Classification prediction of pancreatic cystic neoplasms based on radiomics deep learning models
Abstract Background: Preoperative prediction of pancreatic cystic neoplasm (PCN) differentiation has significant value for implementation of personalized diagnosis and treatment plan. This study aims to build radiomics deep learning (DL) models using computed tomography (CT) data for preoperative differential diagnosis of common cystic tumors of pancreas.Methods: Clinical and CT data of 193 patients with PCN was collected for the study. Among these patients, 99 were pathologically diagnosed with pancreatic serous cystadenoma (SCA), 55 with mucinous cystadenoma (MCA) and 39 with intraductal papillary mucinous neoplasm (IPMN). The regions of interest (ROIs) were obtained based on manual image segmentation of CT scans. The radiomics model and radiomics-DL model were constructed using support vector machines (SVMs). Moreover, in combination with clinical and radiological features, the best combined feature set was obtained by Akaike information criterion (AIC) analysis, and the fused model was constructed using logistic regression.Results: For SCA differential diagnosis, the fused model performed the best and obtained average area under the curve (AUC) of 0.916. It had a best feature set including position, polycystic features (≥6), cystic wall calcification, pancreatic duct dilatation and radiomics-DL score. As for MCA and IPMN differential diagnosis, the fused model with AUC of 0.973 had a best feature set including age, communicate with pancreatic duct and radiomics scores.Conclusions: The radiomics, radiomics-DL and fused model based on CT images have a favorable differential diagnostic performance for SCA, MCA and IPMN. These findings may be beneficial to the formulation of individualized management strategies..
| Medienart: |
|
|---|
Preprint| Erscheinungsjahr: |
2022 |
|---|---|
| Erschienen: |
2022 |
| Enthalten in: |
ResearchSquare.com - (2022) vom: 05. Mai Zur Gesamtaufnahme - year:2022 |
|---|
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Liang, Wenjie [VerfasserIn] |
|---|
| Links: |
Volltext [kostenfrei] |
|---|
| doi: |
10.21203/rs.3.rs-1593820/v1 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
XRA035926317 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | XRA035926317 | ||
| 003 | DE-627 | ||
| 005 | 20230429182650.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 220506s2022 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.21203/rs.3.rs-1593820/v1 |2 doi | |
| 035 | |a (DE-627)XRA035926317 | ||
| 035 | |a (ResearchSquare)10.21203/rs.3.rs-1593820/v1 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | |a 570 |q DE-84 | |
| 100 | 1 | |a Liang, Wenjie |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Classification prediction of pancreatic cystic neoplasms based on radiomics deep learning models |
| 264 | 1 | |c 2022 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a Abstract Background: Preoperative prediction of pancreatic cystic neoplasm (PCN) differentiation has significant value for implementation of personalized diagnosis and treatment plan. This study aims to build radiomics deep learning (DL) models using computed tomography (CT) data for preoperative differential diagnosis of common cystic tumors of pancreas.Methods: Clinical and CT data of 193 patients with PCN was collected for the study. Among these patients, 99 were pathologically diagnosed with pancreatic serous cystadenoma (SCA), 55 with mucinous cystadenoma (MCA) and 39 with intraductal papillary mucinous neoplasm (IPMN). The regions of interest (ROIs) were obtained based on manual image segmentation of CT scans. The radiomics model and radiomics-DL model were constructed using support vector machines (SVMs). Moreover, in combination with clinical and radiological features, the best combined feature set was obtained by Akaike information criterion (AIC) analysis, and the fused model was constructed using logistic regression.Results: For SCA differential diagnosis, the fused model performed the best and obtained average area under the curve (AUC) of 0.916. It had a best feature set including position, polycystic features (≥6), cystic wall calcification, pancreatic duct dilatation and radiomics-DL score. As for MCA and IPMN differential diagnosis, the fused model with AUC of 0.973 had a best feature set including age, communicate with pancreatic duct and radiomics scores.Conclusions: The radiomics, radiomics-DL and fused model based on CT images have a favorable differential diagnostic performance for SCA, MCA and IPMN. These findings may be beneficial to the formulation of individualized management strategies. | ||
| 700 | 1 | |a Wang, Pan |e verfasserin |4 aut | |
| 700 | 1 | |a Tian, Wuwei |e verfasserin |4 aut | |
| 700 | 1 | |a Wang, Yubizhuo |e verfasserin |4 aut | |
| 700 | 1 | |a Zhang, Hongbin |e verfasserin |4 aut | |
| 700 | 1 | |a Ruan, Shijian |e verfasserin |4 aut | |
| 700 | 1 | |a Shao, Jiayuan |e verfasserin |4 aut | |
| 700 | 1 | |a Wang, Yifan |e verfasserin |4 aut | |
| 700 | 1 | |a Zhang, Xiuming |e verfasserin |4 aut | |
| 700 | 1 | |a Huang, Danjiang |e verfasserin |4 aut | |
| 700 | 1 | |a Ding, Yong |e verfasserin |4 aut | |
| 700 | 1 | |a Bai, Xueli |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t ResearchSquare.com |g (2022) vom: 05. Mai |
| 773 | 1 | 8 | |g year:2022 |g day:05 |g month:05 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.21203/rs.3.rs-1593820/v1 |z kostenfrei |3 Volltext |
| 912 | |a GBV_XRA | ||
| 912 | |a SSG-OLC-PHA | ||
| 951 | |a AR | ||
| 952 | |j 2022 |b 05 |c 05 | ||
| 953 | |2 045F |a 570 | ||