Mophological analysis of the human vasculature from medical imaging. Application to cerebral and coranary arteries

In this project, we have investigated new ways of modelling and analysis of human vasculature from Medical images. The research was divided in two main areas: cerebral vasculature analysis and coronary arteries modeling. Regarding cerebral vasculature analysis, we have studed cerebral aneurysms, internal carotid and the Circle of Willis (CoW). Aneurysms are abnormal vessel enlargements that can rupture causing important cerebral damages or death. The understanding of this pathology, together with its virtual treatment, and image diagnosis and prognosis, includes identification and detailed measurement of the aneurysms. In this context, we have proposed two automatic aneurysm isolation method, to separate the abnormal part of the vessel from the healthy part, to homogenize and speed-up the processing pipeline usually employed to study this pathology, [Cardenes2011TMI, arrabide2011MedPhys]. The results obtained from both methods have been also compared and validatied in [Cardenes2012MBEC]. A second important task here the analysis of the internal carotid [Bogunovic2011Media] and the automatic labelling of the CoW, Bogunovic2011MICCAI, Bogunovic2012TMI]. The second area of research covers the study of coronary arteries, specially coronary bifurcations because there is where the formation of atherosclerotic plaque is more common, and where the intervention is more challenging. Therefore, we proposed a novel modelling method from Computed Tomography Angiography (CTA) images, combined with Conventional Coronary Angiography (CCA), to obtain realistic vascular models of coronary bifurcations, presented in [Cardenes2011MICCAI], and fully validated including phantom experiments in [Cardene2013MedPhys]. The realistic models obtained from this method are being used to simulate stenting procedures, and to investigate the hemodynamic variables in coronary bifurcations in the works submitted in [Morlachi2012, Chiastra2012]. Additionally, another preliminary work has been done to reconstruct the coronary tree from rotational angiography, and published in [Cardenes2012ISBI].

Quality Assessment of Cardiac Magnetic Resonance Images at 1.5/3.0 T: Description and Validation of Standardized Criteria

PURPOSE: Cardiovascular magnetic resonance (CMR) has become a robust and important diagnostic imaging modality in cardiovascular medicine. However,insufficient image quality may compromise its diagnostic accuracy. No standardized criteria are available to assess the quality of CMR studies. We aimed todescribe and validate standardized criteria to evaluate the quality of CMR studies including: a) cine steady-state free precession, b) delayed gadoliniumenhancement, and c) adenosine stress first-pass perfusion. These criteria will serve for the assessment of the image quality in the setting of the Euro-CMR registry.METHOD AND MATERIALS: First, a total of 45 quality criteria were defined (35 qualitative criteria with a score from 0-3, and 10 quantitative criteria). Thequalitative score ranged from 0 to 105. The lower the qualitative score, the better the quality. The quantitative criteria were based on the absolute signal intensity (delayed enhancement) and on the signal increase (perfusion) of the anterior/posterior left ventricular wall after gadolinium injection. These criteria were then applied in 30 patients scanned with a 1.5T system and in 15 patients scanned with a 3.0T system. The examinations were jointly interpreted by 3 CMR experts and 1 study nurse. In these 45 patients the correlation between the results of the quality assessment obtained by the different readers was calculated.RESULTS: On the 1.5T machine, the mean quality score was 3.5. The mean difference between each pair of observers was 0.2 (5.7%) with a mean standarddeviation of 1.4. On the 3.0T machine, the mean quality score was 4.4. The mean difference between each pair of onservers was 0.3 (6.4%) with a meanstandard deviation of 1.6. The quantitative quality assessments between observers were well correlated for the 1.5T machine: R was between 0.78 and 0.99 (pCONCLUSION: The described criteria for the assessment of CMR image quality are robust and have a low inter-observer variability, especially on 1.5T systems.CLINICAL RELEVANCE/APPLICATION: These criteria will allow the standardization of CMR examinations. They will help to improve the overall quality ofexaminations and the comparison between clinical studies.

