Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer
Abstract Background Preoperative pelvic lymph node metastasis (PLNM) prediction can help clinicians determine whether to perform pelvic lymph node dissection (PLND). The purpose of this research is to explore the feasibility of diffusion-weighted imaging (DWI)-based radiomics for preoperative PLNM p...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2022-11-01
|
| Series: | BMC Medical Imaging |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12880-022-00905-3 |
| _version_ | 1811233832417886208 |
|---|---|
| author | Xiang Liu Jingyi Tian Jingyun Wu Yaofeng Zhang Xiangpeng Wang Xiaodong Zhang Xiaoying Wang |
| author_facet | Xiang Liu Jingyi Tian Jingyun Wu Yaofeng Zhang Xiangpeng Wang Xiaodong Zhang Xiaoying Wang |
| author_sort | Xiang Liu |
| collection | DOAJ |
| description | Abstract Background Preoperative pelvic lymph node metastasis (PLNM) prediction can help clinicians determine whether to perform pelvic lymph node dissection (PLND). The purpose of this research is to explore the feasibility of diffusion-weighted imaging (DWI)-based radiomics for preoperative PLNM prediction in PCa patients at the nodal level. Methods The preoperative MR images of 1116 pathologically confirmed lymph nodes (LNs) from 84 PCa patients were enrolled. The subjects were divided into a primary cohort (67 patients with 192 positive and 716 negative LNs) and a held-out cohort (17 patients with 43 positive and 165 negative LNs) at a 4:1 ratio. Two preoperative pelvic lymph node metastasis (PLNM) prediction models were constructed based on automatic LN segmentation with quantitative radiological LN features alone (Model 1) and combining radiological and radiomics features (Model 2) via multiple logistic regression. The visual assessments of junior (Model 3) and senior (Model 4) radiologists were compared. Results No significant difference was found between the area under the curve (AUCs) of Models 1 and 2 (0.89 vs. 0.90; P = 0.573) in the held-out cohort. Model 2 showed the highest AUC (0.83, 95% CI 0.76, 0.89) for PLNM prediction in the LN subgroup with a short diameter ≤ 10 mm compared with Model 1 (0.78, 95% CI 0.70, 0.84), Model 3 (0.66, 95% CI 0.52, 0.77), and Model 4 (0.74, 95% CI 0.66, 0.88). The nomograms of Models 1 and 2 yielded C-index values of 0.804 and 0.910, respectively, in the held-out cohort. The C-index of the nomogram analysis (0.91) and decision curve analysis (DCA) curves confirmed the clinical usefulness and benefit of Model 2. Conclusions A DWI-based radiomics nomogram incorporating the LN radiomics signature with quantitative radiological features is promising for PLNM prediction in PCa patients, particularly for normal-sized LNM. |
| first_indexed | 2024-04-12T11:25:57Z |
| format | Article |
| id | doaj.art-ee4e56d3497642c88fcf41ed52d04167 |
| institution | Directory Open Access Journal |
| issn | 1471-2342 |
| language | English |
| last_indexed | 2024-04-12T11:25:57Z |
| publishDate | 2022-11-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Imaging |
| spelling | doaj.art-ee4e56d3497642c88fcf41ed52d041672022-12-22T03:35:13ZengBMCBMC Medical Imaging1471-23422022-11-0122111310.1186/s12880-022-00905-3Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancerXiang Liu0Jingyi Tian1Jingyun Wu2Yaofeng Zhang3Xiangpeng Wang4Xiaodong Zhang5Xiaoying Wang6Department of Radiology, Peking University First HospitalDepartment of Radiology, Beijing Water Conservancy HospitalDepartment of Radiology, Peking University First HospitalBeijing Smart Tree Medical Technology Co. LtdBeijing Smart Tree Medical Technology Co. LtdDepartment of Radiology, Peking University First HospitalDepartment of Radiology, Peking University First HospitalAbstract Background Preoperative pelvic lymph node metastasis (PLNM) prediction can help clinicians determine whether to perform pelvic lymph node dissection (PLND). The purpose of this research is to explore the feasibility of diffusion-weighted imaging (DWI)-based radiomics for preoperative PLNM prediction in PCa patients at the nodal level. Methods The preoperative MR images of 1116 pathologically confirmed lymph nodes (LNs) from 84 PCa patients were enrolled. The subjects were divided into a primary cohort (67 patients with 192 positive and 716 negative LNs) and a held-out cohort (17 patients with 43 positive and 165 negative LNs) at a 4:1 ratio. Two preoperative pelvic lymph node metastasis (PLNM) prediction models were constructed based on automatic LN segmentation with quantitative radiological LN features alone (Model 1) and combining radiological and radiomics features (Model 2) via multiple logistic regression. The visual assessments of junior (Model 3) and senior (Model 4) radiologists were compared. Results No significant difference was found between the area under the curve (AUCs) of Models 1 and 2 (0.89 vs. 0.90; P = 0.573) in the held-out cohort. Model 2 showed the highest AUC (0.83, 95% CI 0.76, 0.89) for PLNM prediction in the LN subgroup with a short diameter ≤ 10 mm compared with Model 1 (0.78, 95% CI 0.70, 0.84), Model 3 (0.66, 95% CI 0.52, 0.77), and Model 4 (0.74, 95% CI 0.66, 0.88). The nomograms of Models 1 and 2 yielded C-index values of 0.804 and 0.910, respectively, in the held-out cohort. The C-index of the nomogram analysis (0.91) and decision curve analysis (DCA) curves confirmed the clinical usefulness and benefit of Model 2. Conclusions A DWI-based radiomics nomogram incorporating the LN radiomics signature with quantitative radiological features is promising for PLNM prediction in PCa patients, particularly for normal-sized LNM.https://doi.org/10.1186/s12880-022-00905-3Prostate cancerLymph nodesRadiomicsNomogramMagnetic resonance imaging |
| spellingShingle | Xiang Liu Jingyi Tian Jingyun Wu Yaofeng Zhang Xiangpeng Wang Xiaodong Zhang Xiaoying Wang Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer BMC Medical Imaging Prostate cancer Lymph nodes Radiomics Nomogram Magnetic resonance imaging |
| title | Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer |
| title_full | Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer |
| title_fullStr | Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer |
| title_full_unstemmed | Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer |
| title_short | Utility of diffusion weighted imaging-based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer |
| title_sort | utility of diffusion weighted imaging based radiomics nomogram to predict pelvic lymph nodes metastasis in prostate cancer |
| topic | Prostate cancer Lymph nodes Radiomics Nomogram Magnetic resonance imaging |
| url | https://doi.org/10.1186/s12880-022-00905-3 |
| work_keys_str_mv | AT xiangliu utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer AT jingyitian utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer AT jingyunwu utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer AT yaofengzhang utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer AT xiangpengwang utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer AT xiaodongzhang utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer AT xiaoyingwang utilityofdiffusionweightedimagingbasedradiomicsnomogramtopredictpelviclymphnodesmetastasisinprostatecancer |