Explore the vast range of titles available.

We hypothesized that machine learning analysis based on texture information from the preoperative MRI can predict IDH mutational status in newly diagnosed WHO grade II and III gliomas. This retrospective study included in total 79 consecutive patients with a newly diagnosed WHO grade II or III glioma. Local binary pattern texture features were generated from preoperative B0 and fractional anisotropy (FA) diffusion tensor imaging. Using a training set of 59 patients, a single hidden layer neural network was then trained on the texture features to predict IDH status. The model was validated based on the prediction accuracy calculated in a previously unseen set of 20 gliomas. Prediction accuracy of the generated model was 92% (54/59 cases; AUC = 0.921) in the training and 95% (19/20; AUC = 0.952) in the validation cohort. The ten most important features were comprised of tumor size and both B0 and FA texture information, underlining the joint contribution of imaging data to classification. Machine learning analysis of DTI texture information and tumor size reliably predicts IDH status in preoperative MRI of gliomas. Such information may increasingly support individualized surgical strategies, supplement pathological analysis and highlight the potential of radiogenomics.

ObjectivesTo compare opportunistic quantitative CT (QCT) with dual energy X-ray absorptiometry (DXA) in their ability to predict incident vertebral fractures.MethodsWe included 84 patients aged 50 years and older, who had routine CT including the lumbar spine and DXA within a 12-month period (baseline) as well as follow-up imaging after at least 12 months or who sustained an incident vertebral fracture documented earlier. Patients with bone disorders aside from osteoporosis were excluded. Fracture status and trabecular bone mineral density (BMD) were retrospectively evaluated in baseline CT and fracture status was reassessed at follow-up. BMDQCT was assessed by opportunistic QCT with asynchronous calibration of multiple MDCT scanners.ResultsSixteen patients had incident vertebral fractures showing lower mean BMDQCT than patients without fracture (p = 0.001). For the risk of incident vertebral fractures, the hazard ratio increased per SD in BMDQCT (4.07; 95% CI, 1.98–8.38), as well as after adjusting for age, sex, and prevalent fractures (2.54; 95% CI, 1.09–5.90). For DXA, a statistically significant increase in relative hazard per SD decrease in T-score was only observed after age and sex adjustment (1.57; 95% CI, 1.04–2.38). The predictability of incident vertebral fractures was good by BMDQCT (AUC = 0.76; 95% CI, 0.64–0.89) and non-significant by T-scores. Asynchronously calibrated CT scanners showed good long-term stability (linear drift ranging from − 0.55 to − 2.29 HU per year).ConclusionsOpportunistic screening of mainly neurosurgical and oncologic patients in CT performed for indications other than densitometry allows for better risk assessment of imminent vertebral fractures than dedicated DXA.Key Points• Opportunistic QCT predicts osteoporotic vertebral fractures better than DXA reference standard in mainly neurosurgical and oncologic patients.• More than every second patient (56%) with an incident vertebral fracture was misdiagnosed not having osteoporosis according to DXA.• Standard ACR QCT-cutoff values for osteoporosis (< 80 mg/cm3) and osteopenia (≤ 120 mg/cm3) can also be applied scanner independently in calibrated opportunistic QCT.

This paper develops a novel mathematical framework for collaborative learning by means of geometrically inspired kernel machines which includes statements on the bounds of generalisation and approximation errors, and sample complexity. For classification problems, this approach allows us to learn bounded geometric structures around given data points and hence solve the global model learning problem in an efficient way by exploiting convexity properties of the related optimisation problem in a Reproducing Kernel Hilbert Space (RKHS). In this way, we can reduce classification problems to determining the closest bounded geometric structure from a given data point. Further advantages that come with our solution is that our approach does not require clients to perform multiple epochs of local optimisation using stochastic gradient descent, nor require rounds of communication between client/server for optimising the global model. We highlight that numerous experiments have shown that the proposed method is a competitive alternative to the state-of-the-art.

