Frontotemporal dementia (FTD) is a heterogeneous neurodegenerative disorder, with a strong genetic component. Previous research has shown that medial temporal lobe atrophy is a common feature of FTD. ...However, no study has so far investigated the differential vulnerability of the hippocampal subfields in FTD.
We aimed to investigate hippocampal subfield volumes in genetic FTD.
We in6/2/2018vestigated hippocampal subfield volumes in a cohort of 75 patients with genetic FTD (age: mean (standard deviation) 59.3 (7.7) years; disease duration: 5.1 (3.4) years; 29 with MAPT, 28 with C9orf72, and 18 with GRN mutations) compared with 97 age-matched controls (age: 62.1 (11.1) years). We performed a segmentation of their volumetric T1-weighted MRI scans to extract hippocampal subfields volumes. Left and right volumes were summed and corrected for total intracranial volumes.
All three groups had smaller hippocampi than controls. The MAPT group had the most atrophic hippocampi, with the subfields showing the largest difference from controls being CA1-4 (24-27%, p < 0.0005). For C9orf72, the CA4, CA1, and dentate gyrus regions (8-11%, p < 0.0005), and for GRN the presubiculum and subiculum (10-14%, p < 0.0005) showed the largest differences from controls.
The hippocampus was affected in all mutation types but a different pattern of subfield involvement was found in the three genetic groups, consistent with differential cortical-subcortical network vulnerability.
Display omitted
•A label fusion framework based on a generative model that works across modalities.•The registrations are not precomputed, but estimated during the fusion.•The registrations are ...explicitly linked in the generative model.
Multi-atlas label fusion is a powerful image segmentation strategy that is becoming increasingly popular in medical imaging. A standard label fusion algorithm relies on independently computed pairwise registrations between individual atlases and the (target) image to be segmented. These registrations are then used to propagate the atlas labels to the target space and fuse them into a single final segmentation. Such label fusion schemes commonly rely on the similarity between intensity values of the atlases and target scan, which is often problematic in medical imaging – in particular, when the atlases and target images are obtained via different sensor types or imaging protocols.
In this paper, we present a generative probabilistic model that yields an algorithm for solving the atlas-to-target registrations and label fusion steps simultaneously. The proposed model does not directly rely on the similarity of image intensities. Instead, it exploits the consistency of voxel intensities within the target scan to drive the registration and label fusion, hence the atlases and target image can be of different modalities. Furthermore, the framework models the joint warp of all the atlases, introducing interdependence between the registrations.
We use variational expectation maximization and the Demons registration framework in order to efficiently identify the most probable segmentation and registrations. We use two sets of experiments to illustrate the approach, where proton density (PD) MRI atlases are used to segment T1-weighted brain scans and vice versa. Our results clearly demonstrate the accuracy gain due to exploiting within-target intensity consistency and integrating registration into label fusion.
Advances in artificial intelligence have cultivated a strong interest in developing and validating the clinical utilities of computer-aided diagnostic models. Machine learning for diagnostic ...neuroimaging has often been applied to detect psychological and neurological disorders, typically on small-scale datasets or data collected in a research setting. With the collection and collation of an ever-growing number of public datasets that researchers can freely access, much work has been done in adapting machine learning models to classify these neuroimages by diseases such as Alzheimer's, ADHD, autism, bipolar disorder, and so on. These studies often come with the promise of being implemented clinically, but despite intense interest in this topic in the laboratory, limited progress has been made in clinical implementation. In this review, we analyze challenges specific to the clinical implementation of diagnostic AI models for neuroimaging data, looking at the differences between laboratory and clinical settings, the inherent limitations of diagnostic AI, and the different incentives and skill sets between research institutions, technology companies, and hospitals. These complexities need to be recognized in the translation of diagnostic AI for neuroimaging from the laboratory to the clinic.
Neuroimaging is essential for detecting spontaneous, nontraumatic intracerebral hemorrhage (ICH). Recent data suggest ICH can be characterized using low-field magnetic resonance imaging (MRI). Our ...primary objective was to investigate the sensitivity and specificity of ICH on a 0.064T portable MRI (pMRI) scanner using a methodology that provided clinical information to inform rater interpretations. As a secondary aim, we investigated whether the incorporation of a deep learning (DL) reconstruction algorithm affected ICH detection.
The pMRI device was deployed at Yale New Haven Hospital to examine patients presenting with stroke symptoms from October 26, 2020 to February 21, 2022. Three raters independently evaluated pMRI examinations. Raters were provided the images alongside the patient's clinical information to simulate real-world context of use. Ground truth was the closest conventional computed tomography or 1.5/3T MRI. Sensitivity and specificity results were grouped by DL and non-DL software to investigate the effects of software advances.
A total of 189 exams (38 ICH, 89 acute ischemic stroke, 8 subarachnoid hemorrhage, 3 primary intraventricular hemorrhage, 51 no intracranial abnormality) were evaluated. Exams were correctly classified as positive or negative for ICH in 185 of 189 cases (97.9% overall accuracy). ICH was correctly detected in 35 of 38 cases (92.1% sensitivity). Ischemic stroke and no intracranial abnormality cases were correctly identified as blood-negative in 139 of 140 cases (99.3% specificity). Non-DL scans had a sensitivity and specificity for ICH of 77.8% and 97.1%, respectively. DL scans had a sensitivity and specificity for ICH of 96.6% and 99.3%, respectively.
