On the usage of average Hausdorff distance for segmentation performance assessment: hidden error when used for ranking

@article{Aydin2021OnTU,
  title={On the usage of average Hausdorff distance for segmentation performance assessment: hidden error when used for ranking},
  author={Orhun Utku Aydin and Abdel Aziz Taha and Adam Hilbert and Ahmed A. Khalil and Ivana Galinovic and Jochen B. Fiebach and Dietmar Frey and Vince Istvan Madai},
  journal={European Radiology Experimental},
  year={2021},
  volume={5}
}
Average Hausdorff distance is a widely used performance measure to calculate the distance between two point sets. In medical image segmentation, it is used to compare ground truth images with segmentations allowing their ranking. We identified, however, ranking errors of average Hausdorff distance making it less suitable for applications in segmentation performance assessment. To mitigate this error, we present a modified calculation of this performance measure that we have coined “balanced… 
Auto-Detection and Segmentation of Involved Lymph Nodes in HPV-Associated Oropharyngeal Cancer Using a Convolutional Deep Learning Neural Network
TLDR
The results from this single-institution study demonstrate the successful automation of lymph node segmentation for patients with HPV-OPC using a DL-CNN.
Survey of Supervised Learning for Medical Image Processing
TLDR
This survey explains the performance metrics of supervised learning methods, summarizes the available medical datasets, and studies the state-of-the-art supervised learning architectures for medical imaging processing, including convolutional neural networks and their corresponding algorithms.
Toward Sharing Brain Images: Differentially Private TOF-MRA Images With Segmentation Labels Using Generative Adversarial Networks
TLDR
This study implemented a Wasserstein GAN with and without differential privacy guarantees to generate privacy-preserving labeled Time-of-Flight Magnetic Resonance Angiography (TOF-MRA) image patches for brain vessel segmentation to ensure the patient's privacy while maintaining the predictive properties of the data.
Towards a Guideline for Evaluation Metrics in Medical Image Segmentation
TLDR
This work provides an overview and interpretation guide on the following metrics for medical image segmentation evaluation in binary as well as multi-class problems: Dice similarity coefficient, Jaccard, Sensitivity, Specificity, Rand index, ROC curves, Cohen’s Kappa, and Hausdorff distance.
An evaluation of performance measures for arterial brain vessel segmentation
TLDR
Average Hausdorff distance-based measures should be used as a standard performance measure in evaluating cerebral vessel segmentation quality because they can identify more relevant segmentation errors, especially in high-quality segmentations.
Automatic segmentation of deep grey nuclei using a high‐resolution 7T magnetic resonance imaging atlas—Quantification of T1 values in healthy volunteers
TLDR
A fully automated segmentation process permitting the extraction of unbiased normative T1 values in a population of young healthy controls as a reference for characterizing subtle structural alterations of deep grey nuclei relevant to a range of neurological diseases is presented.
Bag-of-Vectors Autoencoders for Unsupervised Conditional Text Generation
TLDR
This work extends Emb2Emb to Bag-of-Vectors Autoencoders (BoV-AEs), which encode the text into a variablesize bag of vectors that grows with the size of the text, as in attention-based models, and proposes regularization techniques that facilitate learning meaningful operations in the latent space.
DBSegment: Fast and robust segmentation of deep brain structures - Evaluation of transportability across acquisition domains
TLDR
This paper uses deep learning to provide a one-step, robust and efficient deep brain segmentation solution directly in the patient’s native space and generalizes with high reliability.
...
1
2
...

References

SHOWING 1-10 OF 16 REFERENCES
significantly improved efficiency and reliability
  • Neuroimage 31:1116–1128 https://doi.org/10.1016/j. neuroimage.2006.01.015 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Aydin et al. European Radiology Experimental
  • 2021
BRAVE-NET: Fully Automated Arterial Brain Vessel Segmentation In Patients with Cerebrovascular Disease
TLDR
A new state-of-the-art of arterial brain vessel segmentation tailored to cerebrovascular pathology is presented and an extensive experimental validation of the model is provided using a large aggregated dataset encompassing a large variability of cerebroVascular disease.
BRAVE-NET: Fully Automated Arterial Brain Vessel Segmentation in Patients With Cerebrovascular Disease
TLDR
The BRAVE-NET model is a multiscale 3-D convolutional neural network model developed on a dataset of 264 patients from three different studies enrolling patients with cerebrovascular diseases and validated using high-quality manual labels as ground truth and is the most resistant toward false labelings as revealed by the visual analysis.
Impact of inter-reader contouring variability on textural radiomics of colorectal liver metastases
TLDR
A 2D contouring approach may help mitigate overall inter-reader variability, albeit stable RFs can be extracted from both 3D and 2D segmentations of CRC liver metastases.
Manual prostate cancer segmentation in MRI: interreader agreement and volumetric correlation with transperineal template core needle biopsy
TLDR
A systematic discrepancy between the tumor volume estimate by MRI segmentation and TTSB core length is suggested, especially for large and high-grade tumors.
A U-Net Deep Learning Framework for High Performance Vessel Segmentation in Patients With Cerebrovascular Disease
TLDR
The U-net framework was optimized and evaluated with three metrics: Dice coefficient, 95% Hausdorff distance (95HD) and average Hausdorf distance (AVD) and revealed excellent performance in large vessels and sufficient performance in small vessels.
A UNet Deep Learning Framework for High Performance Vessel Segmentation in Patients With Cerebrovascular Disease
  • Front Neurosci [Internet]
  • 2019
MR Imaging of the Extracranial Facial Nerve with the CISS Sequence
TLDR
The facial nerve can be seen on CISS images from the stylomastoid foramen to the temporofacial and cervicofacial trunks, proximal to the parotid plexus, and use of a CISS protocol is feasible in the clinical setting to determine the location of the nerve relative to tumor.
Standardization of Temporal Bone CT Planes across a Multisite Academic Institution
TLDR
Evaluating whether an automated postprocessing algorithm or an educational intervention with postprocessing by a technologist could result in consistent temporal bone image reformations into planes referenced to the lateral semicircular canal resulted in significantly less imaging plane variance compared with images reformatted in the acquisition plane.
Deep learning for segmentation of brain tumors: Impact of cross‐institutional training and testing
TLDR
There is a very strong effect of selecting data for training on performance of CNNs in a multi-institutional setting, and the reasons behind this effect requires additional comprehensive investigation.
...
1
2
...