Improving interobserver agreement and performance of deep learning models for segmenting acute ischemic stroke by combining DWI with optimized ADC thresholds.
Saved in:
| Title: | Improving interobserver agreement and performance of deep learning models for segmenting acute ischemic stroke by combining DWI with optimized ADC thresholds. |
|---|---|
| Authors: | Juan, Chun-Jung1,2,3,4,5 (AUTHOR), Lin, Shao-Chieh2,6 (AUTHOR), Li, Ya-Hui2,7 (AUTHOR), Chang, Chia-Ching2,8 (AUTHOR), Jeng, Yi-Hung2,5 (AUTHOR), Peng, Hsu-Hsia5 (AUTHOR), Huang, Teng-Yi9 (AUTHOR), Chung, Hsiao-Wen7,10 (AUTHOR), Shen, Wu-Chung2,3 (AUTHOR), Tsai, Chon-Haw11 (AUTHOR), Chang, Ruey-Feng1,7 (AUTHOR) rfchang@csie.ntu.edu.tw, Liu, Yi-Jui12 (AUTHOR) ericliu6@ms35.hinet.net |
| Source: | European Radiology. Aug2022, Vol. 32 Issue 8, p5371-5381. 11p. |
| Subjects: | Ischemic stroke, Deep learning, Inter-observer reliability, Image segmentation, Computer-assisted image analysis (Medicine), Bland-Altman plot |
| Abstract: | Objectives: To examine the role of ADC threshold on agreement across observers and deep learning models (DLMs) plus segmentation performance of DLMs for acute ischemic stroke (AIS). Methods: Twelve DLMs, which were trained on DWI-ADC-ADC combination from 76 patients with AIS using 6 different ADC thresholds with ground truth manually contoured by 2 observers, were tested by additional 67 patients in the same hospital and another 78 patients in another hospital. Agreement between observers and DLMs were evaluated by Bland-Altman plot and intraclass correlation coefficient (ICC). The similarity between ground truth (GT) defined by observers and between automatic segmentation performed by DLMs was evaluated by Dice similarity coefficient (DSC). Group comparison was performed using the Mann-Whitney U test. The relationship between the DSC and ADC threshold as well as AIS lesion size was evaluated by linear regression analysis. A p <.05 was considered statistically significant. Results: Excellent interobserver agreement and intraobserver repeatability in the manual segmentation (all ICC > 0.98, p <.001) were achieved. The 95% limit of agreement was reduced from 11.23 cm2 for GT on DWI to 0.59 cm2 for prediction at an ADC threshold of 0.6 × 10−3 mm2/s combined with DWI. The segmentation performance of DLMs was improved with an overall DSC from 0.738 ± 0.214 on DWI to 0.971 ± 0.021 on an ADC threshold of 0.6 × 10−3 mm2/s combined with DWI. Conclusions: Combining an ADC threshold of 0.6 × 10−3 mm2/s with DWI reduces interobserver and inter-DLM difference and achieves best segmentation performance of AIS lesions using DLMs. Key Points: • Higher Dice similarity coefficient (DSC) in predicting acute ischemic stroke lesions was achieved by ADC thresholds combined with DWI than by DWI alone (all p <.05). • DSC had a negative association with the ADC threshold in most sizes, both hospitals, and both observers (most p <.05) and a positive association with the stroke size in all ADC thresholds, both hospitals, and both observers (all p <.001). • An ADC threshold of 0.6 × 10−3mm2/s eliminated the difference of DSC at any stroke size between observers or between hospitals (p =.07 to >.99). [ABSTRACT FROM AUTHOR] |
| Copyright of European Radiology is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Text: Availability: 1 |
|---|---|
| Header | DbId: egs DbLabel: Engineering Source An: 157956801 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Improving interobserver agreement and performance of deep learning models for segmenting acute ischemic stroke by combining DWI with optimized ADC thresholds. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Juan%2C+Chun-Jung%22">Juan, Chun-Jung</searchLink><relatesTo>1,2,3,4,5</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Lin%2C+Shao-Chieh%22">Lin, Shao-Chieh</searchLink><relatesTo>2,6</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Li%2C+Ya-Hui%22">Li, Ya-Hui</searchLink><relatesTo>2,7</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Chang%2C+Chia-Ching%22">Chang, Chia-Ching</searchLink><relatesTo>2,8</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Jeng%2C+Yi-Hung%22">Jeng, Yi-Hung</searchLink><relatesTo>2,5</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Peng%2C+Hsu-Hsia%22">Peng, Hsu-Hsia</searchLink><relatesTo>5</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Huang%2C+Teng-Yi%22">Huang, Teng-Yi</searchLink><relatesTo>9</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Chung%2C+Hsiao-Wen%22">Chung, Hsiao-Wen</searchLink><relatesTo>7,10</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Shen%2C+Wu-Chung%22">Shen, Wu-Chung</searchLink><relatesTo>2,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Tsai%2C+Chon-Haw%22">Tsai, Chon-Haw</searchLink><relatesTo>11</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Chang%2C+Ruey-Feng%22">Chang, Ruey-Feng</searchLink><relatesTo>1,7</relatesTo> (AUTHOR)<i> rfchang@csie.ntu.edu.tw</i><br /><searchLink fieldCode="AR" term="%22Liu%2C+Yi-Jui%22">Liu, Yi-Jui</searchLink><relatesTo>12</relatesTo> (AUTHOR)<i> ericliu6@ms35.hinet.net</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22European+Radiology%22">European Radiology</searchLink>. Aug2022, Vol. 32 Issue 8, p5371-5381. 