Virtual monoenergetic imaging predicting Ki-67 expression in lung cancer

Ultrasound-based artificial intelligence model for prediction of Ki-67 proliferation index in soft tissue tumors

RATIONALE AND OBJECTIVES

To investigate the value of deep learning (DL) combined with radiomics and clinical and imaging features in predicting the Ki-67 proliferation index of soft tissue tumors (STTs).

MATERIALS AND METHODS

In this retrospective study, a total of 394 patients with STTs admitted from January 2021 to December 2023 in two separate hospitals were collected. Hospital-1 was the training cohort (323 cases, of which 89 and 234 were high and low Ki-67, respectively) and Hospital-2 was the external validation cohort (71 cases, of which 23 and 48 were high and low Ki-67, respectively). Clinical and ultrasound characteristics including age, sex, tumor size, morphology, margins, internal echoes and blood flow were assessed. Risk factors with significant correlations were screened by univariate and multivariate logistic regression analyses. After extracting the radiomics and DL features, the feature fusion model is constructed by Support Vector Machine. The prediction results obtained from separate clinical features, radiomics features and DL features were combined to construct decision fusion models. Finally, the DeLong test was used to compare whether the AUCs between the models were significantly different.

RESULTS

The three feature fusion models and three decision fusion models constructed demonstrated excellent diagnostic performance in predicting Ki-67 expression levels in STTs. Among them, the feature fusion model based on clinical, radiomics, and DL performed the best with an AUC of 0.911 (95% CI: 0.886-0.935) in the training cohort and 0.923 (95% CI: 0.873-0.972) in the validation cohort, and proved to be well-calibrated and clinically useful. The DeLong test showed that the decision fusion models based on clinical, radiomics and DL performed significantly worse than the three feature fusion models on the validation set. There was no statistical difference in diagnostic performance between the other models.

CONCLUSION

The ultrasound-based fusion model of clinical, radiomics, and DL features showed good performance in predicting Ki-67 expression levels in STTs.

Quantitative parameters of dual-layer spectral detector computed tomography for evaluating Ki-67 and human epidermal growth factor receptor 2 expression in colorectal adenocarcinoma

Background

Ki-67 and human epidermal growth factor receptor 2 (HER2) are key biomarkers in evaluating the prognosis of colorectal adenocarcinoma (CRAC). The purpose of this study was to investigate the value of quantitative parameters in dual-layer spectral detector computed tomography (SDCT) for evaluating the expression of Ki-67 and HER2 in CRAC.

Methods

In this retrospective, cross-sectional study, 88 eligible patients with pathologically confirmed CRAC were selected from Taicang Hospital of Traditional Chinese Medicine between May 2021 and April 2023. The study participants underwent enhanced SDCT of the whole abdomen within 2 weeks before to surgery, did not receive antitumor therapy, and had complete immunohistochemical (IHC) indexes. Patients with nonadenocarcinoma pathologic types, poor quality of spectral CT images, or no complete immunohistochemistry results were excluded. Spectral parameters including CT values at 40 and 100 keV, effective atomic number, iodine concentration (IC), the slope of the spectral Hounsfield unit (HU) curve (λHU), and normalized iodine concentration (NIC) in the arterial phase (AP) and venous phase (VP) were analyzed for their value in distinguishing between the high and low expression of Ki-67 and HER2-positive and -negative status in CRAC. The statistical significance of the SDCT parameters between the different groups of Ki-67 expression and those of HER2 status was assessed with the Mann-Whitney test. Spearman correlation analysis was used to analyze the correlation between the SDCT parameters and the extent of Ki-67 expression and HER2 expression status. The receiver operating characteristic (ROC) curve was used, and the area under the curve (AUC) was calculated.

Results

The SDCT parameters of CT values at 40 keV, effective atomic number, IC, and the λHU in the VP showed significant differences between the Ki-67 high- and low-expression groups in CRAC (P=0.035, P=0.041, P=0.036, and P=0.044, respectively), with AUCs of 0.639 [95% confidence interval (CI): 0.512-0.766], 0.634 (95% CI: 0.508-0.761), 0.638 (95% CI: 0.510-0.766), and 0.633 (95% CI: 0.504-0.762), respectively. The expression of CRAC Ki-67 was positively correlated with CT values at 40 keV (r=0.227; P=0.034), effective atomic number (r=0.219; P=0.040), IC (r=0.225; P=0.035), and the λHU in VP (r=0.216; P=0.043). SDCT parameter values showed no statistical difference between negative and positive expression in HER2 (all P values >0.05). There was no significant correlation between SDCT parameters and the expression of HER2 in CRAC (all P values >0.05).

Conclusions

The quantitative parameters of SDCT in the VP provide valuable information for distinguishing between the low expression and high expression of Ki-67 in CRAC.