Value of positron emission tomography in full-thickness chest wall resections for malignancies

Rapport de synthèse : Les tumeurs de la paroi thoracique sont des pathologies graves dont le traitement principal consiste en une résection chirurgicale. L'enjeu majeur de cette intervention ontologique est de réséquer la totalité de la tumeur, ce qui nécessite une planification préopératoire minutieuse. Classiquement, l'identification et la localisation de la tumeur se fait à l'aide de la tomodensitométrie (computed tomography, CT) ou de l'imagerie par résonnance magnétique (1RM). Actuellement, l'imagerie nucléaire fonctionnelle par tomographie par émission de positons (positron emission tomography, PET) qui peut être couplée au CT (PET/CT) est de plus en plus appliquée aux patients présentant une tumeur maligne. Son efficacité a fréquemment été démontrée. Le but de la présente étude est d'évaluer la valeur du PET dans la planification de la résection des tumeurs de la paroi thoracique. Une analyse rétrospective de dix-huit patients opérés entre 2004 et 2006 a été réalisée; Dans ce groupe de patient, la taille de la tumeur mesurée sur la pièce opératoire réséquée a été comparée à la taille de la tumeur mesurée sur le CT et le PET. Les résultats démontrent que le CT surestimait de manière consistante la taille réelle de la tumeur par rapport au PET (+64% par rapport à +1%, P<0.001). De plus, le PET s'est avéré particulièrement performant pour prédire la taille des tumeurs de plus de 5.5 cm de diamètre par rapport au CT (valeur prédictive positive 80% par rapport à 44% et spécificité 93% par apport à 64%, respectivement). Cette étude démontre que le PET permettrait de mesurer la taille des tumeurs de la paroi thoracique de manière plus précise que le CT. Cette nouvelle modalité diagnostique s'avèrerait donc utile pour planifier les résections chirurgicales de telles tumeurs. A notre connaissance, aucune publication ne décrit la valeur du PET dans ce domaine. Les performances accrues du PET permettraient une meilleure délimitation des tumeurs ce qui améliorerait la précision de la résection chirurgicale. En conclusion, cette étude préliminaire rétrospective démontre la faisabilité du PET pour les tumeurs de la paroi thoracique. Ces résultats devraient être confirmés par une étude prospective incluant un plus grand nombre de patients avec la perspective de juger l'impact clinique réel du PET sur la prise en charge thérapeutique des patients.

BIMS: Biomedical Information Management System

This final year project presents the design principles and prototype implementation of BIMS (Biomedical Information Management System), a flexible software system which provides an infrastructure to manage all information required by biomedical research projects.The BIMS project was initiated with the motivation to solve several limitations in medical data acquisition of some research projects, in which Universitat Pompeu Fabra takes part. These limitations,based on the lack of control mechanisms to constraint information submitted by clinicians, impact on the data quality, decreasing it.BIMS can easily be adapted to manage information of a wide variety of clinical studies, not being limited to a given clinical specialty. The software can manage both, textual information, like clinical data (measurements, demographics, diagnostics, etc ...), as well as several kinds of medical images (magnetic resonance imaging, computed tomography, etc ...). Moreover, BIMS provides a web - based graphical user interface and is designed to be deployed in a distributed andmultiuser environment. It is built on top of open source software products and frameworks.Specifically, BIMS has been used to represent all clinical data being currently used within the CardioLab platform (an ongoing project managed by Universitat Pompeu Fabra), demonstratingthat it is a solid software system, which could fulfill requirements of a real production environment.

Increase of [(18)F]FLT tumor uptake in vivo mediated by FdUrd: toward improving cell proliferation positron emission tomography.

