Multitask deep learning on mammography to predict extensive intraductal component in invasive breast cancer
© 2023. The Author(s), under exclusive licence to European Society of Radiology..
OBJECTIVES: To develop a multitask deep learning (DL) algorithm to automatically classify mammography imaging findings and predict the existence of extensive intraductal component (EIC) in invasive breast cancer.
METHODS: Mammograms with invasive breast cancers from 2010 to 2019 were downloaded for two radiologists performing image segmentation and imaging findings annotation. Images were randomly split into training, validation, and test datasets. A multitask approach was performed on the EfficientNet-B0 neural network mainly to predict EIC and classify imaging findings. Three more models were trained for comparison, including a single-task model (predicting EIC), a two-task model (predicting EIC and cell receptor status), and a three-task model (combining the abovementioned tasks). Additionally, these models were trained in a subgroup of invasive ductal carcinoma. The DeLong test was used to examine the difference in model performance.
RESULTS: This study enrolled 1459 breast cancers on 3076 images. The EIC-positive rate was 29.0%. The three-task model was the best DL model with an area under the curve (AUC) of EIC prediction of 0.758 and 0.775 at the image and breast (patient) levels, respectively. Mass was the most accurately classified imaging finding (AUC = 0.915), followed by calcifications and mass with calcifications (AUC = 0.878 and 0.824, respectively). Cell receptor status prediction was less accurate (AUC = 0.625-0.653). The multitask approach improves the model training compared to the single-task model, but without significant effects.
CONCLUSIONS: A mammography-based multitask DL model can perform simultaneous imaging finding classification and EIC prediction.
CLINICAL RELEVANCE STATEMENT: The study results demonstrated the potential of deep learning to extract more information from mammography for clinical decision-making.
KEY POINTS: • Extensive intraductal component (EIC) is an independent risk factor of local tumor recurrence after breast-conserving surgery. • A mammography-based deep learning model was trained to predict extensive intraductal component close to radiologists' reading. • The developed multitask deep learning model could perform simultaneous imaging finding classification and extensive intraductal component prediction.
| Errataetall: | |
|---|---|
| Medienart: |
E-Artikel |
| Erscheinungsjahr: |
2024 |
|---|---|
| Erschienen: |
2024 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:34 |
|---|---|
| Enthalten in: |
European radiology - 34(2024), 4 vom: 09. Apr., Seite 2593-2604 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Tsai, Huei-Yi [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
Area under curve |
|---|
| Anmerkungen: |
Date Completed 22.03.2024 Date Revised 09.04.2024 published: Print-Electronic CommentIn: Eur Radiol. 2023 Nov 16;:. - PMID 37971682 Citation Status MEDLINE |
|---|
| doi: |
10.1007/s00330-023-10254-6 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM363054707 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM363054707 | ||
| 003 | DE-627 | ||
| 005 | 20240409232203.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231226s2024 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s00330-023-10254-6 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1370.xml |
| 035 | |a (DE-627)NLM363054707 | ||
| 035 | |a (NLM)37812297 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Tsai, Huei-Yi |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Multitask deep learning on mammography to predict extensive intraductal component in invasive breast cancer |
| 264 | 1 | |c 2024 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 22.03.2024 | ||
| 500 | |a Date Revised 09.04.2024 | ||
| 500 | |a published: Print-Electronic | ||
| 500 | |a CommentIn: Eur Radiol. 2023 Nov 16;:. - PMID 37971682 | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a © 2023. The Author(s), under exclusive licence to European Society of Radiology. | ||
| 520 | |a OBJECTIVES: To develop a multitask deep learning (DL) algorithm to automatically classify mammography imaging findings and predict the existence of extensive intraductal component (EIC) in invasive breast cancer | ||
| 520 | |a METHODS: Mammograms with invasive breast cancers from 2010 to 2019 were downloaded for two radiologists performing image segmentation and imaging findings annotation. Images were randomly split into training, validation, and test datasets. A multitask approach was performed on the EfficientNet-B0 neural network mainly to predict EIC and classify imaging findings. Three more models were trained for comparison, including a single-task model (predicting EIC), a two-task model (predicting EIC and cell receptor status), and a three-task model (combining the abovementioned tasks). Additionally, these models were trained in a subgroup of invasive ductal carcinoma. The DeLong test was used to examine the difference in model performance | ||
| 520 | |a RESULTS: This study enrolled 1459 breast cancers on 3076 images. The EIC-positive rate was 29.0%. The three-task model was the best DL model with an area under the curve (AUC) of EIC prediction of 0.758 and 0.775 at the image and breast (patient) levels, respectively. Mass was the most accurately classified imaging finding (AUC = 0.915), followed by calcifications and mass with calcifications (AUC = 0.878 and 0.824, respectively). Cell receptor status prediction was less accurate (AUC = 0.625-0.653). The multitask approach improves the model training compared to the single-task model, but without significant effects | ||
| 520 | |a CONCLUSIONS: A mammography-based multitask DL model can perform simultaneous imaging finding classification and EIC prediction | ||
| 520 | |a CLINICAL RELEVANCE STATEMENT: The study results demonstrated the potential of deep learning to extract more information from mammography for clinical decision-making | ||
| 520 | |a KEY POINTS: • Extensive intraductal component (EIC) is an independent risk factor of local tumor recurrence after breast-conserving surgery. • A mammography-based deep learning model was trained to predict extensive intraductal component close to radiologists' reading. • The developed multitask deep learning model could perform simultaneous imaging finding classification and extensive intraductal component prediction | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a Area under curve | |
| 650 | 4 | |a Breast neoplasms | |
| 650 | 4 | |a Mammography | |
| 650 | 4 | |a Neural networks, computer | |
| 700 | 1 | |a Kao, Yu-Wei |e verfasserin |4 aut | |
| 700 | 1 | |a Wang, Jo-Ching |e verfasserin |4 aut | |
| 700 | 1 | |a Tsai, Tsung-Yu |e verfasserin |4 aut | |
| 700 | 1 | |a Chung, Wei-Shiuan |e verfasserin |4 aut | |
| 700 | 1 | |a Hsu, Jui-Sheng |e verfasserin |4 aut | |
| 700 | 1 | |a Hou, Ming-Feng |e verfasserin |4 aut | |
| 700 | 1 | |a Weng, Shih-Feng |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t European radiology |d 1991 |g 34(2024), 4 vom: 09. Apr., Seite 2593-2604 |w (DE-627)NLM087691310 |x 1432-1084 |7 nnns |
| 773 | 1 | 8 | |g volume:34 |g year:2024 |g number:4 |g day:09 |g month:04 |g pages:2593-2604 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1007/s00330-023-10254-6 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 34 |j 2024 |e 4 |b 09 |c 04 |h 2593-2604 | ||