Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network
Graphically prescribed patient-specific imaging volumes and local pre-scan volumes are routinely placed by MRI technologists to optimize image quality. However, manual placement of these volumes by MR technologists is time-consuming, tedious, and subject to intra- and inter-operator variability. Res...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Tomography |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2379-139X/9/3/79 |
| _version_ | 1797592514831581184 |
|---|---|
| author | Timothy J. Allen Leah C. Henze Bancroft Kang Wang Ping Ni Wang Orhan Unal Lloyd D. Estkowski Ty A. Cashen Ersin Bayram Roberta M. Strigel James H. Holmes |
| author_facet | Timothy J. Allen Leah C. Henze Bancroft Kang Wang Ping Ni Wang Orhan Unal Lloyd D. Estkowski Ty A. Cashen Ersin Bayram Roberta M. Strigel James H. Holmes |
| author_sort | Timothy J. Allen |
| collection | DOAJ |
| description | Graphically prescribed patient-specific imaging volumes and local pre-scan volumes are routinely placed by MRI technologists to optimize image quality. However, manual placement of these volumes by MR technologists is time-consuming, tedious, and subject to intra- and inter-operator variability. Resolving these bottlenecks is critical with the rise in abbreviated breast MRI exams for screening purposes. This work proposes an automated approach for the placement of scan and pre-scan volumes for breast MRI. Anatomic 3-plane scout image series and associated scan volumes were retrospectively collected from 333 clinical breast exams acquired on 10 individual MRI scanners. Bilateral pre-scan volumes were also generated and reviewed in consensus by three MR physicists. A deep convolutional neural network was trained to predict both the scan and pre-scan volumes from the 3-plane scout images. The agreement between the network-predicted volumes and the clinical scan volumes or physicist-placed pre-scan volumes was evaluated using the intersection over union, the absolute distance between volume centers, and the difference in volume sizes. The scan volume model achieved a median 3D intersection over union of 0.69. The median error in scan volume location was 2.7 cm and the median size error was 2%. The median 3D intersection over union for the pre-scan placement was 0.68 with no significant difference in mean value between the left and right pre-scan volumes. The median error in the pre-scan volume location was 1.3 cm and the median size error was −2%. The average estimated uncertainty in positioning or volume size for both models ranged from 0.2 to 3.4 cm. Overall, this work demonstrates the feasibility of an automated approach for the placement of scan and pre-scan volumes based on a neural network model. |
| first_indexed | 2024-03-11T01:52:13Z |
| format | Article |
| id | doaj.art-863134151db94fc89294bcb87015ecde |
| institution | Directory Open Access Journal |
| issn | 2379-1381 2379-139X |
| language | English |
| last_indexed | 2024-03-11T01:52:13Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Tomography |
| spelling | doaj.art-863134151db94fc89294bcb87015ecde2023-11-18T12:53:42ZengMDPI AGTomography2379-13812379-139X2023-05-019396798010.3390/tomography9030079Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural NetworkTimothy J. Allen0Leah C. Henze Bancroft1Kang Wang2Ping Ni Wang3Orhan Unal4Lloyd D. Estkowski5Ty A. Cashen6Ersin Bayram7Roberta M. Strigel8James H. Holmes9Department of Medical Physics, University of Wisconsin-Madison, 1111 Highland Avenue, Madison, WI 53705, USADepartment of Radiology, University of Wisconsin-Madison, 600 Highland Avenue, Madison, WI 53792, USAGE Healthcare, 3000 N Grandview Blvd, Waukesha, WI 53188, USAGE Healthcare, 3000 N Grandview Blvd, Waukesha, WI 53188, USADepartment of Medical Physics, University of Wisconsin-Madison, 1111 Highland Avenue, Madison, WI 53705, USAGE Healthcare, 3000 N Grandview Blvd, Waukesha, WI 53188, USAGE Healthcare, 3000 N Grandview Blvd, Waukesha, WI 53188, USAGE Healthcare, 3000 N Grandview Blvd, Waukesha, WI 53188, USADepartment of Medical Physics, University of Wisconsin-Madison, 1111 Highland Avenue, Madison, WI 53705, USADepartment of Radiology, University of Iowa, 169 Newton Road, Iowa City, IA 52242, USAGraphically prescribed patient-specific imaging volumes and local pre-scan volumes are routinely placed by MRI technologists to optimize image quality. However, manual placement of these volumes by MR technologists is time-consuming, tedious, and subject to intra- and inter-operator variability. Resolving these bottlenecks is critical with the rise in abbreviated breast MRI exams for screening purposes. This work proposes an automated approach for the placement of scan and pre-scan volumes for breast MRI. Anatomic 3-plane scout image series and associated scan volumes were retrospectively collected from 333 clinical breast exams acquired on 10 individual MRI scanners. Bilateral pre-scan volumes were also generated and reviewed in consensus by three MR physicists. A deep convolutional neural network was trained to predict both the scan and pre-scan volumes from the 3-plane scout images. The agreement between the network-predicted volumes and the clinical scan volumes or physicist-placed pre-scan volumes was evaluated using the intersection over union, the absolute distance between volume centers, and the difference in volume sizes. The scan volume model achieved a median 3D intersection over union of 0.69. The median error in scan volume location was 2.7 cm and the median size error was 2%. The median 3D intersection over union for the pre-scan placement was 0.68 with no significant difference in mean value between the left and right pre-scan volumes. The median error in the pre-scan volume location was 1.3 cm and the median size error was −2%. The average estimated uncertainty in positioning or volume size for both models ranged from 0.2 to 3.4 cm. Overall, this work demonstrates the feasibility of an automated approach for the placement of scan and pre-scan volumes based on a neural network model.https://www.mdpi.com/2379-139X/9/3/79breast MRIdeep learningartificial intelligenceabbreviated breast MRworkflowautomation |
| spellingShingle | Timothy J. Allen Leah C. Henze Bancroft Kang Wang Ping Ni Wang Orhan Unal Lloyd D. Estkowski Ty A. Cashen Ersin Bayram Roberta M. Strigel James H. Holmes Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network Tomography breast MRI deep learning artificial intelligence abbreviated breast MR workflow automation |
| title | Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network |
| title_full | Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network |
| title_fullStr | Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network |
| title_full_unstemmed | Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network |
| title_short | Automated Placement of Scan and Pre-Scan Volumes for Breast MRI Using a Convolutional Neural Network |
| title_sort | automated placement of scan and pre scan volumes for breast mri using a convolutional neural network |
| topic | breast MRI deep learning artificial intelligence abbreviated breast MR workflow automation |
| url | https://www.mdpi.com/2379-139X/9/3/79 |
| work_keys_str_mv | AT timothyjallen automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT leahchenzebancroft automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT kangwang automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT pingniwang automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT orhanunal automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT lloyddestkowski automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT tyacashen automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT ersinbayram automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT robertamstrigel automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork AT jameshholmes automatedplacementofscanandprescanvolumesforbreastmriusingaconvolutionalneuralnetwork |