PURPOSE: 3'-deoxy-3'-[(18)F]fluorothymidine ([(18)F]FLT), a cell proliferation positron emission tomography (PET) tracer, has been shown in numerous tumors to be more specific than 2-deoxy-2-[(18)F]fluoro-D-glucose ([(18)F]FDG) but less sensitive. We studied the capacity of a nontoxic concentration of 5-fluoro-2'-deoxyuridine (FdUrd), a thymidine synthesis inhibitor, to increase uptake of [(18)F]FLT in tumor xenografts. METHODS: The duration of the FdUrd effect in vivo on tumor cell cycling and thymidine analogue uptake was studied by varying FdUrd pretreatment timing and holding constant the timing of subsequent flow cytometry and 5-[(125)I]iodo-2'-deoxyuridine biodistribution measurements. In [(18)F]FLT studies, FdUrd pretreatment was generally performed 1 h before radiotracer injection. [(18)F]FLT biodistributions were measured 1 to 3 h after radiotracer injection of mice grafted with five different human tumors and pretreated or not with FdUrd and compared with [(18)F]FDG tumor uptake. Using microPET, the dynamic distribution of [(18)F]FLT was followed for 1.5 h in FdUrd pretreated mice. High-field T2-weighted magnetic resonance imaging (MRI) and histology were used comparatively in assessing tumor viability and proliferation. RESULTS: FdUrd induced an immediate increase in tumor uptake of 5-[(125)I]iodo-2'-deoxyuridine, that vanished after 6 h, as also confirmed by flow cytometry. Biodistribution measurements showed that FdUrd pretreatment increased [(18)F]FLT uptake in all tumors by factors of 3.2 to 7.8 compared with controls, while [(18)F]FDG tumor uptake was about fourfold and sixfold lower in breast cancers and lymphoma. Dynamic PET in FdUrd pretreated mice showed that [(18)F]FLT uptake in all tumors increased steadily up to 1.5 h. MRI showed a well-vascularized homogenous lymphoma with high [(18)F]FLT uptake, while in breast cancer, a central necrosis shown by MRI was inactive in PET, consistent with the histomorphological analysis. CONCLUSION: We showed a reliable and significant uptake increase of [(18)F]FLT in different tumor xenografts after low-dose FdUrd pretreatment. These results show promise for a clinical application of FdUrd aimed at increasing the sensitivity of [(18)F]FLT PET.

Selective coronary artery plaque visualization and differentiation by contrast-enhanced inversion prepared MRI.

AIMS: We sought to evaluate the utility of contrast-enhanced coronary magnetic resonance imaging (CE-MRI) for selective visualization and non-invasive differentiation of atherosclerotic coronary plaque in humans. METHODS AND RESULTS: Nine patients with coronary artery disease (CAD) as confirmed by X-ray angiography and multidetector computed tomography (MDCT) were studied by T1-weighted black blood inversion recovery coronary MRI before (N-IR) and after administration of Gd-DTPA (CE-IR). Plaques were categorized as calcified, non-calcified, and mixed based on their Hounsfield number derived from MDCT. With MDCT, a total of 29 plaques were identified, including calcified (n=6), non-calcified (n=6), and mixed calcified/non-calcified (n=17). On N-IR MRI, 26 plaques (90%) were dark, whereas three plaques (two non-calcified and one mixed) appeared bright. On CE-MRI, 13/29 (45%) plaques, 11 of which were mixed, one non-calcified, and one calcified showed contrast uptake. All others remained dark. CONCLUSION: In this preliminary study, we demonstrate the potential utility of CE-IR MRI for selective plaque visualization and differentiation of plaque types. The observed contrast uptake may be associated with endothelial dysfunction, neovascularization, inflammation, and/or fibrosis.

In vivo loading increases mechanical properties of scaffold by affecting bone formation and bone resorption rates.

A successful bone tissue engineering strategy entails producing bone-scaffold constructs with adequate mechanical properties. Apart from the mechanical properties of the scaffold itself, the forming bone inside the scaffold also adds to the strength of the construct. In this study, we investigated the role of in vivo cyclic loading on mechanical properties of a bone scaffold. We implanted PLA/β-TCP scaffolds in the distal femur of six rats, applied external cyclic loading on the right leg, and kept the left leg as a control. We monitored bone formation at 7 time points over 35 weeks using time-lapsed micro-computed tomography (CT) imaging. The images were then used to construct micro-finite element models of bone-scaffold constructs, with which we estimated the stiffness for each sample at all time points. We found that loading increased the stiffness by 60% at 35 weeks. The increase of stiffness was correlated to an increase in bone volume fraction of 18% in the loaded scaffold compared to control scaffold. These changes in volume fraction and related stiffness in the bone scaffold are regulated by two independent processes, bone formation and bone resorption. Using time-lapsed micro-CT imaging and a newly-developed longitudinal image registration technique, we observed that mechanical stimulation increases the bone formation rate during 4-10 weeks, and decreases the bone resorption rate during 9-18 weeks post-operatively. For the first time, we report that in vivo cyclic loading increases mechanical properties of the scaffold by increasing the bone formation rate and decreasing the bone resorption rate.

EuroCMR (European Cardiovascular Magnetic Resonance) registry: results of the German pilot phase.

