Liver steatosis in patients with transfusion-dependent thalassaemia.

We evaluated the prevalence and the clinical associations of liver steatosis (LS) in patients with transfusion-dependent thalassaemia (TDT). We considered 301 TDT patients (177 females, median age = 40.61 years) enrolled in the Extension-Myocardial Iron Overload in Thalassaemia Network, and 25 healthy subjects. Magnetic resonance imaging was used to quantify iron overload and hepatic fat fraction (FF) by T2* technique and cardiac function by cine images. The glucose metabolism was assessed by the oral glucose tolerance test (OGTT). Hepatic FF was significantly higher in TDT patients than in healthy subjects (median value: 1.48% vs. 0.55%; p = 0.013). In TDT, hepatic FF was not associated with age, gender, serum ferritin levels or liver function parameters, but showed a weak inverse correlation with high-density lipoprotein cholesterol. The 36.4% of TDT patients showed LS (FF >3.7%). Active hepatitis C virus (HCV) infection, increased body mass index and hepatic iron were independent determinants of LS. A hepatic FF >3.53% predicted the presence of an abnormal OGTT. Hepatic FF was not correlated with cardiac iron, biventricular volumes or ejection fractions, but was correlated with left ventricular mass index. In TDT, LS is a frequent finding, associated with iron overload, increased weight and HCV, and conveying an increased risk for the alterations of glucose metabolism.

Impact of genotype on multi-organ iron and complications in patients with non-transfusion-dependent ß-thalassemia intermedia.

We evaluated the impact of the genotype on clinical and hematochemical features, hepatic and cardiac iron levels, and endocrine, hepatic, and cardiovascular complications in non-transfusion-dependent (NTD) ß-thalassemia intermedia (TI) patients. Sixty patients (39.09 ± 11.11 years, 29 females) consecutively enrolled in the Myocardial Iron Overload in Thalassemia project underwent Magnetic Resonance Imaging to quantify iron overload, biventricular function parameters, and atrial areas and to detect replacement myocardial fibrosis. Three groups of patients were identified: homozygous ß+ (N = 18), heterozygous ß0ß+ (N = 22), and homozygous ß0 (N = 20). The groups were homogeneous for sex, age, splenectomy, hematochemical parameters, chelation therapy, and iron levels. The homozygous ß° genotype was associated with significantly higher biventricular end-diastolic and end-systolic volume indexes and bi-atrial area indexes. No difference was detected in biventricular ejection fractions or myocardial fibrosis. Extramedullary hematopoiesis and leg ulcers were significantly more frequent in the homozygous ß° group compared to the homozygous ß+ group. No association was detected between genotype and liver cirrhosis, hypogonadism, hypothyroidism, osteoporosis, heart failure, arrhythmias, and pulmonary hypertension. Heart remodelling related to a high cardiac output state cardiomyopathy, extramedullary hematopoiesis, and leg ulcers were more pronounced in patients with the homozygous ß° genotype compared to the other genotypes analyzed. The knowledge of the genotype can assist in the clinical management of NTD ß-TI patients.

Left atrial strain in patients with ß-thalassemia major: a cross-sectional CMR study.

