Conference item
Towards capturing sonographic experience: cognition-inspired ultrasound video saliency prediction
- Abstract:
- For visual tasks like ultrasound (US) scanning, experts direct their gaze towards regions of task-relevant information. Therefore, learning to predict the gaze of sonographers on US videos captures the spatio-temporal patterns that are important for US scanning. The spatial distribution of gaze points on video frames can be represented through heat maps termed saliency maps. Here, we propose a temporally bidirectional model for video saliency prediction (BDS-Net), drawing inspiration from modern theories of human cognition. The model consists of a convolutional neural network (CNN) encoder followed by a bidirectional gated-recurrent-unit recurrent convolutional network (GRU-RCN) decoder. The temporal bidirectionality mimics human cognition, which simultaneously reacts to past and predicts future sensory inputs. We train the BDS-Net alongside spatial and temporally one-directional comparative models on the task of predicting saliency in videos of US abdominal circumference plane detection. The BDS-Net outperforms the comparative models on four out of five saliency metrics. We present a qualitative analysis on representative examples to explain the model’s superior performance.
- Publication status:
- Published
- Peer review status:
- Reviewed (other)
Actions
Access Document
- Files:
-
-
(Preview, Accepted manuscript, pdf, 4.8MB, Terms of use)
-
- Publisher copy:
- 10.1007/978-3-030-39343-4_15
- Publication website:
- https://link.springer.com/chapter/10.1007/978-3-030-39343-4_15
Authors
- Publisher:
- Springer Verlag
- Host title:
- Medical Image Understanding and Analysis
- Journal:
- MIUA: Annual Conference on Medical Image Understanding and Analysis More from this journal
- Pages:
- 174-186
- Series:
- Communications in Computer and Information Science
- Publication date:
- 2020-01-24
- Acceptance date:
- 2019-04-16
- Event title:
- Annual Conference on Medical Image Understanding and Analysis
- Event location:
- Liverpool, UK
- Event start date:
- 2019-07-24
- Event end date:
- 2019-07-26
- DOI:
- EISSN:
-
1865-0937
- ISSN:
-
1865-0929
- ISBN:
- 9783319959214
- Language:
-
English
- Keywords:
- Pubs id:
-
pubs:997166
- UUID:
-
uuid:a14df633-3dc5-4918-ba90-09dda3f51363
- Local pid:
-
pubs:997166
- Source identifiers:
-
997166
- Deposit date:
-
2019-05-09
Terms of use
- Copyright holder:
- Springer Nature
- Copyright date:
- 2020
- Rights statement:
- © Springer Nature Switzerland AG 2020
- Notes:
- This is the accepted manuscript version of the article. The final version is available from Springer at: https://doi.org/10.1007/978-3-030-39343-4_15
If you are the owner of this record, you can report an update to it here: Report update to this record