European Radiology: Multiparametric MRI-based radiomics for preoperative prediction of extraluminal vein invasion in rectal cancer

Colorectal cancer is the second leading cause of cancer-related death worldwide, with rectal cancer alone accounting for one-third of these cases
.

Extravena cava invasion (EMVI) refers to blood vessels located outside of the inherent muscle in the mesorectal fat in which tumor tissue invades .

Colorectal cancer is the second leading cause of cancer-related death worldwide, with rectal cancer alone accounting for one-third of these cases

Multiparametric magnetic resonance imaging (mpMRI) is the imaging method of choice for early noninvasive assessment and EMVI detection of rectal cancer

Radiomics analyzes the correlation between radiological features and pathological data

Recently, a study published in the journal European Radiology applied mpMRI-based radiomics for preoperative prediction of EMVI status in rectal cancer, using different machine learning algorithms to establish optimal radiomics signatures, and combining radiomics with radiomics.
He developed and validated a combined model for EMVI assessment with histological , clinical, and imaging features, providing technical support for prognostic prediction in the formulation of individualized treatment regimens for rectal cancer patients .
 

This study retrospectively analyzed the clinical and imaging data of 317 patients with rectal cancer

The Bayesian-based radiomic features performed well in both training and test sets, with AUCs of 0.

Figure two cases of rectal cancer, histopathological examination showed EMVI positive .
ad One case had significant EMVI on MRI, which was also detected by the joint model .
eh Another is the EMVI detected by the joint model, which is not evident on MRI .
 

Figure two cases of rectal cancer, histopathological examination showed EMVI positive .

ad One case had significant EMVI on MRI, which was also detected by the joint model .
eh Another is the EMVI detected by the joint model, which is not evident on MRI .
Figure two cases of rectal cancer, histopathological examination showed EMVI positive .
ad One case had significant EMVI on MRI, which was also detected by the joint model .
eh Another is the EMVI detected by the joint model, which is not evident on MRI .
 

This study demonstrates that mpMRI-based radiomics can improve the diagnostic performance of preoperative EMVI prediction in patients with rectal cancer, especially for inexperienced radiologists and residents .

Visual nomogram based on radiomic features is a clinically valuable imaging tool that can largely avoid misdiagnosis caused by inexperienced radiologists .
The findings also provide important evidence for future risk stratification of rectal cancer using combined models .

This study demonstrates that mpMRI-based radiomics can improve the diagnostic performance of preoperative EMVI prediction in patients with rectal cancer, especially for inexperienced radiologists and residents .
Visual nomogram based on radiomic features is a clinically valuable imaging tool that can largely avoid misdiagnosis caused by inexperienced radiologists .
The findings also provide important evidence for future risk stratification of rectal cancer using combined models .

Original source :

Original source :

Zhenyu Shu , Dewang Mao , Qiaowei Song .

Multiparameter MRI-based radiomics for preoperative prediction of extramural venous invasion in rectal cancer .
DOI: 10.
1007/s00330-021-08242-9

Zhenyu Shu Dewang Mao Qiaowei Song Multiparameter MRI-based radiomics for preoperative prediction of extramural venous invasion in rectal cancer 10.
1007/s00330-021-08242-9 10.
1007/s00330-021-08242-9 Leave a comment here