These results demonstrate improvements in ICH detection accuracy on pMRI that may be attributed to the integration of clinical information in rater review and the incorporation of a DL-based algorithm. The use of pMRI holds promise in providing diagnostic neuroimaging for patients with ICH.
In the 20th century, the advent of neuroimaging dramatically altered the field of neurologic care. However, despite iterative advances since the invention of CT and MRI, little progress has been made ...to bring MR neuroimaging to the point of care. Recently, the emergence of a low-field (<1 T) portable MRI (pMRI) is setting the stage to revolutionize the landscape of accessible neuroimaging. Users can transport the pMRI into a variety of locations, using a standard 110-220 V wall outlet. In this article, we discuss current applications for pMRI, including in the acute and critical care settings, the barriers to broad implementation, and future opportunities.
Frontotemporal dementia (FTD) is a heterogeneous neurodegenerative disorder with multiple genetic and pathological causes. It is characterized by both cortical and subcortical atrophies, with ...previous studies showing early involvement of the amygdala. However, no prior study has specifically investigated the atrophy of different subnuclei of the amygdala.
Using an automated segmentation tool for T1-weighted volumetric magnetic resonance imaging, we investigated amygdalar subnuclei (AS) involvement in a cohort of 132 patients with genetic or pathologically confirmed FTD (age: mean = 61 years (standard deviation = 8); disease duration: 5 (3) years) compared with 107 age-matched controls.
AS were affected in all genetic and pathological forms of FTD. MAPT mutations/FTDP-17, Pick's disease, and transactive response DNA binding protein 43 kDa type C were the forms with the smallest amygdala (35%–50% smaller than controls in the most affected hemisphere, P < .0005). In most FTD groups, medial subnuclei (particularly the superficial, accessory basal and basal/paralaminar subnuclei) tended to be affected more than the lateral subnuclei, except for the progressive supranuclear palsy group, in which the corticoamygdaloid transition area was the least-affected area.
Differential involvement of the AS was seen in the different genetic and pathological forms of FTD. In general, the most affected subnuclei were the superficial, accessory basal and basal/paralaminar subnuclei, which form part of a network of regions that control reward and emotion regulation, functions known to be particularly affected in FTD.
•Amygdalar subnuclei are affected in all genetic and pathological forms of frontotemporal dementia.•Amygdalar subnuclei are differently involved among genetic and pathological forms.•MAPT/FTDP-17, Pick's disease, and TDP-43 type C showed the smallest amygdala.•The medial subnuclei tended to be affected more than the lateral subnuclei.
Semantic variant of primary progressive aphasia (svPPA) is a subtype of frontotemporal dementia characterized by asymmetric temporal atrophy.
We investigated the pattern of medial temporal lobe ...atrophy in 24 svPPA patients compared to 72 controls using novel approaches to segment the hippocampal and amygdalar subregions on MRIs. Based on semantic knowledge scores, we split the svPPA group into 3 subgroups of early, middle and late disease stage.
Early stage: all left amygdalar and hippocampal subregions (except the tail) were affected in svPPA (21-35% smaller than controls), together with the following amygdalar nuclei in the right hemisphere: lateral, accessory basal and superficial (15-23%). On the right, only the temporal pole was affected among the cortical regions. Middle stage: the left hippocampal tail became affected (28%), together with the other amygdalar nuclei (22-26%), and CA4 (15%) on the right, with orbitofrontal cortex and subcortical structures involvement on the left, and more posterior temporal lobe on the right. Late stage: the remaining right hippocampal regions (except the tail) (19-24%) became affected, with more posterior left cortical and right extra-temporal anterior cortical involvement.
With advanced subregions segmentation, it is possible to detect early involvement of the right medial temporal lobe in svPPA that is not detectable by measuring the amygdala or hippocampus as a whole.
In this paper we present a novel label fusion algorithm suited for scenarios in which different manual delineation protocols with potentially disparate structures have been used to annotate the ...training scans (hereafter referred to as “atlases”). Such scenarios arise when atlases have missing structures, when they have been labeled with different levels of detail, or when they have been taken from different heterogeneous databases. The proposed algorithm can be used to automatically label a novel scan with any of the protocols from the training data. Further, it enables us to generate new labels that are not present in any delineation protocol by defining intersections on the underling labels. We first use probabilistic models of label fusion to generalize three popular label fusion techniques to the multi-protocol setting: majority voting, semi-locally weighted voting and STAPLE. Then, we identify some shortcomings of the generalized methods, namely the inability to produce meaningful posterior probabilities for the different labels (majority voting, semi-locally weighted voting) and to exploit the similarities between the atlases (all three methods). Finally, we propose a novel generative label fusion model that can overcome these drawbacks. We use the proposed method to combine four brain MRI datasets labeled with different protocols (with a total of 102 unique labeled structures) to produce segmentations of 148 brain regions. Using cross-validation, we show that the proposed algorithm outperforms the generalizations of majority voting, semi-locally weighted voting and STAPLE (mean Dice score 83%, vs. 77%, 80% and 79%, respectively). We also evaluated the proposed algorithm in an aging study, successfully reproducing some well-known results in cortical and subcortical structures.
Display omitted
PDF
