Enhancing MR image segmentation with realistic adversarial data augmentation
The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Chen, Chen [verfasserIn] |
|---|
| Format: |
E-Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2022transfer abstract |
|---|
| Schlagwörter: |
|---|
| Übergeordnetes Werk: |
Enthalten in: Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation - Mohammadi, Behzad ELSEVIER, 2019, Amsterdam [u.a.] |
|---|---|
| Übergeordnetes Werk: |
volume:82 ; year:2022 ; pages:0 |
| Links: |
|---|
| DOI / URN: |
10.1016/j.media.2022.102597 |
|---|
| Katalog-ID: |
ELV059224630 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | ELV059224630 | ||
| 003 | DE-627 | ||
| 005 | 20230626052449.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 221103s2022 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1016/j.media.2022.102597 |2 doi | |
| 028 | 5 | 2 | |a /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica |
| 035 | |a (DE-627)ELV059224630 | ||
| 035 | |a (ELSEVIER)S1361-8415(22)00230-4 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 690 |q VZ |
| 084 | |a 52.43 |2 bkl | ||
| 084 | |a 52.52 |2 bkl | ||
| 084 | |a 52.42 |2 bkl | ||
| 084 | |a 50.38 |2 bkl | ||
| 100 | 1 | |a Chen, Chen |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Enhancing MR image segmentation with realistic adversarial data augmentation |
| 264 | 1 | |c 2022transfer abstract | |
| 336 | |a nicht spezifiziert |b zzz |2 rdacontent | ||
| 337 | |a nicht spezifiziert |b z |2 rdamedia | ||
| 338 | |a nicht spezifiziert |b zu |2 rdacarrier | ||
| 520 | |a The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. | ||
| 520 | |a The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. | ||
| 650 | 7 | |a Adversarial training |2 Elsevier | |
| 650 | 7 | |a MR image segmentation |2 Elsevier | |
| 650 | 7 | |a Data augmentation |2 Elsevier | |
| 650 | 7 | |a Model generalization |2 Elsevier | |
| 650 | 7 | |a Adversarial data augmentation |2 Elsevier | |
| 700 | 1 | |a Qin, Chen |4 oth | |
| 700 | 1 | |a Ouyang, Cheng |4 oth | |
| 700 | 1 | |a Li, Zeju |4 oth | |
| 700 | 1 | |a Wang, Shuo |4 oth | |
| 700 | 1 | |a Qiu, Huaqi |4 oth | |
| 700 | 1 | |a Chen, Liang |4 oth | |
| 700 | 1 | |a Tarroni, Giacomo |4 oth | |
| 700 | 1 | |a Bai, Wenjia |4 oth | |
| 700 | 1 | |a Rueckert, Daniel |4 oth | |
| 773 | 0 | 8 | |i Enthalten in |n Elsevier Science |a Mohammadi, Behzad ELSEVIER |t Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation |d 2019 |g Amsterdam [u.a.] |w (DE-627)ELV003074609 |
| 773 | 1 | 8 | |g volume:82 |g year:2022 |g pages:0 |
| 856 | 4 | 0 | |u https://doi.org/10.1016/j.media.2022.102597 |3 Volltext |
| 912 | |a GBV_USEFLAG_U | ||
| 912 | |a GBV_ELV | ||
| 912 | |a SYSFLAG_U | ||
| 936 | b | k | |a 52.43 |j Kältetechnik |q VZ |
| 936 | b | k | |a 52.52 |j Thermische Energieerzeugung |j Wärmetechnik |q VZ |
| 936 | b | k | |a 52.42 |j Heizungstechnik |j Lüftungstechnik |j Klimatechnik |q VZ |
| 936 | b | k | |a 50.38 |j Technische Thermodynamik |q VZ |
| 951 | |a AR | ||
| 952 | |d 82 |j 2022 |h 0 | ||
| author_variant |
c c cc |
|---|---|
| matchkey_str |
chenchenqinchenouyangchenglizejuwangshuo:2022----:nacnmiaeemnainihelsiavrai |
| hierarchy_sort_str |
2022transfer abstract |
| bklnumber |
52.43 52.52 52.42 50.38 |
| publishDate |
2022 |
| allfields |
10.1016/j.media.2022.102597 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica (DE-627)ELV059224630 (ELSEVIER)S1361-8415(22)00230-4 DE-627 ger DE-627 rakwb eng 690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl Chen, Chen verfasserin aut Enhancing MR image segmentation with realistic adversarial data augmentation 2022transfer abstract nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Elsevier Qin, Chen oth Ouyang, Cheng oth Li, Zeju oth Wang, Shuo oth Qiu, Huaqi oth Chen, Liang oth Tarroni, Giacomo oth Bai, Wenjia oth Rueckert, Daniel oth Enthalten in Elsevier Science Mohammadi, Behzad ELSEVIER Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation 2019 Amsterdam [u.a.] (DE-627)ELV003074609 volume:82 year:2022 pages:0 https://doi.org/10.1016/j.media.2022.102597 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U 52.43 Kältetechnik VZ 52.52 Thermische Energieerzeugung Wärmetechnik VZ 52.42 Heizungstechnik Lüftungstechnik Klimatechnik VZ 50.38 Technische Thermodynamik VZ AR 82 2022 0 |
| spelling |
10.1016/j.media.2022.102597 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica (DE-627)ELV059224630 (ELSEVIER)S1361-8415(22)00230-4 DE-627 ger DE-627 rakwb eng 690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl Chen, Chen verfasserin aut Enhancing MR image segmentation with realistic adversarial data augmentation 2022transfer abstract nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Elsevier Qin, Chen oth Ouyang, Cheng oth Li, Zeju oth Wang, Shuo oth Qiu, Huaqi oth Chen, Liang oth Tarroni, Giacomo oth Bai, Wenjia oth Rueckert, Daniel oth Enthalten in Elsevier Science Mohammadi, Behzad ELSEVIER Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation 2019 Amsterdam [u.a.] (DE-627)ELV003074609 volume:82 year:2022 pages:0 https://doi.org/10.1016/j.media.2022.102597 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U 52.43 Kältetechnik VZ 52.52 Thermische Energieerzeugung Wärmetechnik VZ 52.42 Heizungstechnik Lüftungstechnik Klimatechnik VZ 50.38 Technische Thermodynamik VZ AR 82 2022 0 |
