3D G-CNNs reduce false positives in lung nodule detection.
problem Reducing false positives in pulmonary nodule detection.
method Used 3D roto-translation group convolutions (G-Convs) instead of traditional convolutions.
result 3D G-CNNs achieved FROC scores close to those of a CNN trained on ten times more data.
Reduces false positives in lung nodule detection by using unlabeled data.
problem Lack of labeled data for training supervised algorithms in medical imaging.
method Uses pseudo-negative labels from unlabeled data to refine a pulmonary nodule detection network.
result False positive rate reduced from 0.4864 to 0.1266 while maintaining sensitivity.
Propagating uncertainty improves deep learning model performance.
problem Improving computer-aided detection of pulmonary nodules.
method Multi-stage Bayesian CNN architecture with uncertainty propagation.
result Improves overall performance in terms of accuracy and model confidence.
Study uses multi-task Bayesian optimization to speed up SVM hyperparameter tuning for nodules diagnosis.
problem Redundant and time-consuming hyperparameter tuning for SVM classifiers in medical imaging.
method Employed multi-task Bayesian optimization to accelerate hyperparameter search.
result Multi-task Bayesian optimization significantly accelerates hyperparameter search.
System converts 3D lung nodule images into embeddings for retrieval.
problem Retrieving similar 3D lung nodule images for radiologist decision support.
method 3D deep learning, semantic representation, transfer learning, similarity score.
result System can measure similarity between nodule annotations and CBIR results.
Computed tomography (CT) generates a stack of cross-sectional images covering a region of the body. The visual assessment of these images for the identification of potential abnormalities is a challenging and time consuming task due to the large amount of information that needs to be processed. In this article we propo…
Study uses weak labels and visual attention networks to detect lung nodules in chest radiographs.
problem Automated detection of lung nodules in chest radiographs requires large amounts of manually annotated images.
method Proposes two network architectures: one using saliency maps and the other a recurrent attention model trained with reinforcement learning.
result Demonstrates promising nodule detection performance using weak labels and visual attention mechanisms.
Reinforcement learning improves self training for medical image segmentation.
problem Lack of labeled data in medical imaging.
method Integrating reinforcement learning into self training for complex segmentation networks.
result Improved segmentation performance with less labeled data.
Paper classifies lung nodules in CT scans, improving diagnostic accuracy.
problem Early detection of lung nodules for better treatment planning.
method Proposes four 3D neural networks for direct mapping from 3D images to class labels.
result 3D multi-output DenseNet achieves state-of-the-art classification accuracy.
A novel approach for 3D lung nodule segmentation using adaptive ROI and multi-view residual learning.
problem Inaccurate nodule segmentation due to fixed ROI and redundant structures.
method Two-stage approach: 2D ROI patch-wise investigation with adaptive ROI strategy, followed by 2D and 3D VOI investigation with deep residual U-Net.
result Significantly robust and accurate nodule segmentation compared to previous methods.
NoduleX predicts lung nodule malignancy with high accuracy using CT scans.
problem Challenges in accurately predicting lung nodule malignancy from CT scans.
method Deep learning convolutional neural networks (CNN) trained on a large dataset of lung nodules.
result NoduleX achieves an AUC of ~0.99 for nodule malignancy classification, comparable to radiologists.
GD networks improve lung nodule classification accuracy.
problem Difficult classification of lung nodules of varying sizes.
method Proposes Gated-Dilated (GD) networks with Context-Aware sub-network.
result GD network outperforms state-of-the-art models with AUC > 0.95.
U-Det improves lung nodule segmentation in CT images.
problem Challenging shapes and surroundings of lung nodules in CT images.
method End-to-end deep learning with Bi-FPN, Mish activation, and class weights.
result U-Det achieves 82.82% Dice similarity coefficient, comparable to human experts.
3D Axial-Attention improves lung nodule classification accuracy.
problem Limited 3D attention in existing methods.
method Proposes 3D Axial-Attention network with 3D positional encoding.
result 3D Axial-Attention achieves state-of-the-art performance.
Deep Local-Global network improves lung nodule malignancy prediction.
problem Challenging task of classifying lung nodules as benign or malignant.
method Proposes a novel method combining local and global feature extraction.
result Achieved state-of-the-art results with AUC=95.62%.
S4ND detects lung nodules faster and more accurately.
problem Efficient lung nodule detection from CT scans.
method Single-Shot Single-Scale 3D Convolutional Neural Network (CNN) trained end-to-end.
result S4ND outperforms state-of-the-art methods in terms of efficiency and accuracy.
