CASED: Curriculum Adaptive Sampling for Extreme Data Imbalance

Abstract

We introduce CASED, a novel curriculum sampling algorithm that facilitates the optimization of deep learning segmentation or detection models on data sets with extreme class imbalance. We evaluate the CASED learning framework on the task of lung nodule detection in chest CT. In contrast to two-stage solutions, wherein nodule candidates are first proposed by a segmentation model and refined by a second detection stage, CASED improves the training of deep nodule segmentation models (e.g. UNet) to the point where state of the art results are achieved using only a trivial detection stage. CASED improves the optimization of deep segmentation models by allowing them to first learn how to distinguish nodules from their immediate surroundings, while continuously adding a greater proportion of difficult-to-classify global context, until uniformly sampling from the empirical data distribution. Using CASED during training yields a minimalist proposal to the lung nodule detection problem that tops the LUNA16 nodule detection benchmark with an average sensitivity score of 88.35%. Furthermore, we find that models trained using CASED are robust to nodule annotation quality by showing that comparable results can be achieved when only a point and radius for each ground truth nodule are provided during training. Finally, the CASED learning framework makes no assumptions with regard to imaging modality or segmentation target and should generalize to other medical imaging problems where class imbalance is a persistent problem.

Medical Image Segmentation With Limited Supervision: A Review of Deep Network Models

“…It has been empirically demonstrated that this learning paradigm is useful in avoiding bad local minima and in achieving better generalization ability [134]. Data curriculum learning has recently been used in several medical applications, especially location and classification tasks [78], [135]- [137] but few in segmentation tasks [138], [139]. To train a deep network for the classification and location of thoracic diseases on chest radiographs, Tang et al [135] first ranked the training images according to the difficulty (indicated by the severity-levels of the disease) and then fed them to the deep network to boost the representation learning gradually.…”

“…In addition to the predefined curriculum by prior knowledge and keeping it fixed after that, the curriculum can also be dynamically determined to adapt to the feedback of the learner, also known as self-paced curriculum learning [141] or self-paced learning [142]. For lung nodule segmentation/detection with extreme class imbalance, Jesson et al [138] introduced an adaptive sampling strategy, which favors difficult-to-classify examples. For instance-level segmentation of pulmonary nodule, Wang et al [139] employed pseudo labels as the surrogate of ground truth labels on unlabeled data.…”

Deep Learning Application for Analyzing of Constituents and Their Correlations in the Interpretations of Medical Images

“…CASED performs adaptive curriculum sampling to solve the problem of highly data imbalance and makes it possible for the model to distinguish nodules from immediate proximity and subsequently enlarges the hard-declassified global context, up to uniform categories in the empirical data pool. In this way, CASED is the most performant and is used in the detection of pulmonary nodules in thoracic CT [165].…”

Automatic CNN-based detection of cardiac MR motion artefacts using k-space data augmentation and curriculum learning

“…Recently, the idea of curriculum learning has been utilised for medical imaging challenges. Jesson et al (2017) proposed to use patches of different complexity to train a network for lung nodule detection. Their algorithm learnt how to distinguish nodules from the initial surroundings and added difficult patches gradually.…”