| allfields_unstemmed |
10.1016/j.media.2022.102597 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica (DE-627)ELV059224630 (ELSEVIER)S1361-8415(22)00230-4 DE-627 ger DE-627 rakwb eng 690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl Chen, Chen verfasserin aut Enhancing MR image segmentation with realistic adversarial data augmentation 2022transfer abstract nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Elsevier Qin, Chen oth Ouyang, Cheng oth Li, Zeju oth Wang, Shuo oth Qiu, Huaqi oth Chen, Liang oth Tarroni, Giacomo oth Bai, Wenjia oth Rueckert, Daniel oth Enthalten in Elsevier Science Mohammadi, Behzad ELSEVIER Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation 2019 Amsterdam [u.a.] (DE-627)ELV003074609 volume:82 year:2022 pages:0 https://doi.org/10.1016/j.media.2022.102597 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U 52.43 Kältetechnik VZ 52.52 Thermische Energieerzeugung Wärmetechnik VZ 52.42 Heizungstechnik Lüftungstechnik Klimatechnik VZ 50.38 Technische Thermodynamik VZ AR 82 2022 0 |
| allfieldsGer |
10.1016/j.media.2022.102597 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica (DE-627)ELV059224630 (ELSEVIER)S1361-8415(22)00230-4 DE-627 ger DE-627 rakwb eng 690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl Chen, Chen verfasserin aut Enhancing MR image segmentation with realistic adversarial data augmentation 2022transfer abstract nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Elsevier Qin, Chen oth Ouyang, Cheng oth Li, Zeju oth Wang, Shuo oth Qiu, Huaqi oth Chen, Liang oth Tarroni, Giacomo oth Bai, Wenjia oth Rueckert, Daniel oth Enthalten in Elsevier Science Mohammadi, Behzad ELSEVIER Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation 2019 Amsterdam [u.a.] (DE-627)ELV003074609 volume:82 year:2022 pages:0 https://doi.org/10.1016/j.media.2022.102597 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U 52.43 Kältetechnik VZ 52.52 Thermische Energieerzeugung Wärmetechnik VZ 52.42 Heizungstechnik Lüftungstechnik Klimatechnik VZ 50.38 Technische Thermodynamik VZ AR 82 2022 0 |
| allfieldsSound |
10.1016/j.media.2022.102597 doi /cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica (DE-627)ELV059224630 (ELSEVIER)S1361-8415(22)00230-4 DE-627 ger DE-627 rakwb eng 690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl Chen, Chen verfasserin aut Enhancing MR image segmentation with realistic adversarial data augmentation 2022transfer abstract nicht spezifiziert zzz rdacontent nicht spezifiziert z rdamedia nicht spezifiziert zu rdacarrier The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Elsevier Qin, Chen oth Ouyang, Cheng oth Li, Zeju oth Wang, Shuo oth Qiu, Huaqi oth Chen, Liang oth Tarroni, Giacomo oth Bai, Wenjia oth Rueckert, Daniel oth Enthalten in Elsevier Science Mohammadi, Behzad ELSEVIER Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation 2019 Amsterdam [u.a.] (DE-627)ELV003074609 volume:82 year:2022 pages:0 https://doi.org/10.1016/j.media.2022.102597 Volltext GBV_USEFLAG_U GBV_ELV SYSFLAG_U 52.43 Kältetechnik VZ 52.52 Thermische Energieerzeugung Wärmetechnik VZ 52.42 Heizungstechnik Lüftungstechnik Klimatechnik VZ 50.38 Technische Thermodynamik VZ AR 82 2022 0 |
| language |
English |
| source |
Enthalten in Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation Amsterdam [u.a.] volume:82 year:2022 pages:0 |
| sourceStr |
Enthalten in Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation Amsterdam [u.a.] volume:82 year:2022 pages:0 |
| format_phy_str_mv |
Article |
| bklname |
Kältetechnik Thermische Energieerzeugung Wärmetechnik Heizungstechnik Lüftungstechnik Klimatechnik Technische Thermodynamik |
| institution |
findex.gbv.de |
| topic_facet |
Adversarial training MR image segmentation Data augmentation Model generalization Adversarial data augmentation |
| dewey-raw |
690 |
| isfreeaccess_bool |
false |
| container_title |
Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation |
| authorswithroles_txt_mv |
Chen, Chen @@aut@@ Qin, Chen @@oth@@ Ouyang, Cheng @@oth@@ Li, Zeju @@oth@@ Wang, Shuo @@oth@@ Qiu, Huaqi @@oth@@ Chen, Liang @@oth@@ Tarroni, Giacomo @@oth@@ Bai, Wenjia @@oth@@ Rueckert, Daniel @@oth@@ |
| publishDateDaySort_date |
2022-01-01T00:00:00Z |
| hierarchy_top_id |
ELV003074609 |
| dewey-sort |
3690 |
| id |
ELV059224630 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">ELV059224630</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230626052449.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">221103s2022 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1016/j.media.2022.102597</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">/cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)ELV059224630</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ELSEVIER)S1361-8415(22)00230-4</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">690</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">52.43</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">52.52</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">52.42</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">50.38</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Chen, Chen</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Enhancing MR image segmentation with realistic adversarial data augmentation</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022transfer abstract</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zzz</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">z</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zu</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications.