Observer variation-aware medical image segmentation by combining deep learning and surrogate-assisted genetic algorithms

Dushatskiy, Arkadiy; Mendrik, Adriënne M.; Bosman, Peter A. N.; Alderliesten, Tanja

doi:10.1117/12.2547459

Cited by 3 publications

(3 citation statements)

References 10 publications

(12 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multi-observer medical image segmentation pertains to learning automatic segmentation based on delineations provided by multiple expert observers, which may be conflicting due to inter-observer variability [38,42]. We applied our MO learning approach to the multi-observer medical image segmentation scenario mentioned in [7]. The dataset [34] contains Magnetic Resonance Imaging (MRI) scans of prostate regions of 32 patients.…”

Section: Multi-observer Medical Image Segmentationmentioning

confidence: 99%

Multi-Objective Learning to Predict Pareto Fronts Using Hypervolume Maximization

Deist¹,

Grewal²,

Dankers³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Real-world problems are often multi-objective with decision-makers unable to specify a priori which trade-off between the conflicting objectives is preferable. Intuitively, building machine learning solutions in such cases would entail providing multiple predictions that span and uniformly cover the Pareto front of all optimal trade-off solutions. We propose a novel learning approach to estimate the Pareto front by maximizing the dominated hypervolume (HV) of the average loss vectors corresponding to a set of learners, leveraging established multi-objective optimization methods. In our approach, the set of learners are trained multi-objectively with a dynamic loss function, wherein each learner's losses are weighted by their HV maximizing gradients. Consequently, the learners get trained according to different trade-offs on the Pareto front, which otherwise is not guaranteed for fixed linear scalarizations or when optimizing for specific trade-offs per learner without knowing the shape of the Pareto front. Experiments on three different multi-objective tasks show that the outputs of the set of learners are indeed well-spread on the Pareto front. Further, the outputs corresponding to validation samples are also found to closely follow the trade-offs that were learned from training samples for our set of benchmark problems.

show abstract

Section: Multi-observer Medical Image Segmentationmentioning

confidence: 99%

Multi-Objective Learning to Predict Pareto Fronts Using Hypervolume Maximization

Deist¹,

Grewal²,

Dankers³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…To the best of our knowledge, the only similar approach was initially proposed by us in. 3 However, in our previous work substantially large simulated segmentation variations were used, and the approach was mainly suited for only two types of variations. Here, we propose an efficient algorithm capable of capturing an arbitrary number of variations (defined by the user beforehand) and apply it to segmentation variations present in real clinical data.…”

Section: Introductionmentioning

confidence: 99%

Data variation-aware medical image segmentation

Dushatskiy¹,

Lowe²,

Bosman³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Deep learning algorithms have become the golden standard for segmentation of medical imaging data. In most works, the variability and heterogeneity of real clinical data is acknowledged to still be a problem. One way to automatically overcome this is to capture and exploit this variation explicitly. Here, we propose an approach that improves on our previous work in this area and explain how it potentially can improve clinical acceptance of (semi-)automatic segmentation methods.In contrast to a standard neural network that produces one segmentation, we propose to use a multi-path Unet network that produces multiple segmentation variants, presumably corresponding to the variations that reside in the dataset. Different paths of the network are trained on disjoint data subsets. Because a priori it may be unclear what variations exist in the data, the subsets should be automatically determined. This is achieved by searching for the best data partitioning with an evolutionary optimization algorithm. Because each network path can become more specialized when trained on a more homogeneous data subset, better segmentation quality can be achieved. In practical usage, various automatically produced segmentations can be presented to a medical expert, from which the preferred segmentation can be selected. In experiments with a real clinical dataset of CT scans with prostate segmentations, our approach provides an improvement of several percentage points in terms of Dice and surface Dice coefficients compared to when all network paths are trained on all training data. Noticeably, the largest improvement occurs in the upper part of the prostate that is known to be most prone to inter-observer segmentation variation.

show abstract

“…A function evaluation that corresponds to a partial [3] or end-to-end [32] Deep Neural Network (DNN) training procedure can be extremely expensive, taking up to several hours [23]. A related application is finding an ensemble of deep learning models through data partitioning, for instance to tackle the problem of data heterogeneity in medical image analysis [20,29]. Here too, the efficiency of the search algorithm is extremely important as each function evaluation again requires training a DNN.…”

Section: Introductionmentioning

confidence: 99%

A Novel Surrogate-assisted Evolutionary Algorithm Applied to Partition-based Ensemble Learning

Dushatskiy¹,

Alderliesten²,

Bosman³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

We propose a novel surrogate-assisted Evolutionary Algorithm for solving expensive combinatorial optimization problems. We integrate a surrogate model, which is used for fitness value estimation, into a state-of-the-art P3-like variant of the Gene-Pool Optimal Mixing Algorithm (GOMEA) and adapt the resulting algorithm for solving non-binary combinatorial problems. We test the proposed algorithm on an ensemble learning problem. Ensembling several models is a common Machine Learning technique to achieve better performance. We consider ensembles of several models trained on disjoint subsets of a dataset. Finding the best dataset partitioning is naturally a combinatorial non-binary optimization problem. Fitness function evaluations can be extremely expensive if complex models, such as Deep Neural Networks, are used as learners in an ensemble. Therefore, the number of fitness function evaluations is typically limited, necessitating expensive optimization techniques.In our experiments we use five classification datasets from the OpenML-CC18 benchmark and Support-vector Machines as learners in an ensemble. The proposed algorithm demonstrates better performance than alternative approaches, including Bayesian optimization algorithms. It manages to find better solutions using just several thousand fitness function evaluations for an ensemble learning problem with up to 500 variables.

show abstract

Observer variation-aware medical image segmentation by combining deep learning and surrogate-assisted genetic algorithms

Cited by 3 publications

References 10 publications

Multi-Objective Learning to Predict Pareto Fronts Using Hypervolume Maximization

Multi-Objective Learning to Predict Pareto Fronts Using Hypervolume Maximization

Data variation-aware medical image segmentation

A Novel Surrogate-assisted Evolutionary Algorithm Applied to Partition-based Ensemble Learning

Contact Info

Product

Resources

About