A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer
ObjectiveAccurate preoperative evaluation of rectal cancer lung metastases (RCLM) is critical for implementing precise medicine. While artificial intelligence (AI) methods have been successful in detecting liver and lymph node metastases using magnetic resonance (MR) images, research on lung metasta...
Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-02-01
|
| Series: | Frontiers in Oncology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fonc.2025.1496820/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832087401066921984 |
|---|---|
| author | Yin Li Yin Li Yin Li Shuang Li Shuang Li Ruolin Xiao Ruolin Xiao Xi Li Yongju Yi Yongju Yi Liangyou Zhang Liangyou Zhang You Zhou You Zhou Yun Wan Chenhua Wei Liming Zhong Liming Zhong Wei Yang Wei Yang Lin Yao Lin Yao Lin Yao |
| author_facet | Yin Li Yin Li Yin Li Shuang Li Shuang Li Ruolin Xiao Ruolin Xiao Xi Li Yongju Yi Yongju Yi Liangyou Zhang Liangyou Zhang You Zhou You Zhou Yun Wan Chenhua Wei Liming Zhong Liming Zhong Wei Yang Wei Yang Lin Yao Lin Yao Lin Yao |
| author_sort | Yin Li |
| collection | DOAJ |
| description | ObjectiveAccurate preoperative evaluation of rectal cancer lung metastases (RCLM) is critical for implementing precise medicine. While artificial intelligence (AI) methods have been successful in detecting liver and lymph node metastases using magnetic resonance (MR) images, research on lung metastases is still limited. Utilizing MR images to classify RCLM could potentially reduce ionizing radiation exposure and the costs associated with chest CT in patients without metastases. This study aims to develop and validate a transformer-based deep learning (DL) model based on pelvic MR images, integrated with clinical features, to predict RCLM.MethodsA total of 819 patients with histologically confirmed rectal cancer who underwent preoperative pelvis MRI and carcinoembryonic antigen (CEA) tests were enrolled. Six state-of-the-art DL methods (Resnet18, EfficientNetb0, MobileNet, ShuffleNet, DenseNet, and our transformer-based model) were trained and tested on T2WI and DWI to predict RCLM. The predictive performance was assessed using the receiver operating characteristic (ROC) curve.ResultsOur transformer-based DL model achieved impressive results in the independent test set, with an AUC of 83.74% (95% CI, 72.60%-92.83%), a sensitivity of 80.00%, a specificity of 78.79%, and an accuracy of 79.01%. Specifically, for stage T4 and N2 rectal cancer cases, the model achieved AUCs of 96.67% (95% CI, 87.14%-100%, 93.33% sensitivity, 89.04% specificity, 94.74% accuracy), and 96.83% (95% CI, 88.67%-100%, 100% sensitivity, 83.33% specificity, 88.00% accuracy) respectively, in predicting RCLM. Our DL model showed a better predictive performance than other state-of-the-art DL methods.ConclusionThe superior performance demonstrates the potential of our work for predicting RCLM, suggesting its potential assistance in personalized treatment and follow-up plans. |
| format | Article |
| id | doaj-art-9fb0e6f4233b4a0eb7bfd13d878ac15d |
| institution | Kabale University |
| issn | 2234-943X |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Oncology |
| spelling | doaj-art-9fb0e6f4233b4a0eb7bfd13d878ac15d2025-02-06T05:21:53ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2025-02-011510.3389/fonc.2025.14968201496820A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancerYin Li0Yin Li1Yin Li2Shuang Li3Shuang Li4Ruolin Xiao5Ruolin Xiao6Xi Li7Yongju Yi8Yongju Yi9Liangyou Zhang10Liangyou Zhang11You Zhou12You Zhou13Yun Wan14Chenhua Wei15Liming Zhong16Liming Zhong17Wei Yang18Wei Yang19Lin Yao20Lin Yao21Lin Yao22Department of Information, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaBiomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaDepartment of Information, The Sixth Affiliated Hospital, Sun Yat-sen University Yuexi Hospital, Maoming, ChinaBiomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaDepartment of General Practice, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Key Laboratory of Medical Image Processing, Guangzhou, ChinaDepartment of Radiology, The Sixth Affiliated Hospital, Sun Yat-sen University Yuexi Hospital, Maoming, ChinaDepartment of Information, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaBiomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaDepartment of Information, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaBiomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaDepartment of Information, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaBiomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaDepartment of Radiology, The Sixth Affiliated Hospital, Sun Yat-sen University Yuexi Hospital, Maoming, ChinaDepartment of Information, The Sixth Affiliated Hospital, Sun Yat-sen University Yuexi Hospital, Maoming, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Key Laboratory of Medical Image Processing, Guangzhou, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaGuangdong Provincial Key Laboratory of Medical Image Processing, Guangzhou, ChinaBiomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaDepartment of Information, The Sixth Affiliated Hospital, Sun Yat-sen University Yuexi Hospital, Maoming, ChinaDepartment of General Practice, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaObjectiveAccurate preoperative evaluation of rectal cancer lung metastases (RCLM) is critical for implementing precise medicine. While artificial intelligence (AI) methods have been successful in detecting liver and lymph node metastases using magnetic resonance (MR) images, research on lung metastases is still limited. Utilizing MR images to classify RCLM could potentially reduce ionizing