Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study
Background: The purpose of this study was to assess the volume of the pulmonary nodules and masses on serial chest X-rays (CXRs) from deep-learning-based automatic detection algorithm (DLAD)-based parameters. Methods: In a retrospective single-institutional study, 72 patients, who obtained serial CX...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/13/12/2060 |
| _version_ | 1797595316238680064 |
|---|---|
| author | Chae Young Lim Yoon Ki Cha Myung Jin Chung Subin Park Soyoung Park Jung Han Woo Jong Hee Kim |
| author_facet | Chae Young Lim Yoon Ki Cha Myung Jin Chung Subin Park Soyoung Park Jung Han Woo Jong Hee Kim |
| author_sort | Chae Young Lim |
| collection | DOAJ |
| description | Background: The purpose of this study was to assess the volume of the pulmonary nodules and masses on serial chest X-rays (CXRs) from deep-learning-based automatic detection algorithm (DLAD)-based parameters. Methods: In a retrospective single-institutional study, 72 patients, who obtained serial CXRs (<i>n</i> = 147) for pulmonary nodules or masses with corresponding chest CT images as the reference standards, were included. A pre-trained DLAD based on a convolutional neural network was developed to detect and localize nodules using 13,710 radiographs and to calculate a localization map and the derived parameters (e.g., the area and mean probability value of pulmonary nodules) for each CXR, including serial follow-ups. For validation, reference 3D CT volumes were measured semi-automatically. Volume prediction models for pulmonary nodules were established through univariable or multivariable, and linear or non-linear regression analyses with the parameters. A polynomial regression analysis was performed as a method of a non-linear regression model. Results: Of the 147 CXRs and 208 nodules of 72 patients, the mean volume of nodules or masses was measured as 9.37 ± 11.69 cm<sup>3</sup> (mean ± standard deviation). The area and CT volume demonstrated a linear correlation of moderate strength (i.e., R = 0.58, RMSE: 9449.9 mm<sup>3</sup> m<sup>3</sup> in a linear regression analysis). The area and mean probability values exhibited a strong linear correlation (R = 0.73). The volume prediction performance based on a multivariable regression model was best with a mean probability and unit-adjusted area (i.e., RMSE: 7975.6 mm<sup>3</sup>, the smallest among the other variable parameters). Conclusions: The prediction model with the area and the mean probability based on the DLAD showed a rather accurate quantitative estimation of pulmonary nodule or mass volume and the change in serial CXRs. |
| first_indexed | 2024-03-11T02:34:36Z |
| format | Article |
| id | doaj.art-42928d9315cf48a3bc9f34deadb80fea |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-11T02:34:36Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-42928d9315cf48a3bc9f34deadb80fea2023-11-18T10:00:32ZengMDPI AGDiagnostics2075-44182023-06-011312206010.3390/diagnostics13122060Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary StudyChae Young Lim0Yoon Ki Cha1Myung Jin Chung2Subin Park3Soyoung Park4Jung Han Woo5Jong Hee Kim6Department of Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul 06351, Republic of KoreaDepartment of Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul 06351, Republic of KoreaDepartment of Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul 06351, Republic of KoreaDepartment of Health Sciences and Technology, SAIHST, Sungkyunkwan University, Suwon 16419, Republic of KoreaDepartment of Health Sciences and Technology, SAIHST, Sungkyunkwan University, Suwon 16419, Republic of KoreaDepartment of Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul 06351, Republic of KoreaDepartment of Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul 06351, Republic of KoreaBackground: The purpose of this study was to assess the volume of the pulmonary nodules and masses on serial chest X-rays (CXRs) from deep-learning-based automatic detection algorithm (DLAD)-based parameters. Methods: In a retrospective single-institutional study, 72 patients, who obtained serial CXRs (<i>n</i> = 147) for pulmonary nodules or masses with corresponding chest CT images as the reference standards, were included. A pre-trained DLAD based on a convolutional neural network was developed to detect and localize nodules using 13,710 radiographs and to calculate a localization map and the derived parameters (e.g., the area and mean probability value of pulmonary nodules) for each CXR, including serial follow-ups. For validation, reference 3D CT volumes were measured semi-automatically. Volume prediction models for pulmonary nodules were established through univariable or multivariable, and linear or non-linear regression analyses with the parameters. A polynomial regression analysis was performed as a method of a non-linear regression model. Results: Of the 147 CXRs and 208 nodules of 72 patients, the mean volume of nodules or masses was measured as 9.37 ± 11.69 cm<sup>3</sup> (mean ± standard deviation). The area and CT volume demonstrated a linear correlation of moderate strength (i.e., R = 0.58, RMSE: 9449.9 mm<sup>3</sup> m<sup>3</sup> in a linear regression analysis). The area and mean probability values exhibited a strong linear correlation (R = 0.73). The volume prediction performance based on a multivariable regression model was best with a mean probability and unit-adjusted area (i.e., RMSE: 7975.6 mm<sup>3</sup>, the smallest among the other variable parameters). Conclusions: The prediction model with the area and the mean probability based on the DLAD showed a rather accurate quantitative estimation of pulmonary nodule or mass volume and the change in serial CXRs.https://www.mdpi.com/2075-4418/13/12/2060pulmonary noduleschest X-raysdeep-learningconvolutional neural networksvolume estimation |
| spellingShingle | Chae Young Lim Yoon Ki Cha Myung Jin Chung Subin Park Soyoung Park Jung Han Woo Jong Hee Kim Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study Diagnostics pulmonary nodules chest X-rays deep-learning convolutional neural networks volume estimation |
| title | Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study |
| title_full | Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study |
| title_fullStr | Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study |
| title_full_unstemmed | Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study |
| title_short | Estimating the Volume of Nodules and Masses on Serial Chest Radiography Using a Deep-Learning-Based Automatic Detection Algorithm: A Preliminary Study |
| title_sort | estimating the volume of nodules and masses on serial chest radiography using a deep learning based automatic detection algorithm a preliminary study |
| topic | pulmonary nodules chest X-rays deep-learning convolutional neural networks volume estimation |
| url | https://www.mdpi.com/2075-4418/13/12/2060 |
| work_keys_str_mv | AT chaeyounglim estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy AT yoonkicha estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy AT myungjinchung estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy AT subinpark estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy AT soyoungpark estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy AT junghanwoo estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy AT jongheekim estimatingthevolumeofnodulesandmassesonserialchestradiographyusingadeeplearningbasedautomaticdetectionalgorithmapreliminarystudy |