</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications.</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Adversarial training</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">MR image segmentation</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Data augmentation</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Model generalization</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Adversarial data augmentation</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Qin, Chen</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Ouyang, Cheng</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Li, Zeju</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Wang, Shuo</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Qiu, Huaqi</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Chen, Liang</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Tarroni, Giacomo</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Bai, Wenjia</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Rueckert, Daniel</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="n">Elsevier Science</subfield><subfield code="a">Mohammadi, Behzad ELSEVIER</subfield><subfield code="t">Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation</subfield><subfield code="d">2019</subfield><subfield code="g">Amsterdam [u.a.]</subfield><subfield code="w">(DE-627)ELV003074609</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:82</subfield><subfield code="g">year:2022</subfield><subfield code="g">pages:0</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.1016/j.media.2022.102597</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ELV</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_U</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">52.43</subfield><subfield code="j">Kältetechnik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">52.52</subfield><subfield code="j">Thermische Energieerzeugung</subfield><subfield code="j">Wärmetechnik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">52.42</subfield><subfield code="j">Heizungstechnik</subfield><subfield code="j">Lüftungstechnik</subfield><subfield code="j">Klimatechnik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">50.38</subfield><subfield code="j">Technische Thermodynamik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">82</subfield><subfield code="j">2022</subfield><subfield code="h">0</subfield></datafield></record></collection>
|
| author |
Chen, Chen |
| spellingShingle |
Chen, Chen ddc 690 bkl 52.43 bkl 52.52 bkl 52.42 bkl 50.38 Elsevier Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Enhancing MR image segmentation with realistic adversarial data augmentation |
| authorStr |
Chen, Chen |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)ELV003074609 |
| format |
electronic Article |
| dewey-ones |
690 - Buildings |
| delete_txt_mv |
keep |
| author_role |
aut |
| collection |
elsevier |
| remote_str |
true |
| illustrated |
Not Illustrated |
| topic_title |
690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl Enhancing MR image segmentation with realistic adversarial data augmentation Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation Elsevier |
| topic |
ddc 690 bkl 52.43 bkl 52.52 bkl 52.42 bkl 50.38 Elsevier Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation |
| topic_unstemmed |
ddc 690 bkl 52.43 bkl 52.52 bkl 52.42 bkl 50.38 Elsevier Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation |
| topic_browse |
ddc 690 bkl 52.43 bkl 52.52 bkl 52.42 bkl 50.38 Elsevier Adversarial training Elsevier MR image segmentation Elsevier Data augmentation Elsevier Model generalization Elsevier Adversarial data augmentation |
| format_facet |
Elektronische Aufsätze Aufsätze Elektronische Ressource |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
zu |
| author2_variant |
c q cq c o co z l zl s w sw h q hq l c lc g t gt w b wb d r dr |
| hierarchy_parent_title |
Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation |
| hierarchy_parent_id |
ELV003074609 |
| dewey-tens |
690 - Building & construction |
| hierarchy_top_title |
Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)ELV003074609 |
| title |
Enhancing MR image segmentation with realistic adversarial data augmentation |
| ctrlnum |
(DE-627)ELV059224630 (ELSEVIER)S1361-8415(22)00230-4 |
| title_full |
Enhancing MR image segmentation with realistic adversarial data augmentation |
| author_sort |
Chen, Chen |
| journal |
Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation |
| journalStr |
Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
600 - Technology |
| recordtype |
marc |
| publishDateSort |
2022 |
| contenttype_str_mv |
zzz |
| container_start_page |
0 |
| author_browse |
Chen, Chen |
| container_volume |
82 |
| class |
690 VZ 52.43 bkl 52.52 bkl 52.42 bkl 50.38 bkl |
| format_se |
Elektronische Aufsätze |
| author-letter |
Chen, Chen |