OBJECTIVES: During its German pilot phase, the EuroCMR (European Cardiovascular Magnetic Resonance) registry sought to evaluate indications, image quality, safety, and impact on patient management of routine CMR. BACKGROUND: CMR has a broad range of applications and is increasingly used in clinical practice. METHODS: This was a multicenter registry with consecutive enrollment of patients in 20 German centers. RESULTS: A total of 11,040 consecutive patients were enrolled. Eighty-eight percent of patients received gadolinium-based contrast agents. Twenty-one percent underwent adenosine perfusion, and 11% high-dose dobutamine-stress CMR. The most important indications were workup of myocarditis/cardiomyopathies (32%), risk stratification in suspected coronary artery disease/ischemia (31%), as well as assessment of viability (15%). Image quality was good in 90.1%, moderate in 8.1%, and inadequate in 1.8% of cases. Severe complications occurred in 0.05%, and were all associated with stress testing. No patient died during or due to CMR. In nearly two-thirds of patients, CMR findings impacted patient management. Importantly, in 16% of cases the final diagnosis based on CMR was different from the diagnosis before CMR, leading to a complete change in management. In more than 86% of cases, CMR was capable of satisfying all imaging needs so that no further imaging was required. CONCLUSIONS: CMR is frequently performed in clinical practice in many participating centers. The most important indications are workup of myocarditis/cardiomyopathies, risk stratification in suspected coronary artery disease/ischemia, and assessment of viability. CMR imaging as used in the centers of the pilot registry is a safe procedure, has diagnostic image quality in 98% of cases, and its results have strong impact on patient management.

Focal dependent pleural thickening at MDCT: pleural lesion or functional abnormality?

PURPOSE: To describe the characteristics of reversible focal pleural thickenings (PTs) mimicking real plaques, that firstly suggest asbestos exposure or pleural metastasis; to propose an imaging strategy and propose an explanation for their mechanism of formation. PATIENTS AND METHODS: Retrospective review of data from 19 patients with PTs fitting the description of pleural plaques at chest computed tomography (CT) and presenting modifications (clearance or appearance) of at least one PT at an additional chest examination in prone position. RESULTS: A total of 152 PTs were recorded on the first chest CT examinations with a range of two to 19 pleural opacities per patient. All PTs had a posterior distribution in the lower lobes. On the additional acquisitions, 144 PTs disappeared. Seventeen patients presented complete regression of PTs and two patients presented persistence of eight PTs. CONCLUSION: Additional low dose acquisition in prone position should be performed in all patients presenting with focal PT in a dependent and basal location. This may allow to exclude a pleural plaque in case of asbestos exposure but also a pleural metastasis in oncologic patients. These reversible dependent PTs could be related to physiological focal accumulation of lymphatic fluid in subpleural area.

Landmark detection for fusion of fundus and MRI toward a patient-specific multimodal eye model.

Ophthalmologists typically acquire different image modalities to diagnose eye pathologies. They comprise, e.g., Fundus photography, optical coherence tomography, computed tomography, and magnetic resonance imaging (MRI). Yet, these images are often complementary and do express the same pathologies in a different way. Some pathologies are only visible in a particular modality. Thus, it is beneficial for the ophthalmologist to have these modalities fused into a single patient-specific model. The goal of this paper is a fusion of Fundus photography with segmented MRI volumes. This adds information to MRI that was not visible before like vessels and the macula. This paper contributions include automatic detection of the optic disc, the fovea, the optic axis, and an automatic segmentation of the vitreous humor of the eye.

Trabecular bone score is associated with volumetric bone density and microarchitecture as assessed by central QCT and HRpQCT in Chinese American and white women.

