Eur Radiol: How to judge whether lung cancer has invaded the visceral pleura?

Pulmonary pleural infiltration (VPI) is one of the main poor prognostic indicators after adjusting the pathological T classification of lung cancer and the histological types of T1 and T2 tumors.

Pulmonary pleural infiltration (VPI) is one of the main poor prognostic indicators after adjusting the pathological T classification of lung cancer and the histological types of T1 and T2 tumors.

However, at present, it is difficult to accurately diagnose VPI based on imaging features before surgery .

Recently, a study published in the journal European Radiology developed and verified a preoperative CT-based deep learning model for predicting the VPI of early lung cancer.

In this retrospective study, data set 1 (for training, tuning, and internal verification) included 676 patients with clinical stage IA lung adenocarcinoma who underwent surgical resection between 2009 and 2015.

In this retrospective study, data set 1 (for training, tuning, and internal verification) included 676 patients with clinical stage IA lung adenocarcinoma who underwent surgical resection between 2009 and 2015.

The area under the receiver operating characteristic curve (AUC) of this model is 0.

Table 1 The area under the receiver operating characteristic curve of the model and three committee-certified radiologists

Table 1 The area under the receiver operating characteristic curve of the model and three committee-certified radiologists

Figure 1 is used to explain the gradient-weighted class activation mapping of the model output.

Figure 1 is used to explain the gradient-weighted class activation mapping of the model output.

 

    This deep learning model based on preoperative CT extracts independent VPI diagnostic information for surgically adapted patients, and shows expert-level diagnostic performance.

    This deep learning model based on preoperative CT extracts independent VPI diagnostic information for surgically adapted patients, and shows expert-level diagnostic performance.

Original source:

Hyewon Choi,Hyungjin Kim,Wonju Hong,et al.

Hyewon Choi,Hyungjin Kim,Wonju Hong,et al.
Prediction of visceral pleural invasion in lung cancer on CT: deep learning model achieves a radiologist-level performance with adaptive sensitivity and specificity to clinical needs.
DOI: org/10.
1007/s00330-020-07431-2">10.
1007/s00330-020-07431 -2org/10.
1007/s00330-020-07431-2"> 10.
1007/s00330-020-07431-2 Leave a message here