating PET image segmentation allows for easier clinical implementation of biological information.

Imaging biomarkers and radiomics in pediatric oncology: a view from the PRIMAGE (PRedictive In silico Multiscale Analytics to support cancer personalized diaGnosis and prognosis, Empowered by imaging biomarkers) project.

This review paper presents the practical development of imaging biomarkers in the scope of the PRIMAGE (PRedictive In silico Multiscale Analytics to support cancer personalized diaGnosis and prognosis, Empowered by imaging biomarkers) project, as a noninvasive and reliable way to improve the diagnosis and prognosis in pediatric oncology. The PRIMAGE project is a European multi-center research initiative that focuses on developing medical imaging-derived artificial intelligence (AI) solutions designed to enhance overall management and decision-making for two types of pediatric cancer: neuroblastoma and diffuse intrinsic pontine glioma. To allow this, the PRIMAGE project has created an open-cloud platform that combines imaging, clinical, and molecular data together with AI models developed from this data, creating a comprehensive decision support environment for clinicians managing patients with these two cancers. In order to achieve this, a standardized data processing and analysis workflow was implemented to generate robust and reliable predictions for different clinical endpoints. Magnetic resonance (MR) image harmonization and registration was performed as part of the workflow. Subsequently, an automated tool for the detection and segmentation of tumors was trained and internally validated. The Dice similarity coefficient obtained for the independent validation dataset was 0.997, indicating compatibility with the manual segmentation variability. Following this, radiomics and deep features were extracted and correlated with clinical endpoints. Finally, reproducible and relevant imaging quantitative features were integrated with clinical and molecular data to enrich both the predictive models and a set of visual analytics tools, making the PRIMAGE platform a complete clinical decision aid system. In order to ensure the advancement of research in this field and to foster engagement with the wider research community, the PRIMAGE data repository and platform are currently being integrated into the European Federation for Cancer Images (EUCAIM), which is the largest European cancer imaging research infrastructure created to date.

A Gadolinium(III) Complex Based on Pyridoxine Molecule with Single-Ion Magnet and Magnetic Resonance Imaging Properties.

Pyridoxine (pyr) is a versatile molecule that forms part of the family of B vitamins. It is used to treat and prevent vitamin B6 deficiency and certain types of metabolic disorders. Moreover, the pyridoxine molecule has been investigated as a suitable ligand toward metal ions. Nevertheless, the study of the magnetic properties of metal complexes containing lanthanide(III) ions and this biomolecule is unexplored. We have synthesized and characterized a novel pyridoxine-based GdIII complex of formula [GdIII(pyr)2(H2O)4]Cl3 · 2 H2O (1) [pyr = pyridoxine]. 1 crystallizes in the triclinic system and space group Pi. In its crystal packing, cationic [Gd(pyr)2(H2O)4]3+ entities are connected through H-bonding interactions involving non-coordinating water molecules and chloride anions. In addition, Hirshfeld surfaces of 1 were calculated to further investigate their intermolecular interactions in the crystal lattice. Our investigation of the magnetic properties of 1, through ac magnetic susceptibility measurements, reveals the occurrence of a slow relaxation in magnetization in this mononuclear GdIII complex, indicating an unusual single-ion magnet (SIM) behavior for this pseudo-isotropic metal ion at very low temperatures. We also studied the relaxometric properties of 1, as a potential contrast agent for high-field magnetic resonance imaging (MRI), from solutions of 1 prepared in physiological serum (0.0-3.2 mM range) and measured at 3 T on a clinical MRI scanner. The values of relaxivity obtained for 1 are larger than those of some commercial MRI contrast agents based on mononuclear GdIII systems.

Increased femoral anteversion in females with anterior knee pain relates to both the neck and the shaft of the femur.