radiation exposure and the costs associated with chest CT in patients without metastases. This study aims to develop and validate a transformer-based deep learning (DL) model based on pelvic MR images, integrated with clinical features, to predict RCLM.MethodsA total of 819 patients with histologically confirmed rectal cancer who underwent preoperative pelvis MRI and carcinoembryonic antigen (CEA) tests were enrolled. Six state-of-the-art DL methods (Resnet18, EfficientNetb0, MobileNet, ShuffleNet, DenseNet, and our transformer-based model) were trained and tested on T2WI and DWI to predict RCLM. The predictive performance was assessed using the receiver operating characteristic (ROC) curve.ResultsOur transformer-based DL model achieved impressive results in the independent test set, with an AUC of 83.74% (95% CI, 72.60%-92.83%), a sensitivity of 80.00%, a specificity of 78.79%, and an accuracy of 79.01%. Specifically, for stage T4 and N2 rectal cancer cases, the model achieved AUCs of 96.67% (95% CI, 87.14%-100%, 93.33% sensitivity, 89.04% specificity, 94.74% accuracy), and 96.83% (95% CI, 88.67%-100%, 100% sensitivity, 83.33% specificity, 88.00% accuracy) respectively, in predicting RCLM. Our DL model showed a better predictive performance than other state-of-the-art DL methods.ConclusionThe superior performance demonstrates the potential of our work for predicting RCLM, suggesting its potential assistance in personalized treatment and follow-up plans.https://www.frontiersin.org/articles/10.3389/fonc.2025.1496820/fullrectal cancerlung metastasesMRItransformerdeep learning |
| spellingShingle | Yin Li Yin Li Yin Li Shuang Li Shuang Li Ruolin Xiao Ruolin Xiao Xi Li Yongju Yi Yongju Yi Liangyou Zhang Liangyou Zhang You Zhou You Zhou Yun Wan Chenhua Wei Liming Zhong Liming Zhong Wei Yang Wei Yang Lin Yao Lin Yao Lin Yao A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer Frontiers in Oncology rectal cancer lung metastases MRI transformer deep learning |
| title | A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer |
| title_full | A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer |
| title_fullStr | A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer |
| title_full_unstemmed | A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer |
| title_short | A pelvis MR transformer-based deep learning model for predicting lung metastases risk in patients with rectal cancer |
| title_sort | pelvis mr transformer based deep learning model for predicting lung metastases risk in patients with rectal cancer |
| topic | rectal cancer lung metastases MRI transformer deep learning |
| url | https://www.frontiersin.org/articles/10.3389/fonc.2025.1496820/full |
| work_keys_str_mv | AT yinli apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yinli apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yinli apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT shuangli apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT shuangli apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT ruolinxiao apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT ruolinxiao apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT xili apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yongjuyi apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yongjuyi apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT liangyouzhang apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT liangyouzhang apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT youzhou apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT youzhou apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yunwan apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT chenhuawei apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT limingzhong apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT limingzhong apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT weiyang apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT weiyang apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT linyao apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT linyao apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT linyao apelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yinli pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yinli pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yinli pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT shuangli pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT shuangli pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT ruolinxiao pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT ruolinxiao pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT xili pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yongjuyi pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yongjuyi pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT liangyouzhang pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT liangyouzhang pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT youzhou pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT youzhou pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT yunwan pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT chenhuawei pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT limingzhong pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT limingzhong pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT weiyang pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT weiyang pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT linyao pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT linyao pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer AT linyao pelvismrtransformerbaseddeeplearningmodelforpredictinglungmetastasesriskinpatientswithrectalcancer |