OBJECTIVES: The aim of this cross-sectional study was to investigate the association of left atrial (LA) strain parameters with demographics, clinical data, cardiovascular magnetic resonance (CMR) findings, and cardiac complications (heart failure and arrhythmias) in a cohort of patients with ß-thalassemia major (ß-TM). MATERIALS AND METHODS: We considered 264 ß-TM patients (133 females, 36.79 ± 11.95 years) consecutively enrolled in the Extension-Myocardial Iron Overload in Thalassemia (E-MIOT) project. Moreover, we included 35 sex- and age-matched healthy controls (14 females, mean age 37.36 ± 17.52 years). Reservoir, conduit, and booster LA functions were analysed by CMR feature tracking using dedicated software. RESULTS: Compared to the healthy control group, ß-TM patients demonstrated lower LA reservoir strain and booster strains, as well as LA reservoir and booster strain rates. However, no differences were found in LA conduit deformation parameters. In ß-TM patients, ageing, sex, and left ventricle (LV) volume indexes were independent determinants of LA strain parameters. The number of segments with late gadolinium enhancement (LGE) significantly correlated with all LA strain parameters, with the exception of the LA conduit rate. Patients with cardiac complications exhibited significantly impaired strain parameters compared to patients without cardiac complications. CONCLUSION: In patients with ß-TM, LA strain parameters were impaired compared to control subjects, and they exhibited a significant correlation with the number of LV segments with LGE. Furthermore, patients with cardiac complications had impaired left atrial strain parameters. Clinical relevance statement In patients with ß-thalassemia major, left atrial strain parameters were impaired compared to control subjects and emerged as a sensitive marker of cardiac complications, stronger than cardiac iron levels. KEY POINTS: • Compared to healthy subjects, ß-thalassemia major patients demonstrated significantly lower left atrial reservoir strain and booster strains, as well as left atrial reservoir and booster strain rates. • In ß-thalassemia major, ageing, sex, and left ventricular volume indexes were independent determinants of left atrial strain parameters, while left atrial strain parameters were not correlated with myocardial iron overload. • An independent association between reduced left atrial strain parameters and a history of cardiac complications was found in ß-thalassemia major patients.

Pancreatic iron in pediatric transfusion-dependent beta-thalassemia patients: A longitudinal MRI study.

BACKGROUND: In pediatric transfusion-dependent thalassemia (TDT) patients, we evaluated the prevalence, pattern, and clinical associations of pancreatic siderosis and the changes in pancreatic iron levels and their association with baseline and changes in total body iron balance. PROCEDURE: We considered 86 pediatric TDT patients consecutively enrolled in the Extension-Myocardial Iron Overload in Thalassemia Network. Iron overload (IO) was quantified by R2* magnetic resonance imaging (MRI). RESULTS: Sixty-three (73%) patients had pancreatic IO (R2* > 38 Hz). Global pancreas R2* values were significantly correlated with mean serum ferritin levels, MRI liver iron concentration (LIC) values, and global heart R2* values. Global pancreas R2* values were significantly higher in patients with altered versus normal glucose metabolism. Thirty-one patients also performed the follow-up MRI at 18 ± 3 months. Higher pancreatic R2* values were detected at the follow-up, but the difference versus the baseline MRI was not significant. The 20% of patients with baseline pancreatic IO showed no pancreatic IO at the follow-up. The 46% of patients without baseline pancreatic IO developed pancreatic siderosis. The changes in global pancreas R2* between the two MRIs were not correlated with baseline serum ferritin levels, baseline, final, and changes in MRI LIC values, or baseline pancreatic iron levels. CONCLUSIONS: In children with TDT, pancreatic siderosis is a frequent finding associated with hepatic siderosis and represents a risk factor for myocardial siderosis and alterations of glucose metabolism. Iron removal from the pancreas is exceptionally challenging and independent from hepatic iron status.

Liver T1 and T2 mapping in a large cohort of healthy subjects: normal ranges and correlation with age and sex.

OBJECTIVE: We established normal ranges for native T1 and T2 values in the human liver using a 1.5 T whole-body imager (General Electric) and we evaluated their variation across hepatic segments and their association with age and sex. MATERIALS AND METHODS: One-hundred healthy volunteers aged 20-70 years (50% females) underwent MRI. Modified Look-Locker inversion recovery and multi-echo fast-spin-echo sequences were used to measure hepatic native global and segmental T1 and T2 values, respectively. RESULTS: T1 and T2 values exhibited good intra- and inter-observer reproducibility (coefficient of variation < 5%). T1 value over segment 4 was significantly lower than the T1 values over segments 2 and 3 (p < 0.0001). No significant regional T2 variability was detected. Segmental and global T1 values were not associated with age or sex. Global T2 values were independent from age but were significantly lower in males than in females. The lower and upper limits of normal for global T1 values were, respectively, 442 ms and 705 ms. The normal range for global T2 values was 35 ms-54 ms in males and 39 ms-54 ms in females. DISCUSSION: Liver T1 and T2 mapping is feasible and reproducible and the provided normal ranges may help to establish diagnosis and progression of various liver diseases.

