Abstract
Background and study aims
Detecting colorectal neoplasia is the goal of high-quality screening and surveillance colonoscopy, as reflected by high adenoma detection rate (ADR) and adenomas ...per colonoscopy (APC). The aim of our study was to evaluate the performance of a novel artificial intelligence (AI)-aided polyp detection device, Skout, with the primary endpoints of ADR and APC in routine colonoscopy.
Patients and methods
We compared ADR and APC in a cohort of outpatients undergoing routine high-resolution colonoscopy with and without the use of a real-time, AI-aided polyp detection device. Patients undergoing colonoscopy with Skout were enrolled in a single-arm, unblinded, prospective trial and the results were compared with a historical cohort. All resected polyps were examined histologically.
Results
Eighty-three patients undergoing screening and surveillance colonoscopy at an outpatient endoscopy center were enrolled and outcomes compared with 283 historical control patients. Overall, ADR with and without Skout was 54.2 % and 40.6 % respectively (
P
= 0.028) and 53.6 % and 30.8 %, respectively, in screening exams (
P
= 0.024). Overall, APC rate with and without Skout was 1.46 and 1.01, respectively, (
P
= 0.104) and 1.18 and 0.50, respectively, in screening exams (
P
= 0.002). Overall, true histology rate (THR) with and without Skout was 73.8 % and 78.4 %, respectively, (
P
= 0.463) and 75.0 % and 71.0 %, respectively, in screening exams (
P
= 0.731).
Conclusion
We have demonstrated that our novel AI-aided polyp detection device increased the ADR in a cohort of patients undergoing screening and surveillance colonoscopy without a significant concomitant increase in hyperplastic polyp resection. AI-aided colonoscopy has the potential for improving the outcomes of patients undergoing colonoscopy.
Background Timely imaging is essential for patients undergoing mechanical thrombectomy (MT). Our objective was to evaluate the safety and feasibility of low‐field portable magnetic resonance imaging ...(pMRI) for bedside evaluation following MT. Methods Patients with suspected large‐vessel occlusion undergoing MT were screened for eligibility. All pMRI examinations were conducted in the standard ferromagnetic environment of the interventional radiology suite. Clinical characteristics, procedural details, and pMRI features were collected. Subsequent high‐field conventional MRI within 72±12 hours was analyzed. If a conventional MRI was not available for comparison, computed tomography within the same time frame was used for validation. Results Twenty‐four patients were included (63% women; median age, 76 years interquartile range, 69–84 years). MT was performed with a median access to revascularization time of 15 minutes (interquartile range, 8–19 minutes), and with a successful outcome as defined by a thrombolysis in cerebral infarction score of ≥2B in 90% of patients. The median time from the end of the procedure to pMRI was 22 minutes (interquartile range, 16–32 minutes). The median pMRI examination time was 30 minutes (interquartile range, 17–33 minutes). Of 23 patients with available subsequent imaging, 9 had infarct progression compared with immediate post‐MT pMRI and 14 patients did not have progression of their infarct volume. There was no adverse event related to the examination. Conclusion Low‐field pMRI is safe and feasible in a post‐MT environment and enables timely identification of ischemic changes in the interventional radiology suite. This approach can facilitate the assessment of baseline infarct burden and may help guide physiological interventions following MT.
Automatic whole-brain extraction from magnetic resonance images (MRI), also known as skull stripping, is a key component in most neuroimage pipelines. As the first element in the chain, its ...robustness is critical for the overall performance of the system. Many skull stripping methods have been proposed, but the problem is not considered to be completely solved yet. Many systems in the literature have good performance on certain datasets (mostly the datasets they were trained/tuned on), but fail to produce satisfactory results when the acquisition conditions or study populations are different. In this paper we introduce a robust, learning-based brain extraction system (ROBEX). The method combines a discriminative and a generative model to achieve the final result. The discriminative model is a Random Forest classifier trained to detect the brain boundary; the generative model is a point distribution model that ensures that the result is plausible. When a new image is presented to the system, the generative model is explored to find the contour with highest likelihood according to the discriminative model. Because the target shape is in general not perfectly represented by the generative model, the contour is refined using graph cuts to obtain the final segmentation. Both models were trained using 92 scans from a proprietary dataset but they achieve a high degree of robustness on a variety of other datasets. ROBEX was compared with six other popular, publicly available methods (BET, BSE, FreeSurfer, AFNI, BridgeBurner, and GCUT) on three publicly available datasets (IBSR, LPBA40, and OASIS, 137 scans in total) that include a wide range of acquisition hardware and a highly variable population (different age groups, healthy/diseased). The results show that ROBEX provides significantly improved performance measures for almost every method/dataset combination.
