CT-3DFlow : Leveraging 3D Normalizing Flows for Unsupervised Detection of Pathological Pulmonary CT scans
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914730401071104 |
|---|---|
| author | Djahnine, Aissam Popoff, Alexandre Jupin-Delevaux, Emilien Cottin, Vincent Nempont, Olivier Boussel, Loic |
| author_facet | Djahnine, Aissam Popoff, Alexandre Jupin-Delevaux, Emilien Cottin, Vincent Nempont, Olivier Boussel, Loic |
| contents | Unsupervised pathology detection can be implemented by training a model on healthy data only and measuring the deviation from the training set upon inference, for example with CNN-based feature extraction and one-class classifiers, or reconstruction-score-based methods such as AEs, GANs and Diffusion models. Normalizing Flows (NF) have the ability to directly learn the probability distribution of training examples through an invertible architecture. We leverage this property in a novel 3D NF-based model named CT-3DFlow, specifically tailored for patient-level pulmonary pathology detection in chest CT data. Our model is trained unsupervised on healthy 3D pulmonary CT patches, and detects deviations from its log-likelihood distribution as anomalies. We aggregate patches-level likelihood values from a patient's CT scan to provide a patient-level 'normal'/'abnormal' prediction. Out-of-distribution detection performance is evaluated using expert annotations on a separate chest CT test dataset, outperforming other state-of-the-art methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_18514 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | CT-3DFlow : Leveraging 3D Normalizing Flows for Unsupervised Detection of Pathological Pulmonary CT scans Djahnine, Aissam Popoff, Alexandre Jupin-Delevaux, Emilien Cottin, Vincent Nempont, Olivier Boussel, Loic Image and Video Processing Computer Vision and Pattern Recognition Machine Learning Unsupervised pathology detection can be implemented by training a model on healthy data only and measuring the deviation from the training set upon inference, for example with CNN-based feature extraction and one-class classifiers, or reconstruction-score-based methods such as AEs, GANs and Diffusion models. Normalizing Flows (NF) have the ability to directly learn the probability distribution of training examples through an invertible architecture. We leverage this property in a novel 3D NF-based model named CT-3DFlow, specifically tailored for patient-level pulmonary pathology detection in chest CT data. Our model is trained unsupervised on healthy 3D pulmonary CT patches, and detects deviations from its log-likelihood distribution as anomalies. We aggregate patches-level likelihood values from a patient's CT scan to provide a patient-level 'normal'/'abnormal' prediction. Out-of-distribution detection performance is evaluated using expert annotations on a separate chest CT test dataset, outperforming other state-of-the-art methods. |
| title | CT-3DFlow : Leveraging 3D Normalizing Flows for Unsupervised Detection of Pathological Pulmonary CT scans |
| topic | Image and Video Processing Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2403.18514 |