Machine Learning for the Design and the Simulation of Radiofrequency Magnetic Resonance Coils: Literature Review, Challenges, and Perspectives.

Radiofrequency (RF) coils for magnetic resonance imaging (MRI) applications serve to generate RF fields to excite the nuclei in the sample (transmit coil) and to pick up the RF signals emitted by the nuclei (receive coil). For the purpose of optimizing the image quality, the performance of RF coils has to be maximized. In particular, the transmit coil has to provide a homogeneous RF magnetic field, while the receive coil has to provide the highest signal-to-noise ratio (SNR). Thus, particular attention must be paid to the coil simulation and design phases, which can be performed with different computer simulation techniques. Being largely used in many sectors of engineering and sciences, machine learning (ML) is a promising method among the different emerging strategies for coil simulation and design. Starting from the applications of ML algorithms in MRI and a short description of the RF coil's performance parameters, this narrative review describes the applications of such techniques for the simulation and design of RF coils for MRI, by including deep learning (DL) and ML-based algorithms for solving electromagnetic problems.

Three-Dimensional Multi-Modality Registration for Orthopaedics and Cardiovascular Settings: State-of-the-Art and Clinical Applications.

The multimodal and multidomain registration of medical images have gained increasing recognition in clinical practice as a powerful tool for fusing and leveraging useful information from different imaging techniques and in different medical fields such as cardiology and orthopedics. Image registration could be a challenging process, and it strongly depends on the correct tuning of registration parameters. In this paper, the robustness and accuracy of a landmarks-based approach have been presented for five cardiac multimodal image datasets. The study is based on 3D Slicer software and it is focused on the registration of a computed tomography (CT) and 3D ultrasound time-series of post-operative mitral valve repair. The accuracy of the method, as a function of the number of landmarks used, was performed by analysing root mean square error (RMSE) and fiducial registration error (FRE) metrics. The validation of the number of landmarks resulted in an optimal number of 10 landmarks. The mean RMSE and FRE values were 5.26 ± 3.17 and 2.98 ± 1.68 mm, respectively, showing comparable performances with respect to the literature. The developed registration process was also tested on a CT orthopaedic dataset to assess the possibility of reconstructing the damaged jaw portion for a pre-operative planning setting. Overall, the proposed work shows how 3D Slicer and registration by landmarks can provide a useful environment for multimodal/unimodal registration.

A Novel Workflow for Electrophysiology Studies in Patients with Brugada Syndrome.

Brugada Syndrome (BrS) is a primary electrical epicardial disease characterized by ST-segment elevation followed by a negative T-wave in the right precordial leads on the surface electrocardiogram (ECG), also known as the 'type 1' ECG pattern. The risk stratification of asymptomatic individuals with spontaneous type 1 ECG pattern remains challenging. Clinical and electrocardiographic prognostic markers are known. As none of these predictors alone is highly reliable in terms of arrhythmic prognosis, several multi-factor risk scores have been proposed for this purpose. This article presents a new workflow for processing endocardial signals acquired with high-density RV electro-anatomical mapping (HDEAM) from BrS patients. The workflow, which relies solely on Matlab software, calculates various electrical parameters and creates multi-parametric maps of the right ventricle. The workflow, but it has already been employed in several research studies involving patients carried out by our group, showing its potential positive impact in clinical studies. Here, we will provide a technical description of its functionalities, along with the results obtained on a BrS patient who underwent an endocardial HDEAM.

Multiparametric cardiac magnetic resonance in patients with thalassemia intermedia: new insights from the E-MIOT network.