The hippocampal formation is a complex, heterogeneous structure that consists of a number of distinct, interacting subregions. Atrophy of these subregions is implied in a variety of neurodegenerative ...diseases, most prominently in Alzheimer's disease (AD). Thanks to the increasing resolution of MR images and computational atlases, automatic segmentation of hippocampal subregions is becoming feasible in MRI scans. Here we introduce a generative model for dedicated longitudinal segmentation that relies on subject-specific atlases. The segmentations of the scans at the different time points are jointly computed using Bayesian inference. All time points are treated the same to avoid processing bias. We evaluate this approach using over 4700 scans from two publicly available datasets (ADNI and MIRIAD). In test–retest reliability experiments, the proposed method yielded significantly lower volume differences and significantly higher Dice overlaps than the cross-sectional approach for nearly every subregion (average across subregions: 4.5% vs. 6.5%, Dice overlap: 81.8% vs. 75.4%). The longitudinal algorithm also demonstrated increased sensitivity to group differences: in MIRIAD (69 subjects: 46 with AD and 23 controls), it found differences in atrophy rates between AD and controls that the cross sectional method could not detect in a number of subregions: right parasubiculum, left and right presubiculum, right subiculum, left dentate gyrus, left CA4, left HATA and right tail. In ADNI (836 subjects: 369 with AD, 215 with early cognitive impairment — eMCI — and 252 controls), all methods found significant differences between AD and controls, but the proposed longitudinal algorithm detected differences between controls and eMCI and differences between eMCI and AD that the cross sectional method could not find: left presubiculum, right subiculum, left and right parasubiculum, left and right HATA. Moreover, many of the differences that the cross-sectional method already found were detected with higher significance. The presented algorithm will be made available as part of the open-source neuroimaging package FreeSurfer.
•A segmentation method for the hippocampal substructures in longitudinal MRI scans•Increased test–retest reliability compared with cross-sectional analysis•Increased power to detect group differences in atrophy rates in LME framework•Algorithm will be made publicly available as part of FreeSurfer
Brainstem segmentation has been useful in identifying potential imaging biomarkers for diagnosis and progression in atypical parkinsonian syndromes (APS). However, the majority of work has been ...performed using manual segmentation, which is time consuming for large cohorts.
We investigated brainstem involvement in APS using an automated method. We measured the volume of the medulla, pons, superior cerebellar peduncle (SCP) and midbrain from T1-weighted MRIs in 67 patients and 42 controls. Diagnoses were corticobasal syndrome (CBS, n = 14), multiple system atrophy (MSA, n = 16: 8 with parkinsonian syndrome, MSA-P; 8 with cerebellar syndrome, MSA-C), progressive supranuclear palsy with a Richardson's syndrome (PSP-RS, n = 12), variant PSP (n = 18), and APS not otherwise specified (APS-NOS, n = 7).
All brainstem regions were smaller in MSA-C (19-42% volume difference, p < 0.0005) and in both PSP groups (18-33%, p < 0.0005) than in controls. MSA-P showed lower volumes in all regions except the SCP (15-26%, p < 0.0005). The most affected region in MSA-C and MSA-P was the pons (42% and 26%, respectively), while the most affected regions in both the PSP-RS and variant PSP groups were the SCP (33% and 23%, respectively) and midbrain (26% and 24%, respectively). The brainstem was less affected in CBS, but nonetheless, the pons (14%, p < 0.0005), midbrain (14%, p < 0.0005) and medulla (10%, p = 0.001) were significantly smaller in CBS than in controls. The brainstem was unaffected in APS-NOS.
Automated methods can accurately quantify the involvement of brainstem structures in APS. This will be important in future trials with large patient numbers where manual segmentation is unfeasible.
We introduce a strategy for learning image registration without acquired imaging data, producing powerful networks agnostic to contrast introduced by magnetic resonance imaging (MRI). While classical ...registration methods accurately estimate the spatial correspondence between images, they solve an optimization problem for every new image pair. Learning-based techniques are fast at test time but limited to registering images with contrasts and geometric content similar to those seen during training. We propose to remove this dependency on training data by leveraging a generative strategy for diverse synthetic label maps and images that exposes networks to a wide range of variability, forcing them to learn more invariant features. This approach results in powerful networks that accurately generalize to a broad array of MRI contrasts. We present extensive experiments with a focus on 3D neuroimaging, showing that this strategy enables robust and accurate registration of arbitrary MRI contrasts even if the target contrast is not seen by the networks during training. We demonstrate registration accuracy surpassing the state of the art both within and across contrasts, using a single model. Critically, training on arbitrary shapes synthesized from noise distributions results in competitive performance, removing the dependency on acquired data of any kind. Additionally, since anatomical label maps are often available for the anatomy of interest, we show that synthesizing images from these dramatically boosts performance, while still avoiding the need for real intensity images. Our code is available at doic https://w3id.org/synthmorph .
