Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,051 papers · 148 categories

Trend · papers per month

12.5%25.0%37.5%50.0% · Jul 199319922001200920182026
48 results for surface Dice similarity coefficient

Bayesian variational inference improves medical image segmentation confidence.

problem Improving interpretability and confidence in deep learning models for medical image segmentation.
method Encoder-decoder architecture based on variational inference for segmenting brain tumor images.
result The model segments brain tumors with both aleatoric and epistemic uncertainty.

Deep learning improves radiotherapy by automating head and neck cancer organ segmentation.

problem Manual delineation of radio-sensitive organs in head and neck cancer radiotherapy is time-consuming and variably performed.
method A 3D U-Net deep learning architecture trained on clinical CT scans and expert segmentations.
result Deep learning model achieves expert-level performance in delineating 21 distinct OARs.

Deep BV automates brain ventricle segmentation from 3D ultrasound images of mouse embryos.

problem Manual segmentation of brain ventricles from high-frequency ultrasound images is tedious, time-consuming, and requires specialized expertise.
method A fully automated system using two modules: localization and segmentation. Localization identifies a 3D bounding box containing the entire brain ventricle. Segmentation then segments the detected bounding box into brain ventricle and background.
result Achieves a Dice Similarity Coefficient (DSC) of 0.8956 for BV segmentation on an unseen test set, surpassing the previous state-of-the-art method by 25%

Lung segmentation accuracy varies little across diverse datasets.

problem Limited clinical applicability of automated lung segmentation methods.
method Comparison of four deep learning approaches and two standard algorithms on diverse datasets.
result Standard U-net approach yields higher accuracy on routine imaging data.

Enhanced deep learning model improves tumor segmentation in ultrasound images.

problem Challenges in integrating patient-specific medical priors into deep learning models.
method Integrates visual saliency into a U-Net architecture with attention blocks.
result Achieved a Dice similarity coefficient of 90.5 percent on a dataset of 510 images.

Post-processing improves WMH segmentation accuracy from randomly-initialized U-nets.

problem Improving WMH segmentation accuracy from randomly-initialized U-nets.
method Post-processing technique involving thresholding and averaging U-net outputs.
result Superior performance of the proposed post-processing method.

Paper reduces Hausdorff Distance in medical image segmentation.

problem Reduction of Hausdorff Distance in medical image segmentation.
method Three novel loss functions for training CNNs to estimate and reduce HD.
result Approximately 18-45% reduction in HD without degrading other performance metrics.

A novel 3D U-Net approach improves kidney tumor segmentation in medical imaging.

problem Challenging manual annotation and great medical impact of kidney tumor segmentation.
method End-to-end cascaded U-Nets with a localization network.
result Achieves Sørensen-Dice coefficients of 0.902 for kidney and 0.408 for tumor segmentation.

AI tool automates blood segmentation from head CT scans after SAH.

problem Accurate volumetric assessment of SAH patients for clinical and prognostic implications.
method Transformer-based Swin UNETR architecture for noncontrast CT scans.
result High accuracy and robust performance across internal and external validation cohorts.

Model infers mineral locations from geospatial data, improving predictions with auxiliary data.

problem Challenges in characterizing hidden mineral deposits underground.
method Generative modeling approach using masked and infilled geospatial maps.
result Models achieve Dice coefficients of 0.31 and recalls of 0.22 at 1×1 mi² resolution.

MI-GAN generates synthetic medical images for supervised analysis.

problem Lack of large datasets and overfitting in medical image analysis.
method Generative Adversarial Network (GAN) for generating synthetic medical images and masks.
result MI-GAN achieves state-of-the-art performance with dice coefficient of 0.837 on STARE and 0.832 on DRIVE datasets.

New biomarker predicts MRgFUS treatment outcome without contrast agents.

problem Inaccurate assessment of treated tissue viability after MRgFUS.
method Deep learning on noncontrast multiparametric MRI images, voxel-wise registration.
result Predicted follow-up NPV with DICE coefficient 0.71, outperforming current standard.

Study compares shallow and deep learning for MS lesion segmentation.

problem Automated segmentation of white matter lesions in early-stage MS patients.
method Training and testing shallow and deep learning architectures on 32 patients.
result Combining shallow and deep architectures improves lesion-wise metrics.

Few-shot brain segmentation achieved with weak labels and deep networks.

problem Efficient brain segmentation from limited labeled data.
method Heteroscedastic multi-task networks with Monte-Carlo inference and direct probability learning.
result Significant improvements in segmentation accuracy with minimal labeled data.