INTRODUCTION: Increased femoral anteversion (FAV) can have many clinical manifestations, including anterior knee pain (AKP). To our knowledge, no studies have measured the location of FAV in a cohort of female AKP patients. The objective of this research is to determine whether the increased FAV in AKP females originates above the lesser trochanter, below the lesser trochanter or at both levels. MATERIALS AND METHODS: Thrity-seven consecutive AKP female patients (n = 66 femurs) were recruited prospectively. There were 17 patients (n = 26 femurs; mean age of 28 years) in whom the suspicion for the increased FAV of the femur was based on the clinical examination (pathological group-PG). The control group (CG) consisted of 20 patients (n = 40 femurs; mean age of 29 years) in whom there was no increased FAV from the clinical standpoint. All of them underwent a torsional computed tomography of the lower limbs. FAV was measured according to Murphy´s method. A segmental analysis of FAV was performed using the lesser trochanter as a landmark. RESULTS: Significant differences in the total FAV (18.7 ± 5.52 vs. 42.46 ± 6.33; p < 0.001), the neck version (54.88 ± 9.64 vs. 64.27 ± 11.25; p = 0.0006) and the diaphysis version (- 36.17 ± 8.93 vs. - 21.81 ± 11.73; p < 0.001) were observed between the CG and the PG. The difference in the diaphyseal angle between CG and PG accounts for 60% of the total difference between healthy and pathological groups, while the difference between both groups in the angle of the neck accounts for 40%. CONCLUSION: In chronic AKP female patients with increased FAV, the two segments of the femur contribute to the total FAV, with a different pattern among patients and controls, being the compensation mechanism of the diaphysis much lower in the pathological femurs than in the controls.

Quantification of Liver Iron Overload with MRI: Review and Guidelines from the ESGAR and SAR.

Accumulation of excess iron in the body, or systemic iron overload, results from a variety of causes. The concentration of iron in the liver is linearly related to the total body iron stores and, for this reason, quantification of liver iron concentration (LIC) is widely regarded as the best surrogate to assess total body iron. Historically assessed using biopsy, there is a clear need for noninvasive quantitative imaging biomarkers of LIC. MRI is highly sensitive to the presence of tissue iron and has been increasingly adopted as a noninvasive alternative to biopsy for detection, severity grading, and treatment monitoring in patients with known or suspected iron overload. Multiple MRI strategies have been developed in the past 2 decades, based on both gradient-echo and spin-echo imaging, including signal intensity ratio and relaxometry strategies. However, there is a general lack of consensus regarding the appropriate use of these methods. The overall goal of this article is to summarize the current state of the art in the clinical use of MRI to quantify liver iron content and to assess the overall level of evidence of these various methods. Based on this summary, expert consensus panel recommendations on best practices for MRI-based quantification of liver iron are provided.

Automated prostate multi-regional segmentation in magnetic resonance using fully convolutional neural networks.

OBJECTIVE: Automatic MR imaging segmentation of the prostate provides relevant clinical benefits for prostate cancer evaluation such as calculation of automated PSA density and other critical imaging biomarkers. Further, automated T2-weighted image segmentation of central-transition zone (CZ-TZ), peripheral zone (PZ), and seminal vesicle (SV) can help to evaluate clinically significant cancer following the PI-RADS v2.1 guidelines. Therefore, the main objective of this work was to develop a robust and reproducible CNN-based automatic prostate multi-regional segmentation model using an intercontinental cohort of prostate MRI. METHODS: A heterogeneous database of 243 T2-weighted prostate studies from 7 countries and 10 machines of 3 different vendors, with the CZ-TZ, PZ, and SV regions manually delineated by two experienced radiologists (ground truth), was used to train (n = 123) and test (n = 120) a U-Net-based model with deep supervision using a cyclical learning rate. The performance of the model was evaluated by means of dice similarity coefficient (DSC), among others. Segmentation results with a DSC above 0.7 were considered accurate. RESULTS: The proposed method obtained a DSC of 0.88 ± 0.01, 0.85 ± 0.02, 0.72 ± 0.02, and 0.72 ± 0.02 for the prostate gland, CZ-TZ, PZ, and SV respectively in the 120 studies of the test set when comparing the predicted segmentations with the ground truth. No statistically significant differences were found in the results obtained between manufacturers or continents. CONCLUSION: Prostate multi-regional T2-weighted MR images automatic segmentation can be accurately achieved by U-Net like CNN, generalizable in a highly variable clinical environment with different equipment, acquisition configurations, and population. KEY POINTS: • Deep learning techniques allows the accurate segmentation of the prostate in three different regions on MR T2w images. • Multi-centric database proved the generalization of the CNN model on different institutions across different continents. • CNN models can be used to aid on the diagnosis and follow-up of patients with prostate cancer.