Background: Osteoporosis is a major risk factor for procedure related complications in neurosurgical operations at the spine [1]. With statistical parametric mapping (SPM) regional bone loss at the thoracolumbar spine can be analyzed using routine CT data [2]. The aim of this study was to compare regional trabecular bone mineral density (BMD) between patients with normal, low, and osteoporotic bone mass as defined by dual energy X-ray absorptiometry (DXA). Methods: In this retrospective study, 252 patients with lumbar DXA and a CT scan within 12 months were included. Clinical CT scans were performed for reasons other than densitometry on four different CT scanners each asynchronously calibrated to yield volumetric BMD. SPM of thoracolumbar vertebral bodies were calculated by registration, normalization and voxel-wise statistics (gender-adjusted ANOVA and t-test, p = 0.05 corrected for multiple testing). Patients with spinal metastasis and vertebrae with compression fractures or degenerative sclerosis were excluded from the analysis. Result: According to DXA, 84 patients had osteoporosis, 91 had low bone mass, and 77 were healthy with normal bone mass. In healthy patients, mean BMD was lowest at L3 with 129.0 [+ or -] 54.3 mg/[cm.sup.3] and highest at T1 with 208.7 [+ or -] 54.8 mg/[cm.sup.3] (Fig. 1). Compared to healthy patients, statistically significant decreases in BMD were observed in osteoporotic patients at all vertebral levels with an emphasis on the upper thoracic (minimum at T4 with T-score=-3.2 and 91% affected volume) and on the thoracolumbar junction/upper lumbar spine (minimum at L2 with T-score = -0.94 and 83% affected volume; Fig. 2, Fig. 3). Similar trends were observed for patients with low bone mass. Discussion: SPM revealed reduced trabecular BMD in osteoporotic patients pronounced at the upper thoracic spine and thoracolumbar junction. The mid-thoracic spine (T6-T10) with the least overall motion range due to ribcage stiffness seems to be less prone to osteoporosis-related deterioration of trabecular BMD. Conclusion: SPM can quantify and visualize regional bone loss at the spine for opportunistic osteoporosis screening and may impact neurosurgical planning.

Alzheimers disease (AD) is the most common cause of dementia. Although the underlying pathology is still not completely understood, several diagnostic methods are available. Frequently, the most accurate methods are also the most invasive. The present work investigates the diagnostic potential of Pulsed Arterial Spin Labeling (PASL) for AD: a non-invasive, MRI-based technique for the quantification of regional cerebral blood flow (rCBF). In particular, we propose a pilot computer aided diagnostic (CAD) procedure able to discriminate between healthy and diseased subjects, and at the same time, providing visual informative results. This method encompasses the creation of a healthy model, the computation of a voxel-wise likelihood function as comparison between the healthy model and the subject under examination, and the correction of the likelihood function via prior distributions. The discriminant analysis is carried out to maximize the accuracy of the classification. The algorithm has been trained on a dataset of 81 subjects and achieved a sensitivity of 0.750 and a specificity of 0.875. Moreover, in accordance with the current pathological knowledge, the parietal lobe, and limbic system are shown to be the main discriminant factors.

This paper develops a novel mathematical framework for collaborative learning by means of geometrically inspired kernel machines which includes statements on the bounds of generalisation and approximation errors, and sample complexity. For classification problems, this approach allows us to learn bounded geometric structures around given data points and hence solve the global model learning problem in an efficient way by exploiting convexity properties of the related optimisation problem in a Reproducing Kernel Hilbert Space (RKHS). In this way, we can reduce classification problems to determining the closest bounded geometric structure from a given data point. Further advantages that come with our solution is that our approach does not require clients to perform multiple epochs of local optimisation using stochastic gradient descent, nor require rounds of communication between client/server for optimising the global model. We highlight that numerous experiments have shown that the proposed method is a competitive alternative to the state-of-the-art.

Federated learning must address heterogeneity, strict communication and computation limits, and privacy while ensuring performance. We propose an operator-theoretic framework that maps the \\(L^2\\)-optimal solution into a reproducing kernel Hilbert space (RKHS) via a forward operator, approximates it using available data, and maps back with the inverse operator, yielding a gradient-free scheme. Finite-sample bounds are derived using concentration inequalities over operator norms, and the framework identifies a data-dependent hypothesis space with guarantees on risk, error, robustness, and approximation. Within this space we design efficient kernel machines leveraging the space folding property of Kernel Affine Hull Machines. Clients transfer knowledge via a scalar space folding measure, reducing communication and enabling a simple differentially private protocol: summaries are computed from noise-perturbed data matrices in one step, avoiding per-round clipping and privacy accounting. The induced global rule requires only integer minimum and equality-comparison operations per test point, making it compatible with fully homomorphic encryption (FHE). Across four benchmarks, the gradient-free FL method with fixed encoder embeddings matches or outperforms strong gradient-based fine-tuning, with gains up to 23.7 points. In differentially private experiments, kernel smoothing mitigates accuracy loss in high-privacy regimes. The global rule admits an FHE realization using \\(Q C\\) encrypted minimum and \\(C\\) equality-comparison operations per test point, with operation-level benchmarks showing practical latencies. Overall, the framework provides provable guarantees with low communication, supports private knowledge transfer via scalar summaries, and yields an FHE-compatible prediction rule offering a mathematically grounded alternative to gradient-based federated learning under heterogeneity.