Generates synthetic lung nodule images for training.
problem Lack of source image data for training machine learning models.
method Autoencoder techniques for 3D shape generation.
result Produces high-quality synthetic 3D images.
End-to-end lung nodule detection system improves sensitivity and performance.
problem Detecting subtle lung nodules in raw CT data.
method Deep reconstruction network followed by 3D-CNN for nodule detection.
result End-to-end detector achieves comparable sensitivity to fully-sampled data.
Paper reduces false positives in lung nodule detection using deep learning on point clouds.
problem Reduces false positives in lung nodule detection.
method Uses deep learning models for point clouds to transform 3D CT scan data.
result Achieved 85.98 FROC compared to 77.26 FROC for baseline models.
Characterization of lung nodules as benign or malignant is one of the most important tasks in lung cancer diagnosis, staging and treatment planning. While the variation in the appearance of the nodules remains large, there is a need for a fast and robust computer aided system. In this work, we propose an end-to-end tra…
Clearing algorithm improves CT image segmentation accuracy by merging confident annotations.
problem Inaccurate predictions due to noisy annotations from different annotators.
method 3-stage algorithm: scoring annotators, scoring nodules, merging annotations.
result Improves prediction accuracy in CT image segmentation tasks.
GAN normalizes CT scans for consistent radiomic feature values.
problem Variations in dose levels and slice thickness affect radiomic features sensitivity.
method Used a 3D generative adversarial network (GAN) to normalize reduced dose, thick slice images to normal dose, thinner slice images.
result GAN-based approach led to significantly smaller error in radiomic features.
End-to-end CAD system for thyroid nodule classification using multimodal data and expert guidance.
problem Improving accuracy in thyroid nodule classification for clinicians.
method Knowledge-driven DenseNet framework using multimodal ultrasound data and expert cues.
result The proposed system achieves relevant performances in thyroid nodule classification.
Pipeline detects pulmonary embolisms from sparse CT images.
problem Manual diagnosis of pulmonary embolisms is laborious and prone to errors.
method Two-stage pipeline using AI, sparse annotations, and robust models.
result Achieved AUC scores of 0.94 on validation and 0.85 on test sets for severe PEs.
Synthetic sampling improves per-class performance in multi-label malignancy prediction.
problem Imbalanced multi-label classification problem in CADx systems.
method Synthetic oversampling techniques using random forest classifier.
result Average 7.22% point increase in sensitivity for minority classes.
This work introduces an integrative approach based on Q-analysis with machine learning. The new approach, called Neural Hypernetwork, has been applied to a case study of pulmonary embolism diagnosis. The objective of the application of neural hyper-network to pulmonary embolism (PE) is to improve diagnose for reducing …
Modeling disease progression in irregularly observed patients.
problem Irregular patient observation in healthcare databases.
method Continuous-time hidden Markov model with generalized linear model.
result Interpretable model of healthcare utilization events.
Smartphone app diagnoses pulmonary diseases from chest X-rays.
problem Scarcity of training data and class imbalance issues.
method Data Augmentation Generative Adversarial Network (DAGAN) and Convolutional Siamese Network with attention mechanism.
result Achieved 99.30% and 98.40% testing accuracy on Binary/Multiclass scenarios.
New method uses probabilistic independence to discover disease signatures from medical records.
problem Insufficiently precise diagnosis of clinical disease leading to treatment failures.
method Unsupervised machine learning using probabilistic independence to disentangle disease patterns.
result Inferred 2000 clinical disease signatures from medical records, improving cancer prediction.
UNet++ improves medical image segmentation with deep supervision.
problem Improving accuracy in medical image segmentation.
method Nested U-Net architecture with deep supervision.
result UNet++ achieves significant improvements in IoU scores.
CNNs accurately measure airways and vessels on CT images, improving lung disease diagnosis.
problem Accurately characterizing small pulmonary structures from CT images for disease diagnosis.
method Generative model combined with Convolutional Neural Regressor (CNR) for cross-sectional measurements.
result CNNs provide accurate measurements with physiological correlates.
Improves Bayesian optimization for expensive function evaluations.
problem Optimizing expensive functions with noisy, non-differentiable outputs.
method Proposes a new improvement-based acquisition function.
result Performs at least as well as state-of-the-art acquisition functions, often better.