Predictive value of texture analysis on lumbar MRI in patients with chronic low back pain.

PURPOSE: The aim of this study was to determine whether MRI texture analysis could predict the prognosis of patients with non-specific chronic low back pain. METHODS: A prospective observational study was conducted on 100 patients with non-specific chronic low back pain, who underwent a conventional MRI, followed by rehabilitation treatment, and revisited after 6 months. Sociodemographic variables, numeric pain scale (NPS) value, and the degree of disability as measured by the Roland-Morris disability questionnaire (RMDQ), were collected. The MRI analysis included segmentation of regions of interest (vertebral endplates and intervertebral disks from L3-L4 to L5-S1, paravertebral musculature at the L4-L5 space) to extract texture variables (PyRadiomics software). The classification random forest algorithm was applied to identify individuals who would improve less than 30% in the NPS or would score more than 4 in the RMDQ at the end of the follow-up. Sensitivity, specificity, and the area under the ROC curve were calculated. RESULTS: The final series included 94 patients. The predictive model for classifying patients whose pain did not improve by 30% or more offered a sensitivity of 0.86, specificity 0.57, and area under the ROC curve 0.71. The predictive model for classifying patients with a RMDQ score 4 or more offered a sensitivity of 0.83, specificity of 0.20, and area under the ROC curve of 0.52. CONCLUSION: The texture analysis of lumbar MRI could help identify patients who are more likely to improve their non-specific chronic low back pain through rehabilitation programs, allowing a personalized therapeutic plan to be established.

Feasibility and outcomes of transjugular intrahepatic portosystemic shunts in infants.

BACKGROUND: Experience with transjugular intrahepatic portosystemic shunts (TIPS) in the pediatric population, especially in infants, is limited. OBJECTIVE: To evaluate the feasibility, efficacy and safety of TIPS placement in infants. MATERIALS AND METHODS: This retrospective non-comparative observational cohort study analyzed all pediatric patients < 12 months of age treated with TIPS while waiting for liver transplant between October 2018 and April 2021. The sample consisted of 10 infants with chronic liver disease. All had refractory ascites and decreased portal vein size. Their mean age ± standard deviation was 5 ± 1 months and their mean weight was 5.4 ± 1.0 kg. We calculated the pediatric end-stage liver disease score and portosystemic gradients before and after TIPS placement. We used ultrasound to check for complications and to assess the presence of ascites. We used paired-sample t-test for the mean comparison of paired variables. RESULTS: Ten TIPS procedures were performed that were technically and hemodynamically successful except for one, in which an extrahepatic portal puncture required surgical repair. Ascites resolved in three infants and was reduced in six. The portal vein size remained stable after TIPS placement. Four infants had early stent thrombosis and two had late stent thrombosis treated with angioplasty or covered stents. CONCLUSION: TIPS placement in infants is a feasible, safe and effective procedure.

Objective Image Quality Comparison Between Brain-Dedicated PET and PET/CT Scanners.

As part of a clinical validation of a new brain-dedicated PET system (CMB), image quality of this scanner has been compared to that of a whole-body PET/CT scanner. To that goal, Hoffman phantom and patient data were obtined with both devices. Since CMB does not use a CT for attenuation correction (AC) which is crucial for PET images quality, this study includes the evaluation of CMB PET images using emission-based or CT-based attenuation maps. PET images were compared using 34 image quality metrics. Moreover, a neural network was used to evaluate the degree of agreement between both devices on the patients diagnosis prediction. Overall, results showed that CMB images have higher contrast and recovery coefficient but higher noise than PET/CT images. Although SUVr values presented statistically significant differences in many brain regions, relative differences were low. An asymmetry between left and right hemispheres, however, was identified. Even so, the variations between the two devices were minor. Finally, there is a greater similarity between PET/CT and CMB CT-based AC PET images than between PET/CT and the CMB emission-based AC PET images.