PURPOSE: In a relatively large cohort of thalassemia intermedia (TI) patients, we systematically investigated myocardial iron overload (MIO), function, and replacement fibrosis using cardiac magnetic resonance (CMR), we assessed the clinical determinants of global heart T2* values, and we explored the association between multiparametric CMR findings and cardiac complications. MATERIALS AND METHODS: We considered 254 beta-TI patients (43.14 ± 13.69 years, 138 females) consecutively enrolled in the Extension-Myocardial Iron Overload in Thalassemia project. MIO was quantified by T2* technique and biventricular function and atrial areas by cine images. Macroscopic myocardial fibrosis was detected by late gadolinium enhancement technique. RESULTS: Compared to never/sporadically transfused patients, regularly transfused (RT)-TI patients exhibited significantly lower global heart T2* values, biventricular end-diastolic volume indexes, left ventricular mass index, and cardiac index. In RT-TI patients, age and serum ferritin levels were the strongest predictors of global heart T2* values. Independently from the transfusional state, cardiac T2* values were not associated with biventricular function. Of the 103 (40.6%) patients in whom the contrast medium was administrated, 27 (26.2%) had replacement myocardial fibrosis. Age, sex distribution, cardiac iron, and biventricular function parameters were comparable between patients without and without replacement myocardial fibrosis. Twenty-five (9.8%) patients had a history of cardiac complications (heart failure and arrhythmias). Increased age and replacement myocardial fibrosis emerged as significant risk markers for cardiac complications. CONCLUSIONS: In TI, regular transfusions are associated with less pronounced cardiac remodeling but increase the risk of MIO. Replacement myocardial fibrosis is a frequent finding associated with cardiac complications.

Deep Learning Staging of Liver Iron Content From Multiecho MR Images.

BACKGROUND: MRI represents the most established liver iron content (LIC) evaluation approach by estimation of liver T2* value, but it is dependent on the choice of the measurement region and the software used for image analysis. PURPOSE: To develop a deep-learning method for unsupervised classification of LIC from magnitude T2* multiecho MR images. STUDY TYPE: Retrospective. POPULATION/SUBJECTS: A total of 1069 thalassemia major patients enrolled in the core laboratory of the Myocardial Iron Overload in Thalassemia (MIOT) network, which were included in the training (80%) and test (20%) sets. Twenty patients from different MRI vendors included in the external test set. FIELD STRENGTH/SEQUENCE: A5 T, T2* multiecho magnitude images. ASSESSMENT: Four deep-learning convolutional neural networks (HippoNet-2D, HippoNet-3D, HippoNet-LSTM, and an ensemble network HippoNet-Ensemble) were used to achieve unsupervised staging of LIC using five classes (normal, borderline, middle, moderate, severe). The training set was employed to construct the deep-learning model. The performance of the LIC staging model was evaluated in the test set and in the external test set. The model's performances were assessed by evaluating the accuracy, sensitivity, and specificity with respect to the ground truth labels obtained by T2* measurements and by comparison with operator-induced variability originating from different region of interest (ROI) placements. STATISTICAL TESTS: The network's performances were evaluated by single-class accuracy, specificity, and sensitivity and compared by one-way repeated measures analysis of variance (ANOVA) and one-way ANOVA. RESULTS: HippoNet-Ensemble reached an accuracy significantly higher than the other networks, and a sensitivity and specificity higher than HippoNet-LSTM. Accuracy, sensitivity, and specificity values for the LIC stages were: normal: 0.96/0.93/0.97, borderline: 0.95/0.85/0.98, mild: 0.96/0.88/0.98, moderate: 0.95/0.89/0.97, severe: 0.97/0.95/0.98. Correctly staging of cases was in the range of 85%-95%, depending on the LIC class. Multiclass accuracy was 0.90 against 0.92 for the interobserver variability. DATA CONCLUSION: The proposed HippoNet-Ensemble network can perform unsupervised LIC staging and achieves good prognostic performance. EVIDENCE LEVEL: 4 TECHNICAL EFFICACY: Stage 2.

Prognostic value of multiparametric cardiac magnetic resonance in sickle cell patients.

