Efficient slice anomaly detection network for 3D brain MRI Volume.
Current anomaly detection methods excel with benchmark industrial data but struggle with natural images and medical data due to varying definitions of 'normal' and 'abnormal.' This makes accurate identification of deviations in these fields particularly challenging. Especially fo...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2025-06-01
|
| Series: | PLOS Digital Health |
| Online Access: | https://doi.org/10.1371/journal.pdig.0000874 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849682856568160256 |
|---|---|
| author | Zeduo Zhang Yalda Mohsenzadeh |
| author_facet | Zeduo Zhang Yalda Mohsenzadeh |
| author_sort | Zeduo Zhang |
| collection | DOAJ |
| description | Current anomaly detection methods excel with benchmark industrial data but struggle with natural images and medical data due to varying definitions of 'normal' and 'abnormal.' This makes accurate identification of deviations in these fields particularly challenging. Especially for 3D brain MRI data, all the state-of-the-art models are reconstruction-based with 3D convolutional neural networks which are memory-intensive, time-consuming and producing noisy outputs that require further post-processing. We propose a framework called Simple Slice-based Network (SimpleSliceNet), which utilizes a model pre-trained on ImageNet and fine-tuned on a separate MRI dataset as a 2D slice feature extractor to reduce computational cost. We aggregate the extracted features to perform anomaly detection tasks on 3D brain MRI volumes. Our model integrates a conditional normalizing flow to calculate log likelihood of features and employs the contrastive loss to enhance anomaly detection accuracy. The results indicate improved performance, showcasing our model's remarkable adaptability and effectiveness when addressing the challenges exists in brain MRI data. In addition, for the large-scale 3D brain volumes, our model SimpleSliceNet outperforms the state-of-the-art 2D and 3D models in terms of accuracy, memory usage and time consumption. Code is available at: https://github.com/Jarvisarmy/SimpleSliceNet. |
| format | Article |
| id | doaj-art-37a2e784b09d45cb8d408f11840b19d6 |
| institution | DOAJ |
| issn | 2767-3170 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLOS Digital Health |
| spelling | doaj-art-37a2e784b09d45cb8d408f11840b19d62025-08-20T03:24:03ZengPublic Library of Science (PLoS)PLOS Digital Health2767-31702025-06-0146e000087410.1371/journal.pdig.0000874Efficient slice anomaly detection network for 3D brain MRI Volume.Zeduo ZhangYalda MohsenzadehCurrent anomaly detection methods excel with benchmark industrial data but struggle with natural images and medical data due to varying definitions of 'normal' and 'abnormal.' This makes accurate identification of deviations in these fields particularly challenging. Especially for 3D brain MRI data, all the state-of-the-art models are reconstruction-based with 3D convolutional neural networks which are memory-intensive, time-consuming and producing noisy outputs that require further post-processing. We propose a framework called Simple Slice-based Network (SimpleSliceNet), which utilizes a model pre-trained on ImageNet and fine-tuned on a separate MRI dataset as a 2D slice feature extractor to reduce computational cost. We aggregate the extracted features to perform anomaly detection tasks on 3D brain MRI volumes. Our model integrates a conditional normalizing flow to calculate log likelihood of features and employs the contrastive loss to enhance anomaly detection accuracy. The results indicate improved performance, showcasing our model's remarkable adaptability and effectiveness when addressing the challenges exists in brain MRI data. In addition, for the large-scale 3D brain volumes, our model SimpleSliceNet outperforms the state-of-the-art 2D and 3D models in terms of accuracy, memory usage and time consumption. Code is available at: https://github.com/Jarvisarmy/SimpleSliceNet.https://doi.org/10.1371/journal.pdig.0000874 |
| spellingShingle | Zeduo Zhang Yalda Mohsenzadeh Efficient slice anomaly detection network for 3D brain MRI Volume. PLOS Digital Health |
| title | Efficient slice anomaly detection network for 3D brain MRI Volume. |
| title_full | Efficient slice anomaly detection network for 3D brain MRI Volume. |
| title_fullStr | Efficient slice anomaly detection network for 3D brain MRI Volume. |
| title_full_unstemmed | Efficient slice anomaly detection network for 3D brain MRI Volume. |
| title_short | Efficient slice anomaly detection network for 3D brain MRI Volume. |
| title_sort | efficient slice anomaly detection network for 3d brain mri volume |
| url | https://doi.org/10.1371/journal.pdig.0000874 |
| work_keys_str_mv | AT zeduozhang efficientsliceanomalydetectionnetworkfor3dbrainmrivolume AT yaldamohsenzadeh efficientsliceanomalydetectionnetworkfor3dbrainmrivolume |