SAPSAM trains CNNs on lung CTs with binary labels, improving CPA detection and localization.
problem Chronic Pulmonary Aspergillosis (CPA) detection and localization on CT scans using binary labels.
method Binary labels, average intensity projections, 2D RGB-like images, hierarchical CNN architectures.
result High classification accuracy, precise localization, predictive power of 2-year survival.
Anatomical and biophysical modeling of left atrium (LA) and proximal pulmonary veins (PPVs) is important for clinical management of several cardiac diseases. Magnetic resonance imaging (MRI) allows qualitative assessment of LA and PPVs through visualization. However, there is a strong need for an advanced image segment…
Generative Adversarial Model improves RV segmentation accuracy in MRI.
problem Accurate segmentation of the right ventricular blood pool from cine MRI sequences.
method Combines FCNN, Gated Recurrent Units (GRU), Generative Adversarial Networks (GAN), and L1 loss function.
result Improves Dice Index and Hausdorff Distance by 0.05 and 3.49 mm respectively.
WEST uses EHRs and expert cases to improve rare disease phenotyping.
problem Limited labeled data for rare diseases.
method Weakly supervised transformer model trained on probabilistic silver-standard labels.
result WEST outperforms existing methods in phenotype classification and subphenotyping.
Proposes a tool to contrast global vs personalized models in clinical prediction.
problem Balancing global vs personalized models in clinical prediction.
method Localized regression approach using autoencoder for dimension reduction.
result Identification of patient subgroups where global models fall short.
Study uses data to analyze COPD patients' impact on hospital systems.
problem Understanding and quantifying resource requirements for COPD patients.
method Combines segmentation, queuing theory, and data recovery techniques.
result Finding useful operational results from incomplete administrative data.
The paper uses geometric methods to classify medical data histograms.
problem Classifying medical data histograms for disease diagnosis.
method Information geometry of beta distributions for comparing and classifying histograms.
result Geometric tools, particularly negatively curved Fisher information, enable unique mean calculation and K-means classification.
Improves uncertainty estimates in lung node segmentation.
problem Inaccurate uncertainty estimates in medical segmentation.
method Supervised learning using multi-grader annotation variability.
result Improves predictive uncertainty estimates and sample diversity.
Lung segmentation from abnormal CXRs using data imputation.
problem Segmenting lungs from CXRs with high opacity caused by respiratory ailments.
method Modified CNN-based segmentation network with deep generative model for data imputation.
result The model can segment lungs from abnormal CXRs, extending to cases with extreme abnormalities.
We correct for sampling bias in training models to improve real-world performance.
problem Sampling bias causes discrepancies between lab and real-world model performance.
method Bayesian risk minimization and derived bias-corrected loss functions.
result Our approach integrates seamlessly into current learning paradigms and improves model performance.
Deep learning model predicts CXR findings from patient scans.
problem Insufficient radiologists for accurate CXR interpretation.
method Tagged 959k studies, trained deep learning model on 2M reports.
result Model performs similarly to radiologists on 12 findings.
Chronic obstructive pulmonary disease (COPD) is a lung disease where early detection benefits the survival rate. COPD can be quantified by classifying patches of computed tomography images, and combining patch labels into an overall diagnosis for the image. As labeled patches are often not available, image labels are p…
New methods improve genetic studies of complex diseases.
problem Improving genetic studies of complex diseases using high-dimensional clinical data.
method Evaluation of unsupervised disentangled representation learning methods (autoencoders, VAE, beta-VAE, FactorVAE) for genetic association studies.
result FactorVAEs and beta-VAEs outperform standard VAEs and non-variational autoencoders in genetic studies of asthma and COPD.
Proposes a novel anomaly detection method for echocardiogram videos.
problem Anomaly detection in echocardiogram videos.
method Dynamic Variational Trajectory Models (TVAE-C, TVAE-R, TVAE-S) trained on healthy infant echocardiogram videos.
result Superior performance in detecting congenital heart defects and pulmonary hypertension.
Method learns feature maps from deep CNN layers for weakly supervised chest pathology localization.
problem Localization of chest pathologies in X-ray images is challenging due to varying sizes and appearances.
method Class-aware deep multiscale feature learning using intermediate feature maps from CNN layers.
result Improves localization performance of small pathologies like nodules and masses.
V-FCNN neural network segments atria from clinical images.
problem Automatic segmentation of atria from high-resolution images.
method Volumetric Fully Convolution Neural Network (V-FCNN) with a mixed loss function.
result 92.5% Dice similarity coefficient in 54 patients.