The aim of this multicenter study was to prospectively assess the predictive value of multiparametric cardiac magnetic resonance (CMR) for cardiovascular complications in sickle cell disease (SCD) patients. Among all patients with hemoglobinopathies consecutively enrolled in the Myocardial Iron Overload in Thalassemia (MIOT) Network, we selected 102 SCD patients (34.38 ± 12.67 years, 49 females). Myocardial iron overload (MIO) was measured by the multislice multiecho T2* technique. Atrial dimensions and biventricular function parameters were quantified by cine images. Late gadolinium enhancement (LGE) images were acquired to detect focal myocardial fibrosis. At baseline CMR, only two patients had significant MIO (global heart T2* < 20 ms). During a mean follow-up of 63.01 ± 24.95 months, 11 cardiovascular events (10.8%) were registered: 3 pulmonary hypertension, 2 supraventricular arrhythmias, 1 heart failure, 1 death for heart failure, 1 pulmonary embolism, 1 peripheral vascular disease, 1 transient ischemic attack, and 1 death after acute chest syndrome. In the multivariate analysis, the independent CMR predictors of cardiovascular events were left ventricular (LV) ejection fraction (hazard ratio-HR = 0.88; p = 0.025) and right ventricular (RV) mass index (HR = 1.09; p = 0.047). According to the receiver-operating characteristic curve analysis for adverse events, an LV ejection fraction < 58.9% and an RV mass index > 31 g/m2 were optimal cut-off values. Reduced left ventricular ejection fraction and increased right ventricular mass index showed a significant prognostic value in patients with SCD. Our data seem to suggest that CMR may be added as a screening tool for identifying SCD patients at high risk for cardiopulmonary and vascular diseases.

Increased myocardial extracellular volume is associated with myocardial iron overload and heart failure in thalassemia major.

OBJECTIVES: Myocardial extracellular volume (ECV) by cardiovascular magnetic resonance (CMR) is a surrogate marker of diffuse fibrosis. We evaluated the association between ECV and demographics, CMR findings, and cardiac involvement in patients with thalassemia major (TM). METHODS: A total of 108 ß-TM patients (62 females, 40.16 ± 8.83 years), consecutively enrolled in the Extension-Myocardial Iron Overload in Thalassemia Network, and 16 healthy subjects (6 females, 37.12 ± 16.13 years) underwent CMR. The protocol included assessment of T2*, native T1, and T2 values in all 16 myocardial segments for myocardial iron overload (MIO) quantification, cine images for left ventricular (LV) function quantification, post-contrast T1 mapping for ECV calculation, and late gadolinium enhancement (LGE) technique for replacement myocardial fibrosis detection. RESULTS: Global ECV values were significantly higher in females than in males. Global ECV values were significantly higher in patients with significant MIO (global heart T2* < 20 ms) than in patients without significant MIO, and both groups exhibited higher global ECV values than healthy subjects. No association was detected between native T1 and ECV values, while patients with reduced global heart T2 values showed significantly higher global ECV values than patients with normal and increased global heart T2. Global ECV values were not correlated with LV function/size and were comparable between patients with and without LGE. Compared to patients without heart failure, patients with a history of heart failure (N = 10) showed significantly higher global heart ECV values. CONCLUSION: In TM, increased myocardial ECV, potentially reflecting diffuse interstitial fibrosis, is associated with MIO and heart failure. KEY POINTS: • CMR-derived myocardial extracellular volume is increased in thalassemia major patients, irrespective of the presence of late gadolinium enhancement. • In thalassemia major, myocardial iron overload contributes to the increase in myocardial ECV, which potentially reflects diffuse interstitial fibrosis and is significantly associated with a history of heart failure.

Pancreatic fatty replacement as risk marker for altered glucose metabolism and cardiac iron and complications in thalassemia major.