Automated Whole-Liver MRI Segmentation to Assess Steatosis and Iron Quantification in Chronic Liver Disease.

Background Standardized manual region of interest (ROI) sampling strategies for hepatic MRI steatosis and iron quantification are time consuming, with variable results. Purpose To evaluate the performance of automatic MRI whole-liver segmentation (WLS) for proton density fat fraction (PDFF) and iron estimation (transverse relaxometry [R2*]) versus manual ROI, with liver biopsy as the reference standard. Materials and Methods This prospective, cross-sectional, multicenter study recruited participants with chronic liver disease who underwent liver biopsy and chemical shift-encoded 3.0-T MRI between January 2017 and January 2021. Biopsy evaluation included histologic grading and digital pathology. MRI liver sampling strategies included manual ROI (two observers) and automatic whole-liver (deep learning algorithm) segmentation for PDFF- and R2*-derived measurements. Agreements between segmentation methods were measured using intraclass correlation coefficients (ICCs), and biases were evaluated using Bland-Altman analyses. Linear regression analyses were performed to determine the correlation between measurements and digital pathology. Results A total of 165 participants were included (mean age ± standard deviation, 55 years ± 12; 96 women; 101 of 165 participants [61%] with nonalcoholic fatty liver disease). Agreements between mean measurements were excellent, with ICCs of 0.98 for both PDFF and R2*. The median bias was 0.5% (interquartile range, -0.4% to 1.2%) for PDFF and 2.7 sec-1 (interquartile range, 0.2-5.3 sec-1) for R2* (P < .001 for both). Margins of error were lower for WLS than ROI-derived parameters (-0.03% for PDFF and -0.3 sec-1 for R2*). ROI and WLS showed similar performance for steatosis (ROI AUC, 0.96; WLS AUC, 0.97; P = .53) and iron overload (ROI AUC, 0.85; WLS AUC, 0.83; P = .09). Correlations with digital pathology were high (P < .001) between the fat ratio and PDFF (ROI r = 0.89; WLS r = 0.90) and moderate (P < .001) between the iron ratio and R2* (ROI r = 0.65; WLS r = 0.64). Conclusion Proton density fat fraction and transverse relaxometry measurements derived from MRI automatic whole-liver segmentation (WLS) were accurate for steatosis and iron grading in chronic liver disease and correlated with digital pathology. Automated WLS estimations were higher, with a lower margin of error than manual region of interest estimations. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Moura Cunha and Fowler in this issue.

Bridging gaps between images and data: a systematic update on imaging biobanks.

BACKGROUND AND OBJECTIVE: The systematic collection of medical images combined with imaging biomarkers and patient non-imaging data is the core concept of imaging biobanks, a key element for fuelling the development of modern precision medicine. Our purpose is to review the existing image repositories fulfilling the criteria for imaging biobanks. METHODS: Pubmed, Scopus and Web of Science were searched for articles published in English from January 2010 to July 2021 using a combination of the terms: "imaging" AND "biobanks" and "imaging" AND "repository". Moreover, the Community Research and Development Information Service (CORDIS) database ( https://cordis.europa.eu/projects ) was searched using the terms: "imaging" AND "biobanks", also including collections, projects, project deliverables, project publications and programmes. RESULTS: Of 9272 articles retrieved, only 54 related to biobanks containing imaging data were finally selected, of which 33 were disease-oriented (61.1%) and 21 population-based (38.9%). Most imaging biobanks were European (26/54, 48.1%), followed by American biobanks (20/54, 37.0%). Among disease-oriented biobanks, the majority were focused on neurodegenerative (9/33, 27.3%) and oncological diseases (9/33, 27.3%). The number of patients enrolled ranged from 240 to 3,370,929, and the imaging modality most frequently involved was MRI (40/54, 74.1%), followed by CT (20/54, 37.0%), PET (13/54, 24.1%), and ultrasound (12/54, 22.2%). Most biobanks (38/54, 70.4%) were accessible under request. CONCLUSIONS: Imaging biobanks can serve as a platform for collecting, sharing and analysing medical images integrated with imaging biomarkers, biological and clinical data. A relatively small proportion of current biobanks also contain images and can thus be classified as imaging biobanks. KEY POINTS: • Imaging biobanks are a powerful tool for large-scale collection and processing of medical images integrated with imaging biomarkers and patient non-imaging data. • Most imaging biobanks retrieved were European, disease-oriented and accessible under user request. • While many biobanks have been developed so far, only a relatively small proportion of them are imaging biobanks.