Despite advances in data augmentation and transfer learning, convolutional neural networks (CNNs) difficultly generalise to unseen domains. When segmenting brain scans, CNNs are highly sensitive to ...changes in resolution and contrast: even within the same MRI modality, performance can decrease across datasets. Here we introduce SynthSeg, the first segmentation CNN robust against changes in contrast and resolution. SynthSeg is trained with synthetic data sampled from a generative model conditioned on segmentations. Crucially, we adopt a domain randomisation strategy where we fully randomise the contrast and resolution of the synthetic training data. Consequently, SynthSeg can segment real scans from a wide range of target domains without retraining or fine-tuning, which enables straightforward analysis of huge amounts of heterogeneous clinical data. Because SynthSeg only requires segmentations to be trained (no images), it can learn from labels obtained by automated methods on diverse populations (e.g., ageing and diseased), thus achieving robustness to a wide range of morphological variability. We demonstrate SynthSeg on 5,000 scans of six modalities (including CT) and ten resolutions, where it exhibits unparallelled generalisation compared with supervised CNNs, state-of-the-art domain adaptation, and Bayesian segmentation. Finally, we demonstrate the generalisability of SynthSeg by applying it to cardiac MRI and CT scans.
Display omitted
•A CNN to segment brain MRI scans of any contrast and resolution without retraining.•Domain-independence is achieved by training on randomised unrealistic synthetic data.•SynthSeg sustains almost the accuracy of supervised CNNs across all tested domains.•It also outperforms state-of-the-art domain adaptation, without being retrained.•The model is implemented in FreeSurfer for easy distribution and deployment.
Every year, millions of brain MRI scans are acquired in hospitals, which is a figure considerably larger than the size of any research dataset. Therefore, the ability to analyze such scans could ...transform neuroimaging research. Yet, their potential remains untapped since no automated algorithm is robust enough to cope with the high variability in clinical acquisitions (MR contrasts, resolutions, orientations, artifacts, and subject populations). Here, we present
, an AI segmentation suite that enables robust analysis of heterogeneous clinical datasets. In addition to whole-brain segmentation,
also performs cortical parcellation, intracranial volume estimation, and automated detection of faulty segmentations (mainly caused by scans of very low quality). We demonstrate
in seven experiments, including an aging study on 14,000 scans, where it accurately replicates atrophy patterns observed on data of much higher quality.
is publicly released as a ready-to-use tool to unlock the potential of quantitative morphometry.
Reconstructing 3D MR volumes from multiple motion-corrupted stacks of 2D slices has shown promise in imaging of moving subjects, e.g ., fetal MRI. However, existing slice-to-volume reconstruction ...methods are time-consuming, especially when a high-resolution volume is desired. Moreover, they are still vulnerable to severe subject motion and when image artifacts are present in acquired slices. In this work, we present NeSVoR, a resolution-agnostic slice-to-volume reconstruction method, which models the underlying volume as a continuous function of spatial coordinates with implicit neural representation. To improve robustness to subject motion and other image artifacts, we adopt a continuous and comprehensive slice acquisition model that takes into account rigid inter-slice motion, point spread function, and bias fields. NeSVoR also estimates pixel-wise and slice-wise variances of image noise and enables removal of outliers during reconstruction and visualization of uncertainty. Extensive experiments are performed on both simulated and in vivo data to evaluate the proposed method. Results show that NeSVoR achieves state-of-the-art reconstruction quality while providing two to ten-fold acceleration in reconstruction times over the state-of-the-art algorithms.
Subcortical structures play a critical role in brain function. However, options for assessing electrophysiological activity in these structures are limited. Electromagnetic fields generated by ...neuronal activity in subcortical structures can be recorded noninvasively, using magnetoencephalography (MEG) and electroencephalography (EEG). However, these subcortical signals are much weaker than those generated by cortical activity. In addition, we show here that it is difficult to resolve subcortical sources because distributed cortical activity can explain the MEG and EEG patterns generated by deep sources. We then demonstrate that if the cortical activity is spatially sparse, both cortical and subcortical sources can be resolved with M/EEG. Building on this insight, we develop a hierarchical sparse inverse solution for M/EEG. We assess the performance of this algorithm on realistic simulations and auditory evoked response data, and show that thalamic and brainstem sources can be correctly estimated in the presence of cortical activity. Our work provides alternative perspectives and tools for characterizing electrophysiological activity in subcortical structures in the human brain.