OBJECTIVES: This multicenter study assessed the extent of pancreatic fatty replacement and its correlation with demographics, iron overload, glucose metabolism, and cardiac complications in a cohort of well-treated patients with thalassemia major (TM). METHODS: We considered 308 TM patients (median age: 39.79 years; 182 females) consecutively enrolled in the Extension-Myocardial Iron Overload in Thalassemia Network. Magnetic resonance imaging was used to quantify iron overload (IO) and pancreatic fat fraction (FF) by T2* technique, cardiac function by cine images, and to detect replacement myocardial fibrosis by late gadolinium enhancement technique. The glucose metabolism was assessed by the oral glucose tolerance test. RESULTS: Pancreatic FF was associated with age, body mass index, and history of hepatitis C virus infection. Patients with normal glucose metabolism showed a significantly lower pancreatic FF than patients with impaired fasting glucose (p = 0.030), impaired glucose tolerance (p < 0.0001), and diabetes (p < 0.0001). A normal pancreatic FF (< 6.6%) showed a negative predictive value of 100% for abnormal glucose metabolism. A pancreatic FF > 15.33% predicted the presence of abnormal glucose metabolism. Pancreas FF was inversely correlated with global pancreas and heart T2* values. A normal pancreatic FF showed a negative predictive value of 100% for cardiac iron. Pancreatic FF was significantly higher in patients with myocardial fibrosis (p = 0.002). All patients with cardiac complications had fatty replacement, and they showed a significantly higher pancreatic FF than complications-free patients (p = 0.002). CONCLUSION: Pancreatic FF is a risk marker not only for alterations of glucose metabolism, but also for cardiac iron and complications, further supporting the close link between pancreatic and cardiac disease. KEY POINTS: • In thalassemia major, pancreatic fatty replacement by MRI is a frequent clinical entity, predicted by a pancreas T2* < 20.81 ms and associated with a higher risk of alterations in glucose metabolism. • In thalassemia major, pancreatic fatty replacement is a strong risk marker for cardiac iron, replacement fibrosis, and complications, highlighting a deep connection between pancreatic and cardiac impairment.

National networking in rare diseases and reduction of cardiac burden in thalassemia major.

AIMS: A tailored chelation therapy guided by magnetic resonance imaging (MRI) is a strategy to improve the prognosis in iron-loaded patients, in many cases still hampered by limited MRI availability. In order to address this issue, the Myocardial Iron Overload in Thalassemia (MIOT) network was established in Italy and we aimed to describe the impact of 10-year activity of this network on cardiac burden in thalassemia major (TM). METHODS AND RESULTS: Within the MIOT network, 1746 TM patients (911 females; mean age 31.2 ± 9.1 years) were consecutively enrolled and prospectively followed by 70 thalassemia and 10 MRI centres. Patients were scanned using a multiparametric approach for assessing myocardial iron overload (MIO), biventricular function, and myocardial fibrosis. At the last MRI scan, a significant increase in global heart T2* values and a significantly higher frequency of patients with no MIO (all segmental T2* ≥20 ms) were detected, with a concordant improvement in biventricular function, particularly in patients with baseline global heart T2* <20 ms. Forty-seven percentage of patients changed the chelation regimen based on MRI. The frequency of heart failure (HF) significantly decreased after baseline MRI from 3.5 to 0.8% (P < 0.0001). Forty-six patients died during the study, and HF accounted for 34.8% of deaths. CONCLUSION: Over 10 years, continuous monitoring of cardiac iron and a tailored chelation therapy allowed MIO reduction, with consequent improvement of cardiac function and reduction of cardiac complications and mortality from MIO-related HF. A national networking for rare diseases therefore proved effective in improving the care and reducing cardiac outcomes of TM patients.

Left Ventricle Detection from Cardiac Magnetic Resonance Relaxometry Images Using Visual Transformer.

Left Ventricle (LV) detection from Cardiac Magnetic Resonance (CMR) imaging is a fundamental step, preliminary to myocardium segmentation and characterization. This paper focuses on the application of a Visual Transformer (ViT), a novel neural network architecture, to automatically detect LV from CMR relaxometry sequences. We implemented an object detector based on the ViT model to identify LV from CMR multi-echo T2* sequences. We evaluated performances differentiated by slice location according to the American Heart Association model using 5-fold cross-validation and on an independent dataset of CMR T2*, T2, and T1 acquisitions. To the best of our knowledge, this is the first attempt to localize LV from relaxometry sequences and the first application of ViT for LV detection. We collected an Intersection over Union (IoU) index of 0.68 and a Correct Identification Rate (CIR) of blood pool centroid of 0.99, comparable with other state-of-the-art methods. IoU and CIR values were significantly lower in apical slices. No significant differences in performances were assessed on independent T2* dataset (IoU = 0.68, p = 0.405; CIR = 0.94, p = 0.066). Performances were significantly worse on the T2 and T1 independent datasets (T2: IoU = 0.62, CIR = 0.95; T1: IoU = 0.67, CIR = 0.98), but still encouraging considering the different types of acquisition. This study confirms the feasibility of the application of ViT architectures in LV detection and defines a benchmark for relaxometry imaging.