This paper develops a novel mathematical framework for collaborative learning by means of geometrically inspired kernel machines which includes statements on the bounds of generalisation and approximation errors, and sample complexity. For classification problems, this approach allows us to learn bounded geometric structures around given data points and hence solve the global model learning problem in an efficient way by exploiting convexity properties of the related optimisation problem in a Reproducing Kernel Hilbert Space (RKHS). In this way, we can reduce classification problems to determining the closest bounded geometric structure from a given data point. Further advantages that come with our solution is that our approach does not require clients to perform multiple epochs of local optimisation using stochastic gradient descent, nor require rounds of communication between client/server for optimising the global model. We highlight that numerous experiments have shown that the proposed method is a competitive alternative to the state-of-the-art.

Vertebral labelling and segmentation are two fundamental tasks in an automated spine processing pipeline. Reliable and accurate processing of spine images is expected to benefit clinical decision-support systems for diagnosis, surgery planning, and population-based analysis on spine and bone health. However, designing automated algorithms for spine processing is challenging predominantly due to considerable variations in anatomy and acquisition protocols and due to a severe shortage of publicly available data. Addressing these limitations, the Large Scale Vertebrae Segmentation Challenge (VerSe) was organised in conjunction with the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) in 2019 and 2020, with a call for algorithms towards labelling and segmentation of vertebrae. Two datasets containing a total of 374 multi-detector CT scans from 355 patients were prepared and 4505 vertebrae have individually been annotated at voxel-level by a human-machine hybrid algorithm (https://osf.io/nqjyw/, https://osf.io/t98fz/). A total of 25 algorithms were benchmarked on these datasets. In this work, we present the the results of this evaluation and further investigate the performance-variation at vertebra-level, scan-level, and at different fields-of-view. We also evaluate the generalisability of the approaches to an implicit domain shift in data by evaluating the top performing algorithms of one challenge iteration on data from the other iteration. The principal takeaway from VerSe: the performance of an algorithm in labelling and segmenting a spine scan hinges on its ability to correctly identify vertebrae in cases of rare anatomical variations. The content and code concerning VerSe can be accessed at: https://github.com/anjany/verse.

Purpose: To use and test a labelling algorithm that operates on two-dimensional (2D) reformations, rather than three-dimensional (3D) data to locate and identify vertebrae. Methods: We improved the Btrfly Net (described by Sekuboyina et al) that works on sagittal and coronal maximum intensity projections (MIP) and augmented it with two additional components: spine-localization and adversarial a priori-learning. Furthermore, we explored two variants of adversarial training schemes that incorporated the anatomical a priori knowledge into the Btrfly Net. We investigated the superiority of the proposed approach for labelling vertebrae on three datasets: a public benchmarking dataset of 302 CT scans and two in-house datasets with a total of 238 CT scans. We employed Wilcoxon signed-rank test to compute the statistical significance of the improvement in performance observed due to various architectural components in our approach. Results: On the public dataset, our approach using the described Btrfly(pe-eb) network performed on par with current state-of-the-art methods achieving a statistically significant (p < .001) vertebrae identification rate of 88.5+/-0.2 % and localization distances of less than 7-mm. On the in-house datasets that had a higher inter-scan data variability, we obtained an identification rate of 85.1+/-1.2%. Conclusion: An identification performance comparable to existing 3D approaches was achieved when labelling vertebrae on 2D MIPs. The performance was further improved using the proposed adversarial training regime that effectively enforced local spine a priori knowledge during training. Lastly, spine-localization increased the generalizability of our approach by homogenizing the content in the MIPs.