Virtual conferences: results of an international survey on radiologist preferences and perspectives.

BACKGROUND: We explored perceptions and preferences regarding the conversion of in-person to virtual conferences as necessitated by travel and in-person meeting restrictions. METHODS: A 16-question online survey to assess preferences regarding virtual conferences during the COVID-19 pandemic and future perspectives on this subject was disseminated internationally online between June and August 2020. FINDINGS: A total of 508 responses were received from 73 countries. The largest number of responses came from Italy and the USA. The majority of respondents had already attended a virtual conference (80%) and would like to attend future virtual meetings (97%). The ideal duration of such an event was 2-3 days (42%). The preferred time format was a 2-4-h session (43%). Most respondents also noted that they would like a significant fee reduction and the possibility to attend a conference partly in-person and partly online. Respondents indicated educational sessions as the most valuable sections of virtual meetings. The reported positive factor of the virtual meeting format is the ability to re-watch lectures on demand. On the other hand, the absence of networking and human contact was recognized as a significant loss. In the future, people expressed a preference to attend conferences in person for networking purposes, but only in safer conditions. CONCLUSIONS: Respondents appreciated the opportunity to attend the main radiological congresses online and found it a good opportunity to stay updated without having to travel. However, in general, they would prefer these conferences to be structured differently. The lack of networking opportunities was the main reason for preferring an in-person meeting. KEY POINTS: • Respondents appreciated the opportunity to attend the main radiological meetings online, considering it a good opportunity to stay updated without having to travel. • In the future, it is likely for congresses to offer attendance options both in person and online, making them more accessible to a larger audience. • Respondents indicated that networking represents the most valuable advantage of in-person conferences compared to online ones.

Intravenous SPION-labeled adipocyte-derived stem cells targeted to the brain by magnetic attraction in a rat stroke model: An ultrastructural insight into cell fate within the brain.

Mesenchymal stem cell therapy after stroke is a promising option investigated in animal models and clinical trials. The intravenous route is commonly used in clinical settings guaranteeing an adequate safety profile although low yields of engraftment. In this report, rats subjected to ischemic stroke were injected with adipose-derived stem cells (ADSCs) labeled with superparamagnetic iron oxide nanoparticles (SPIONs) applying an external magnetic field in the skull to retain the cells. Although most published studies demonstrate viability of ADSCs, only a few have used ultrastructural techniques. In our study, the application of a local magnetic force resulted in a tendency for higher yields of SPION-ADSCs targeting the brain. However, grafted cells displayed morphological signs of death, one day after administration, and correlative microscopy showed active microglia and astrocytes associated in the process of scavenging. Thus, we conclude that, although successfully targeted within the brain, SPION-ADSCs viability was rapidly compromised.

Automated Cervical Spinal Cord Segmentation in Real-World MRI of Multiple Sclerosis Patients by Optimized Hybrid Residual Attention-Aware Convolutional Neural Networks.