An Explainable Convolutional Neural Network for the Early Diagnosis of Alzheimer's Disease from 18F-FDG PET.

Convolutional Neural Networks (CNN) which support the diagnosis of Alzheimer's Disease using 18F-FDG PET images are obtaining promising results; however, one of the main challenges in this domain is the fact that these models work as black-box systems. We developed a CNN that performs a multiclass classification task of volumetric 18F-FDG PET images, and we experimented two different post hoc explanation techniques developed in the field of Explainable Artificial Intelligence: Saliency Map (SM) and Layerwise Relevance Propagation (LRP). Finally, we quantitatively analyze the explanations returned and inspect their relationship with the PET signal. We collected 2552 scans from the Alzheimer's Disease Neuroimaging Initiative labeled as Cognitively Normal (CN), Mild Cognitive Impairment (MCI), and Alzheimer's Disease (AD) and we developed and tested a 3D CNN that classifies the 3D PET scans into its final clinical diagnosis. The model developed achieves, to the best of our knowledge, performances comparable with the relevant literature on the test set, with an average Area Under the Curve (AUC) for prediction of CN, MCI, and AD 0.81, 0.63, and 0.77 respectively. We registered the heatmaps with the Talairach Atlas to perform a regional quantitative analysis of the relationship between heatmaps and PET signals. With the quantitative analysis of the post hoc explanation techniques, we observed that LRP maps were more effective in mapping the importance metrics in the anatomic atlas. No clear relationship was found between the heatmap and the PET signal.

Bayesian Convolutional Neural Networks in Medical Imaging Classification: A Promising Solution for Deep Learning Limits in Data Scarcity Scenarios.

Deep neural networks (DNNs) have already impacted the field of medicine in data analysis, classification, and image processing. Unfortunately, their performance is drastically reduced when datasets are scarce in nature (e.g., rare diseases or early-research data). In such scenarios, DNNs display poor capacity for generalization and often lead to highly biased estimates and silent failures. Moreover, deterministic systems cannot provide epistemic uncertainty, a key component to asserting the model's reliability. In this work, we developed a probabilistic system for classification as a framework for addressing the aforementioned criticalities. Specifically, we implemented a Bayesian convolutional neural network (BCNN) for the classification of cardiac amyloidosis (CA) subtypes. We prepared four different CNNs: base-deterministic, dropout-deterministic, dropout-Bayesian, and Bayesian. We then trained them on a dataset of 1107 PET images from 47 CA and control patients (data scarcity scenario). The Bayesian model achieved performances (78.28 (1.99) % test accuracy) comparable to the base-deterministic, dropout-deterministic, and dropout-Bayesian ones, while showing strongly increased "Out of Distribution" input detection (validation-test accuracy mismatch reduction). Additionally, both the dropout-Bayesian and the Bayesian models enriched the classification through confidence estimates, while reducing the criticalities of the dropout-deterministic and base-deterministic approaches. This in turn increased the model's reliability, also providing much needed insights into the network's estimates. The obtained results suggest that a Bayesian CNN can be a promising solution for addressing the challenges posed by data scarcity in medical imaging classification tasks.

Frequency, pattern, and associations of renal iron accumulation in sickle/ß-thalassemia patients.

