Curriculum learning for improved femur fracture classification: Scheduling data with prior knowledge and uncertainty

  title={Curriculum learning for improved femur fracture classification: Scheduling data with prior knowledge and uncertainty},
  author={Amelia Jim{\'e}nez-S{\'a}nchez and Diana Mateus and Sonja Kirchhoff and Chlodwig Kirchhoff and Peter Biberthaler and Nassir Navab and Miguel A. Gonz{\'a}lez Ballester and Gemma Piella},
  journal={Medical image analysis},

Trustworthy clinical AI solutions: a unified review of uncertainty quantification in deep learning models for medical image analysis

An overview of the existing methods to quantify uncertainty associated to DL predictions is proposed, which focuses on applications to medical image analysis, which present specific challenges due to the high dimensionality of images and their quality variability, as well as constraints associated to real-life clinical routine.

Appointment Scheduling Problem in Complexity Systems of the Healthcare Services: A Comprehensive Review

This paper provides a comprehensive review of Appointment Scheduling (AS) in healthcare service while we propose appointment scheduling problems and various applications and solution approaches in



Precise proximal femur fracture classification for interactive training and surgical planning

The feasibility of a fully automatic computer-aided diagnosis (CAD) tool that localizes and classifies proximal femur fractures on X-ray images according to the AO classification is demonstrated, achieving results comparable to expert-level and state-of-the-art performance.

Medical-based Deep Curriculum Learning for Improved Fracture Classification

This work proposes and compares several strategies relying on curriculum learning, to support the classification of proximal femur fracture from X-ray images, a challenging problem as reflected by existing intra- and inter-expert disagreement.

Automatic Classification of Proximal Femur Fractures Based on Attention Models

This work adapts an attention model known as Spatial Transformer to learn an image-dependent localization of the ROI trained only from image classification labels and reports high accuracy with regard to inter-expert correlation values reported in the literature.

Using Convolutional Neural Networks and Transfer Learning for Bone Age Classification

The proposed approach outperforms the current state-of-the-art classification methods in BAA with small dataset and uses transfer learning within DCNNs to perform bone age classifications making full use of advantages ofDCNNs.

Weakly Supervised Universal Fracture Detection in Pelvic X-rays

A two-stage hip and pelvic fracture detection method that executes localized fracture classification using weakly supervised ROI mining and can perform comparably to human physicians (even outperforming emergency physicians and surgeons), in a preliminary reader study of 23 readers.

Attention-Guided Curriculum Learning for Weakly Supervised Classification and Localization of Thoracic Diseases on Chest Radiographs

A convolutional neural network (CNN) based attention-guided curriculum learning (AGCL) framework is presented, which leverages the severity-level attributes mined from radiology reports to improve the classification and localization performance of thoracic diseases from chest radiographs.

Detecting intertrochanteric hip fractures with orthopedist-level accuracy using a deep convolutional neural network

The performance of the convolutional neural network exceeded that of orthopedic surgeons in detecting intertrochanteric hip fractures from proximal femoral radiographs under limited conditions and has a significant potential to be a useful tool for screening for fractures on plain radiographs, especially in the emergency room.

Quantifying and Leveraging Classification Uncertainty for Chest Radiograph Assessment

It is argued that explicitly learning the classification uncertainty as an orthogonal measure to the predicted output, is essential to account for the inherent variability characteristic of this data.