Magnetic resonance (MR) imaging is the most sensitive clinical tool in the diagnosis and monitoring of multiple sclerosis (MS) alterations. Spinal cord evaluation has gained interest in this clinical scenario in recent years, but, unlike the brain, there is a more limited choice of algorithms to assist spinal cord segmentation. Our goal was to investigate and develop an automatic MR cervical cord segmentation method, enabling automated and seamless spinal cord atrophy assessment and setting the stage for the development of an aggregated algorithm for the extraction of lesion-related imaging biomarkers. The algorithm was developed using a real-world MR imaging dataset of 121 MS patients (96 cases used as a training dataset and 25 cases as a validation dataset). Transversal, 3D T1-weighted gradient echo MR images (TE/TR/FA = 1.7-2.7 ms/5.6-8.2 ms/12°) were acquired in a 3 T system (Signa HD, GEHC) as standard of care in our clinical practice. Experienced radiologists supervised the manual labelling, which was considered the ground-truth. The 2D convolutional neural network consisted of a hybrid residual attention-aware segmentation method trained to delineate the cervical spinal cord. The training was conducted using a focal loss function, based on the Tversky index to address label imbalance, and an automatic optimal learning rate finder. Our automated model provided an accurate segmentation, achieving a validation DICE coefficient of 0.904 ± 0.101 compared with the manual delineation. An automatic method for cervical spinal cord segmentation on T1-weighted MR images was successfully implemented. It will have direct implications serving as the first step for accelerating the process for MS staging and follow-up through imaging biomarkers.

Artificial Intelligence on FDG PET Images Identifies Mild Cognitive Impairment Patients with Neurodegenerative Disease.

The purpose of this project is to develop and validate a Deep Learning (DL) FDG PET imaging algorithm able to identify patients with any neurodegenerative diseases (Alzheimer's Disease (AD), Frontotemporal Degeneration (FTD) or Dementia with Lewy Bodies (DLB)) among patients with Mild Cognitive Impairment (MCI). A 3D Convolutional neural network was trained using images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. The ADNI dataset used for the model training and testing consisted of 822 subjects (472 AD and 350 MCI). The validation was performed on an independent dataset from La Fe University and Polytechnic Hospital. This dataset contained 90 subjects with MCI, 71 of them developed a neurodegenerative disease (64 AD, 4 FTD and 3 DLB) while 19 did not associate any neurodegenerative disease. The model had 79% accuracy, 88% sensitivity and 71% specificity in the identification of patients with neurodegenerative diseases tested on the 10% ADNI dataset, achieving an area under the receiver operating characteristic curve (AUC) of 0.90. On the external validation, the model preserved 80% balanced accuracy, 75% sensitivity, 84% specificity and 0.86 AUC. This binary classifier model based on FDG PET images allows the early prediction of neurodegenerative diseases in MCI patients in standard clinical settings with an overall 80% classification balanced accuracy.

Data harmonisation for information fusion in digital healthcare: A state-of-the-art systematic review, meta-analysis and future research directions.

Removing the bias and variance of multicentre data has always been a challenge in large scale digital healthcare studies, which requires the ability to integrate clinical features extracted from data acquired by different scanners and protocols to improve stability and robustness. Previous studies have described various computational approaches to fuse single modality multicentre datasets. However, these surveys rarely focused on evaluation metrics and lacked a checklist for computational data harmonisation studies. In this systematic review, we summarise the computational data harmonisation approaches for multi-modality data in the digital healthcare field, including harmonisation strategies and evaluation metrics based on different theories. In addition, a comprehensive checklist that summarises common practices for data harmonisation studies is proposed to guide researchers to report their research findings more effectively. Last but not least, flowcharts presenting possible ways for methodology and metric selection are proposed and the limitations of different methods have been surveyed for future research.

Machine Learning-Based Integration of Prognostic Magnetic Resonance Imaging Biomarkers for Myometrial Invasion Stratification in Endometrial Cancer.