11p. – Name: Subject Label: Subjects Group: Su Data: <searchLink fieldCode="DE" term="%22Ischemic+stroke%22">Ischemic stroke</searchLink><br /><searchLink fieldCode="DE" term="%22Deep+learning%22">Deep learning</searchLink><br /><searchLink fieldCode="DE" term="%22Inter-observer+reliability%22">Inter-observer reliability</searchLink><br /><searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Computer-assisted+image+analysis+%28Medicine%29%22">Computer-assisted image analysis (Medicine)</searchLink><br /><searchLink fieldCode="DE" term="%22Bland-Altman+plot%22">Bland-Altman plot</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Objectives: To examine the role of ADC threshold on agreement across observers and deep learning models (DLMs) plus segmentation performance of DLMs for acute ischemic stroke (AIS). Methods: Twelve DLMs, which were trained on DWI-ADC-ADC combination from 76 patients with AIS using 6 different ADC thresholds with ground truth manually contoured by 2 observers, were tested by additional 67 patients in the same hospital and another 78 patients in another hospital. Agreement between observers and DLMs were evaluated by Bland-Altman plot and intraclass correlation coefficient (ICC). The similarity between ground truth (GT) defined by observers and between automatic segmentation performed by DLMs was evaluated by Dice similarity coefficient (DSC). Group comparison was performed using the Mann-Whitney U test. The relationship between the DSC and ADC threshold as well as AIS lesion size was evaluated by linear regression analysis. A p <.05 was considered statistically significant. Results: Excellent interobserver agreement and intraobserver repeatability in the manual segmentation (all ICC > 0.98, p <.001) were achieved. The 95% limit of agreement was reduced from 11.23 cm2 for GT on DWI to 0.59 cm2 for prediction at an ADC threshold of 0.6 × 10−3 mm2/s combined with DWI. The segmentation performance of DLMs was improved with an overall DSC from 0.738 ± 0.214 on DWI to 0.971 ± 0.021 on an ADC threshold of 0.6 × 10−3 mm2/s combined with DWI. Conclusions: Combining an ADC threshold of 0.6 × 10−3 mm2/s with DWI reduces interobserver and inter-DLM difference and achieves best segmentation performance of AIS lesions using DLMs. Key Points: • Higher Dice similarity coefficient (DSC) in predicting acute ischemic stroke lesions was achieved by ADC thresholds combined with DWI than by DWI alone (all p <.05). • DSC had a negative association with the ADC threshold in most sizes, both hospitals, and both observers (most p <.05) and a positive association with the stroke size in all ADC thresholds, both hospitals, and both observers (all p <.001). • An ADC threshold of 0.6 × 10−3mm2/s eliminated the difference of DSC at any stroke size between observers or between hospitals (p =.07 to >.99). [ABSTRACT FROM AUTHOR] – Name: AbstractSuppliedCopyright Label: Group: Ab Data: <i>Copyright of European Radiology is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.) |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=157956801 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1007/s00330-022-08633-6 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 11 StartPage: 5371 Subjects: – SubjectFull: Ischemic stroke Type: general – SubjectFull: Deep learning Type: general – SubjectFull: Inter-observer reliability Type: general – SubjectFull: Image segmentation Type: general – SubjectFull: Computer-assisted image analysis (Medicine) Type: general – SubjectFull: Bland-Altman plot Type: general Titles: – TitleFull: Improving interobserver agreement and performance of deep learning models for segmenting acute ischemic stroke by combining DWI with optimized ADC thresholds. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Juan, Chun-Jung – PersonEntity: Name: NameFull: Lin, Shao-Chieh – PersonEntity: Name: NameFull: Li, Ya-Hui – PersonEntity: Name: NameFull: Chang, Chia-Ching – PersonEntity: Name: NameFull: Jeng, Yi-Hung – PersonEntity: Name: NameFull: Peng, Hsu-Hsia – PersonEntity: Name: NameFull: Huang, Teng-Yi – PersonEntity: Name: NameFull: Chung, Hsiao-Wen – PersonEntity: Name: NameFull: Shen, Wu-Chung – PersonEntity: Name: NameFull: Tsai, Chon-Haw – PersonEntity: Name: NameFull: Chang, Ruey-Feng – PersonEntity: Name: NameFull: Liu, Yi-Jui IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 08 Text: Aug2022 Type: published Y: 2022 Identifiers: – Type: issn-print Value: 09387994 Numbering: – Type: volume Value: 32 – Type: issue Value: 8 Titles: – TitleFull: European Radiology Type: main |
| ResultId | 1 |