We evaluated frequency, pattern, and associations of renal iron accumulation in sickle/ß-thalassemia. Thirty-three sickle/ß-thalassemia patients (36.5 ± 14.7 years; 13 females), 14 homozygous sickle cell disease (SCD) patients, and 71 thalassemia major (TM) patients, enrolled in the E-MIOT Network, underwent magnetic resonance imaging. Iron overload (IO) was quantified by the T2* technique. Sickle/ß-thalassemia patients had a significantly lower frequency of renal IO (T2* < 31 ms) than homozygous SCD patients (9.1% vs. 57.1%; P = 0.001), besides having similar hepatic, cardiac and pancreatic IO. Kidney T2* values were comparable between regularly transfused sickle/ß-thalassemia and TM patients but were significantly lower in regularly transfused homozygous SCD patients than in the other two groups. In sickle/ß-thalassemia patients, global renal T2* values were not associated with age, gender, splenectomy, and presence of regular transfusions or chelation. No correlation was detected between renal T2* values and serum ferritin levels or iron load in the other organs. Global renal T2* values were not associated with serum creatinine levels but showed a significant inverse correlation with serum lactate dehydrogenase (R = - 0.709; P < 0.0001) and indirect bilirubin (R = - 0.462; P = 0.012). Renal IO is not common in sickle/ß-thalassemia patients, with a prevalence significantly lower compared to that of homozygous SCD patients, but with a similar underlying mechanism due to the chronic hemolysis.

Adipose tissue dysfunction and visceral fat are associated with hepatic insulin resistance and severity of NASH even in lean individuals.

BACKGROUND & AIMS: Non-alcoholic fatty liver disease (NAFLD) is a heterogeneous disorder, but the factors that determine this heterogeneity remain poorly understood. Adipose tissue dysfunction is causally linked to NAFLD since it causes intrahepatic triglyceride (IHTG) accumulation through increased hepatic lipid flow, due to insulin resistance and pro-inflammatory adipokines release. While many studies in NAFLD have looked at total adiposity (i.e. mainly subcutaneous fat, SC-AT), it is still unclear the possible impact of visceral fat (VF). Thus, we investigated how VF versus SC-AT was related to NAFLD severity in lean, overweight and obese individuals versus lean controls. METHODS: Thirty-two non-diabetic NAFLD with liver biopsy (BMI 21.4-34.7 kg/m2 ) and eight lean individuals (BMI 19.6-22.8 kg/m2 ) were characterized for fat distribution (VF, SC-AT and IHTG by magnetic resonance imaging), lipolysis and insulin resistance by tracer infusion, free fatty acids (FFAs) and triglyceride (TAG) concentration and composition (by mass spectrometry). RESULTS: Intrahepatic triglyceride was positively associated with lipolysis, adipose tissue insulin resistance (Adipo-IR), TAG concentrations, and increased saturated/unsaturated FFA ratio. Compared to controls VF was higher in NAFLD (including lean individuals), increased with fibrosis stage and associated with insulin resistance in liver, muscle and adipose tissue, increased lipolysis and decreased adiponectin levels. Collectively, our results suggest that VF accumulation, given its location close to the liver, is one of the major risk factors for NAFLD. CONCLUSIONS: These findings propose VF as an early indicator of NAFLD progression independently of BMI, which may allow for evidence-based prevention and intervention strategies.

Relationship between pancreatic iron overload, glucose metabolism and cardiac complications in sickle cell disease: An Italian multicentre study.

OBJECTIVES: Evidence about the cross-talk between iron, glucose metabolism, and cardiac disease is increasing. We aimed to explore the link of pancreatic iron by Magnetic Resonance Imaging (MRI) with glucose metabolism and cardiac complications (CC) in sickle cell disease (SCD) patients. METHODS: We considered 70 SCD patients consecutively enrolled in the Extension-Myocardial Iron Overload in Thalassemia Network. Iron overload was quantified by R2* technique and biventricular function by cine images. Macroscopic myocardial fibrosis was evaluated by late gadolinium enhancement technique. Glucose metabolism was assessed by the oral glucose tolerance test. RESULTS: Patients with an altered glucose metabolism showed a significantly higher pancreas R2* than patients with normal glucose metabolism. Pancreatic siderosis emerged as a risk factor for the development of metabolic alterations (OddsRatio 8.25, 95%confidence intervals 1.51-45.1; p = .015). Global pancreas R2* values were directly correlated with mean serum ferritin levels and liver iron concentration. Global pancreas R2* was not significantly associated with global heart R2* and macroscopic myocardial fibrosis. Patients with history of CC showed a significantly higher global pancreas R2* than patients with no CC. CONCLUSIONS: Our findings support the evaluation of pancreatic R2* by MRI in SCD patients to prevent the development of metabolic and cardiac disorders.