BACKGROUND: Estimation of the depth of myometrial invasion (MI) in endometrial cancer is pivotal in the preoperatively staging. Magnetic resonance (MR) reports suffer from human subjectivity. Multiparametric MR imaging radiomics and parameters may improve the diagnostic accuracy. PURPOSE: To discriminate between patients with MI ≥ 50% using a machine learning-based model combining texture features and descriptors from preoperatively MR images. STUDY TYPE: Retrospective. POPULATION: One hundred forty-three women with endometrial cancer were included. The series was split into training (n = 107, 46 with MI ≥ 50%) and test (n = 36, 16 with MI ≥ 50%) cohorts. FIELD STRENGTH/SEQUENCES: Fast spin echo T2-weighted (T2W), diffusion-weighted (DW), and T1-weighted gradient echo dynamic contrast-enhanced (DCE) sequences were obtained at 1.5 or 3 T magnets. ASSESSMENT: Tumors were manually segmented slice-by-slice. Texture metrics were calculated from T2W and ADC map images. Also, the apparent diffusion coefficient (ADC), wash-in slope, wash-out slope, initial area under the curve at 60 sec and at 90 sec, initial slope, time to peak and peak amplitude maps from DCE sequences were obtained as parameters. MR diagnostic models using single-sequence features and a combination of features and parameters from the three sequences were built to estimate MI using Adaboost methods. The pathological depth of MI was used as gold standard. STATISTICAL TEST: Area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, accuracy, positive predictive value, negative predictive value, precision and recall were computed to assess the Adaboost models performance. RESULTS: The diagnostic model based on the features and parameters combination showed the best performance to depict patient with MI ≥ 50% in the test cohort (accuracy = 86.1% and AUROC = 87.1%). The rest of diagnostic models showed a worse accuracy (accuracy = 41.67%-63.89% and AUROC = 41.43%-63.13%). DATA CONCLUSION: The model combining the texture features from T2W and ADC map images with the semi-quantitative parameters from DW and DCE series allow the preoperative estimation of myometrial invasion. EVIDENCE LEVEL: 4 TECHNICAL EFFICACY: Stage 3.

Precise whole liver automatic segmentation and quantification of PDFF and R2* on MR images.

OBJECTIVE: To automate the segmentation of whole liver parenchyma on multi-echo chemical shift encoded (MECSE) MR examinations using convolutional neural networks (CNNs) to seamlessly quantify precise organ-related imaging biomarkers such as the fat fraction and iron load. METHODS: A retrospective multicenter collection of 183 MECSE liver MR examinations was conducted. An encoder-decoder CNN was trained (107 studies) following a 5-fold cross-validation strategy to improve the model performance and ensure lack of overfitting. Proton density fat fraction (PDFF) and R2* were quantified on both manual and CNN segmentation masks. Different metrics were used to evaluate the CNN performance over both unseen internal (46 studies) and external (29 studies) validation datasets to analyze reproducibility. RESULTS: The internal test showed excellent results for the automatic segmentation with a dice coefficient (DC) of 0.93 ± 0.03 and high correlation between the quantification done with the predicted mask and the manual segmentation (rPDFF = 1 and rR2* = 1; p values < 0.001). The external validation was also excellent with a different vendor but the same magnetic field strength, proving the generalization of the model to other manufacturers with DC of 0.94 ± 0.02. Results were lower for the 1.5-T MR same vendor scanner with DC of 0.87 ± 0.06. Both external validations showed high correlation in the quantification (rPDFF = 1 and rR2* = 1; p values < 0.001). In both internal and external validation datasets, the relative error for the PDFF and R2* quantification was below 4% and 1% respectively. CONCLUSION: Liver parenchyma can be accurately segmented with CNN in a vendor-neutral virtual approach, allowing to obtain reproducible automatic whole organ virtual biopsies. KEY POINTS: • Whole liver parenchyma can be automatically segmented using convolutional neural networks. • Deep learning allows the creation of automatic pipelines for the precise quantification of liver-related imaging biomarkers such as PDFF and R2*. • MR "virtual biopsy" can become a fast and automatic procedure for the assessment of chronic diffuse liver diseases in clinical practice.