| doi_str_mv |
10.1016/j.media.2022.102597 |
| dewey-full |
690 |
| title_sort |
enhancing mr image segmentation with realistic adversarial data augmentation |
| title_auth |
Enhancing MR image segmentation with realistic adversarial data augmentation |
| abstract |
The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. |
| abstractGer |
The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. |
| abstract_unstemmed |
The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications. |
| collection_details |
GBV_USEFLAG_U GBV_ELV SYSFLAG_U |
| title_short |
Enhancing MR image segmentation with realistic adversarial data augmentation |
| url |
https://doi.org/10.1016/j.media.2022.102597 |
| remote_bool |
true |
| author2 |
Qin, Chen Ouyang, Cheng Li, Zeju Wang, Shuo Qiu, Huaqi Chen, Liang Tarroni, Giacomo Bai, Wenjia Rueckert, Daniel |
| author2Str |
Qin, Chen Ouyang, Cheng Li, Zeju Wang, Shuo Qiu, Huaqi Chen, Liang Tarroni, Giacomo Bai, Wenjia Rueckert, Daniel |
| ppnlink |
ELV003074609 |
| mediatype_str_mv |
z |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| author2_role |
oth oth oth oth oth oth oth oth oth |
| doi_str |
10.1016/j.media.2022.102597 |
| up_date |
2024-07-06T21:20:56.833Z |
| _version_ |
1803866191254519808 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">ELV059224630</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230626052449.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">221103s2022 xx |||||o 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1016/j.media.2022.102597</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">/cbs_pica/cbs_olc/import_discovery/elsevier/einzuspielen/GBV00000000001933.pica</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)ELV059224630</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ELSEVIER)S1361-8415(22)00230-4</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">690</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">52.43</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">52.52</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">52.42</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">50.38</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Chen, Chen</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Enhancing MR image segmentation with realistic adversarial data augmentation</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022transfer abstract</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zzz</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">z</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zu</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications.</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications.</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Adversarial training</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">MR image segmentation</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Data augmentation</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Model generalization</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="650" ind1=" " ind2="7"><subfield code="a">Adversarial data augmentation</subfield><subfield code="2">Elsevier</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Qin, Chen</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Ouyang, Cheng</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Li, Zeju</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Wang, Shuo</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Qiu, Huaqi</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Chen, Liang</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Tarroni, Giacomo</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Bai, Wenjia</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Rueckert, Daniel</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="n">Elsevier Science</subfield><subfield code="a">Mohammadi, Behzad ELSEVIER</subfield><subfield code="t">Exergoeconomic analysis and multi-objective optimization of a semi-solar greenhouse with experimental validation</subfield><subfield code="d">2019</subfield><subfield code="g">Amsterdam [u.a.]</subfield><subfield code="w">(DE-627)ELV003074609</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:82</subfield><subfield code="g">year:2022</subfield><subfield code="g">pages:0</subfield></datafield><datafield tag="856" ind1="4" ind2="0"><subfield code="u">https://doi.org/10.1016/j.media.2022.102597</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ELV</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_U</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">52.43</subfield><subfield code="j">Kältetechnik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">52.52</subfield><subfield code="j">Thermische Energieerzeugung</subfield><subfield code="j">Wärmetechnik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">52.42</subfield><subfield code="j">Heizungstechnik</subfield><subfield code="j">Lüftungstechnik</subfield><subfield code="j">Klimatechnik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">50.38</subfield><subfield code="j">Technische Thermodynamik</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">82</subfield><subfield code="j">2022</subfield><subfield code="h">0</subfield></datafield></record></collection>
|
| score |
7.4000654 |