Although high-resolution peripheral quantitative computed tomography (HRpQCT) and central quantitative computed tomography (QCT) studies have shown bone structural differences between Chinese American (CH) and white (WH) women, these techniques are not readily available in the clinical setting. The trabecular bone score (TBS) estimates trabecular microarchitecture from dual-energy X-ray absorptiometry spine images. We assessed TBS in CH and WH women and investigated whether TBS is associated with QCT and HRpQCT indices. Areal bone mineral density (aBMD) by dual-energy X-ray absorptiometry, lumbar spine (LS) TBS, QCT of the LS and hip, and HRpQCT of the radius and tibia were performed in 71 pre- (37 WH and 34 CH) and 44 postmenopausal (21 WH and 23 CH) women. TBS did not differ by race in either pre- or postmenopausal women. In the entire cohort, TBS positively correlated with LS trabecular volumetric bone mineral density (vBMD) (r = 0.664), femoral neck integral (r = 0.651), trabecular (r = 0.641) and cortical vBMD (r = 0.346), and cortical thickness (C/I; r = 0.540) by QCT (p < 0.001 for all). TBS also correlated with integral (r = 0.643), trabecular (r = 0.574) and cortical vBMD (r = 0.491), and C/I (r = 0.541) at the total hip (p < 0.001 for all). The combination of TBS and LS aBMD predicted more of the variance in QCT measures than aBMD alone. TBS was associated with all HRpQCT indices (r = 0.20-0.52) except radial cortical thickness and tibial trabecular thickness. Significant associations between TBS and measures of HRpQCT and QCT in WH and CH pre- and postmenopausal women demonstrated here suggest that TBS may be a useful adjunct to aBMD for assessing bone quality.

Use of 2D Sensitivity Encoding for Slow-Infusion Contrast-Enhanced Isotropic 3-T Whole-Heart Coronary MR Angiography.

OBJECTIVE. The purpose of this study was to improve the blood-pool signal-to-noise ratio (SNR) and blood-myocardium contrast-to-noise ratio (CNR) of slow-infusion 3-T whole-heart coronary MR angiography (MRA).SUBJECTS AND METHODS. In 2D sensitivity encoding (SENSE), the number of acquired k-space lines is reduced, allowing less radiofrequency excitation per cardiac cycle and a longer TR. The former can be exploited for signal enhancement with a higher radiofrequency excitation angle, and the latter leads to noise reduction due to lower data-sampling bandwidth. Both effects contribute to SNR gain in coronary MRA when spatial and temporal resolution and acquisition time remain identical. Numeric simulation was performed to select the optimal 2D SENSE pulse sequence parameters and predict the SNR gain. Eleven patients underwent conventional unenhanced and the proposed 2D SENSE contrast-enhanced coronary MRA acquisition. Blood-pool SNR, blood-myocardium CNR, visible vessel length, vessel sharpness, and number of side branches were evaluated.RESULTS. Consistent with the numeric simulation, using 2D SENSE in contrast-enhanced coronary MRA resulted in significant improvement in aortic blood-pool SNR (unenhanced vs contrast-enhanced, 37.5 +/- 14.7 vs 121.3 +/- 44.0; p < 0.05) and CNR (14.4 +/- 6.9 vs 101.5 +/- 40.8; p < 0.05) in the patient sample. A longer length of left anterior descending coronary artery was visualized, but vessel sharpness, coronary artery coverage, and image quality score were not improved with the proposed approach.CONCLUSION. In combination with contrast administration, 2D SENSE was found effective in improving SNR and CNR in 3-T whole-heart coronary MRA. Further investigation of cardiac motion compensation is necessary to exploit the SNR and CNR advantages and to achieve submillimeter spatial resolution.

Quality assessment of cardiovascular magnetic resonance in the setting of the European CMR registry: description and validation of standardized criteria.

BACKGROUND: Cardiovascular magnetic resonance (CMR) has become an important diagnostic imaging modality in cardiovascular medicine. However, insufficient image quality may compromise its diagnostic accuracy. We aimed to describe and validate standardized criteria to evaluate a) cine steady-state free precession (SSFP), b) late gadolinium enhancement (LGE), and c) stress first-pass perfusion images. These criteria will serve for quality assessment in the setting of the Euro-CMR registry. METHODS: Thirty-five qualitative criteria were defined (scores 0-3) with lower scores indicating better image quality. In addition, quantitative parameters were measured yielding 2 additional quality criteria, i.e. signal-to-noise ratio (SNR) of non-infarcted myocardium (as a measure of correct signal nulling of healthy myocardium) for LGE and % signal increase during contrast medium first-pass for perfusion images. These qualitative and quantitative criteria were assessed in a total of 90 patients (60 patients scanned at our own institution at 1.5T (n=30) and 3T (n=30) and in 30 patients randomly chosen from the Euro-CMR registry examined at 1.5T). Analyses were performed by 2 SCMR level-3 experts, 1 trained study nurse, and 1 trained medical student. RESULTS: The global quality score was 6.7±4.6 (n=90, mean of 4 observers, maximum possible score 64), range 6.4-6.9 (p=0.76 between observers). It ranged from 4.0-4.3 for 1.5T (p=0.96 between observers), from 5.9-6.9 for 3T (p=0.33 between observers), and from 8.6-10.3 for the Euro-CMR cases (p=0.40 between observers). The inter- (n=4) and intra-observer (n=2) agreement for the global quality score, i.e. the percentage of assignments to the same quality tertile ranged from 80% to 88% and from 90% to 98%, respectively. The agreement for the quantitative assessment for LGE images (scores 0-2 for SNR <2, 2-5, >5, respectively) ranged from 78-84% for the entire population, and 70-93% at 1.5T, 64-88% at 3T, and 72-90% for the Euro-CMR cases. The agreement for perfusion images (scores 0-2 for %SI increase >200%, 100%-200%,<100%, respectively) ranged from 81-91% for the entire population, and 76-100% at 1.5T, 67-96% at 3T, and 62-90% for the Euro-CMR registry cases. The intra-class correlation coefficient for the global quality score was 0.83. CONCLUSIONS: The described criteria for the assessment of CMR image quality are robust with a good inter- and intra-observer agreement. Further research is needed to define the impact of image quality on the diagnostic and prognostic yield of CMR studies.