A new framework for consistent segmentation evaluation reduces operating losses.

problem Inconsistent thresholding-based segmentation methods lead to suboptimal solutions.
method Developed a consistent ranking-based framework (RankDice/RankIoU) using Bayes rules and Dice-/IoU-calibration.
result The proposed framework is Dice-/IoU-calibrated and provides excess risk bounds and convergence rates.

Study shows federated learning can segment brain tumors without data sharing.

problem Lack of sufficient medical data for deep learning models.
method Federated learning for multi-institutional collaboration without sharing patient data.
result Federated semantic segmentation models perform similarly to those trained with shared data.

DiCE uses diverse agents to explore and learn, avoiding local minima.

problem Local minima in RL due to limited exploration and correlated behavior.
method DiCE employs a group of heterogeneous agents to explore simultaneously and share experiences, with a diversity regularization mechanism.
result DiCE achieves substantial improvement over baselines in MuJoCo locomotion tasks.

Unified DICE estimators as regularized Lagrangians for improved off-policy evaluation.

problem Improving off-policy evaluation from behavior-agnostic data.
method Unified derivation of DICE estimators as regularized Lagrangians of a linear program.
result Dual solutions offer greater flexibility and provide superior estimates in practice.

Proposes a method to capture similarities and covariances between related tasks in multivariate regression.

problem Predicting multiple response variables with shared explanatory variables and capturing within-group similarities.
method Uses multivariate linear mixed models to estimate coefficients and errors, modeling within-group similarities through joint estimation of covariance matrices.
result The proposed MrRCE method outperforms natural competitors and alternative estimators in various model settings.

Defines an implied CO2-price to cover climate change costs, finding it significantly higher than the SCC.

problem The social cost of carbon (SCC) does not fully cover climate change costs.
method Defines an implied CO2-price as a 'polluter pays principle' and calculates its value using a DICE model.
result The cost-implied CO2 price is around 500/tCO2,comparedto50/tCO2, compared to 50/tCO2 for SCC.

Bounding twist number of surface links using polynomial coefficients.

problem Bounding the twist number of alternating surface links.
method Introducing a generalized homological Kauffman bracket and applying it to surface link diagrams.
result A bound for the twist number of alternating surface links in terms of polynomial coefficients.

Quantum cluster algebras for surfaces with coefficients defined using skein theory.

problem Defining quantum cluster algebras for surfaces with coefficients.
method Introducing a skein algebra and proving it has a quantum cluster structure.
result The skein algebra of a walled surface naturally generalizes quantum cluster algebras of marked surfaces.

Study homology groups for non-orientable surfaces with boundary.

problem Determine the first homology group for mapping class groups of non-orientable surfaces.
method Calculate the first homology group with twisted coefficients.
result Explicitly determined the first homology group for various mapping class groups.

New coefficient detects irrational rotation behavior on infinite-type surfaces.

problem Detecting irrational rotation behavior on surfaces of infinite type.
method Introducing a new quasimorphism, the Dehn twist coefficient, and proving its properties.
result The Dehn twist coefficient can have image all of R for some infinite-type surfaces.

This work shows cosine similarity is equivalent to Pearson correlation for word vectors, but not all vectors are suitable for cosine.

problem The use of cosine similarity for semantic textual similarity is often taken for granted, despite its limitations.
method Characterized cases where Pearson correlation is unfit and introduced rank correlation as an alternative.
result Pearson correlation is equivalent to cosine similarity for many word vectors but not all, and rank correlation can improve performance.

We present a novel approach to automatically segment magnetic resonance (MR) images of the human brain into anatomical regions. Our methodology is based on a deep artificial neural network that assigns each voxel in an MR image of the brain to its corresponding anatomical region. The inputs of the network capture infor…

2015-02-09abs ↗pdf ↗

Study the first homology group for a specific mapping class group.

problem Calculating the first homology group for a specific mapping class group.
method Determined using coefficients in H1(N;Z)H_1(N;\mathbb{Z}) for a non-orientable surface.
result Determined the first homology group for the mapping class group of a specific surface.

This paper adapts PATE for semantic segmentation while maintaining privacy.

problem Preserving privacy in medical machine learning, especially for sensitive information.
method Adapting PATE for semantic segmentation using low-dimensional representations and low-sensitivity queries.
result An Autoencoder-based PATE variant achieves a higher Dice coefficient for the same privacy guarantee.

Study shows Dehn twist coefficients are consistent across different actions on surfaces.

problem Consistency of fractional Dehn twist coefficients under various actions.
method Analyzes left orderings of mapping class groups and uses cofinality properties.
result Fractional Dehn twist coefficients are independent of the underlying action for surfaces with genus > 1.