Positron Emission Tomography/Magnetic Resonance Imaging Radiomics in Predicting Lung Adenocarcinoma and Squamous Cell Carcinoma

Abstract

ObjectiveTo investigate the diagnostic value of positron emission tomography (PET)/magnetic resonance imaging (MRI) radiomics in predicting the histological classification of lung adenocarcinoma and lung squamous cell carcinoma.MethodsPET/MRI radiomics and clinical data were retrospectively collected from 61 patients with lung cancer. According to the pathological results of surgery or fiberscope, patients were divided into two groups, lung adenocarcinoma and squamous cell carcinoma group, which were set as positive for adenocarcinoma (40 cases) and negative for squamous cell carcinoma (21 cases). The radiomics characteristics most related to lung cancer classification were calculated and selected using radiomics software, and the two lung cancer groups were randomly assigned into a training set (70%) and a test set (30%). Maximum relevance and minimum redundancy (mRMR) and least absolute shrinkage and selection operator (LASSO) methods in the uAI Research Portal software (United Imaging Intelligence, China) were used to select the desired characteristics from 2600 features extracted from MRI and PET. Eight optimal features were finally retained through 5-fold cross-validation, and a PET/MRI fusion model was constructed. The predictive ability of this model was evaluated by the difference in area under the curve (AUC) obtained from the receiver operating characteristic (ROC) curve.ResultsAUC of PET/MRI model for the training group and test group were 0.886 (0.787-0.985) and 0.847 (0.648-1.000), respectively. PET/MRI radiomics features revealed different degrees of correlation with the classification of lung adenocarcinoma and squamous cell carcinoma, with significant differences.ConclusionThe prediction model constructed based on PET/MRI radiomics features can predict the preoperative histological classification of lung adenocarcinoma and squamous cell carcinoma without seminality and repeatability. It can also provide an objective basis for accurate clinical diagnosis and individualized treatment, thus having important guiding significance for clinical treatment.

The value of combined PET/MRI, CT and clinical metabolic parameters in differentiating lung adenocarcinoma from squamous cell carcinoma

“…The results were consistent with those of Orlhac et al (21), who found that scaly cell carcinoma had lower homogeneity and higher entropy by comparing the texture characteristics of ADC and SCC. Moreover, it was consistent with our previous findings (15) in that PET/MRI was used to create a prediction model for the pathological subtypes of ADC and lung SCC. It was also found that the GLSZM-GLN feature value accounted for the maximum weight ratio, indicating that ADC was more homogeneous than SCC.…”

“…Caiyue et al (28) found that the machine learning-integrated 18 F-FDG PET/CT radiomics model based on the clinical characteristics of 315 patients with NSCLC could efficiently predict the pathological set of SCC and ADC, with an AUC of 0.932 (95% CI: 0.900-0.964) and 0.901 (95% CI: 0.840-0.957) in the training set and testing set, respectively. In our previous study (15), 61 patients with ADC or SCC were divided into a training group and a testing group at the ratio of 7:3, and the features selected from preoperative PET/MRI images were applied to create a prediction model. It was found that the AUC value for classifying ADC and SCC was 0.886 (95% CI: 0.787-0.985) and 0.847 (95% CI: 0.648-1.000) in the training and testing groups, respectively.…”

“…PET/MR scans of the region from the apex to the base of the lung were then performed, and axial, coronal, and sagittal images were obtained using a dedicated MRI coil for the thoracic region to obtain PET, MRI and PET/MR fusion maps of the whole body and region. [Please refer to our previous study ( 15 ) for detailed PET/MRI scanning methods.]. All data were acquired from the same CT and PET/MR instruments.…”

“…All data were acquired from the same CT and PET/MR instruments. Chest CT axial images (1.25 mm), MRI axial T2WI images and PET images were selected as imaging feature extraction sequences in this study ( 15 , 17 ).…”

“…The significant advantages of radiomics include quantifying morphological parameters of tumors, revealing the characteristics of tissue heterogeneity, reflecting the subtle differences between different tissues and linking the image features with tumor characteristics, thus providing objective and valuable information and suggestions for diagnosis, treatment and prognosis (9)(10)(11). In the last few decades, many prediction models based on a single imaging modality (CT, PET and MRI) have been developed for quantifying the tumor microenvironment or predicting tumor pathological type, survival and prognosis (12)(13)(14)(15)(16)(17). However, composite radiomics models based on PET/MRI, CT and clinical metabolic parameters were rarely reported in the literature.…”