Quantitative analysis of crystal/grain sizes and their distributions in 2D and 3D

We review methods to estimate the average crystal (grain) size and the crystal (grain) size distribution in solid rocks. Average grain sizes often provide the base for stress estimates or rheological calculations requiring the quantification of grain sizes in a rock's microstructure. The primary data for grain size data are either 1D (i.e. line intercept methods), 2D (area analysis) or 3D (e.g., computed tomography, serial sectioning). These data have been used for different data treatments over the years, whereas several studies assume a certain probability function (e.g., logarithm, square root) to calculate statistical parameters as the mean, median, mode or the skewness of a crystal size distribution. The finally calculated average grain sizes have to be compatible between the different grain size estimation approaches in order to be properly applied, for example, in paleo-piezometers or grain size sensitive flow laws. Such compatibility is tested for different data treatments using one- and two-dimensional measurements. We propose an empirical conversion matrix for different datasets. These conversion factors provide the option to make different datasets compatible with each other, although the primary calculations were obtained in different ways. In order to present an average grain size, we propose to use the area-weighted and volume-weighted mean in the case of unimodal grain size distributions, respectively, for 2D and 3D measurements. The shape of the crystal size distribution is important for studies of nucleation and growth of minerals. The shape of the crystal size distribution of garnet populations is compared between different 2D and 3D measurements, which are serial sectioning and computed tomography. The comparison of different direct measured 3D data; stereological data and direct presented 20 data show the problems of the quality of the smallest grain sizes and the overestimation of small grain sizes in stereological tools, depending on the type of CSD. (C) 2011 Published by Elsevier Ltd.

Prognostication in comatose survivors of cardiac arrest: An advisory statement from the European Resuscitation Council and the European Society of Intensive Care Medicine.

OBJECTIVES: To review and update the evidence on predictors of poor outcome (death, persistent vegetative state or severe neurological disability) in adult comatose survivors of cardiac arrest, either treated or not treated with controlled temperature, to identify knowledge gaps and to suggest a reliable prognostication strategy. METHODS: GRADE-based systematic review followed by expert consensus achieved using Web-based Delphi methodology, conference calls and face-to-face meetings. Predictors based on clinical examination, electrophysiology, biomarkers and imaging were included. RESULTS AND CONCLUSIONS: Evidence from a total of 73 studies was reviewed. The quality of evidence was low or very low for almost all studies. In patients who are comatose with absent or extensor motor response at ≥72 h from arrest, either treated or not treated with controlled temperature, bilateral absence of either pupillary and corneal reflexes or N20 wave of short-latency somatosensory evoked potentials were identified as the most robust predictors. Early status myoclonus, elevated values of neuron-specific enolase at 48-72 h from arrest, unreactive malignant EEG patterns after rewarming, and presence of diffuse signs of postanoxic injury on either computed tomography or magnetic resonance imaging were identified as useful but less robust predictors. Prolonged observation and repeated assessments should be considered when results of initial assessment are inconclusive. Although no specific combination of predictors is sufficiently supported by available evidence, a multimodal prognostication approach is recommended in all patients.