Deep learning combined with radiomics may optimize the prediction in differentiating high-grade lung adenocarcinomas in ground glass opacity lesions on CT scans

Wang, Xing; Zhang, Li; Yang, Xin; Tang, Lei; Zhao, Jie; Chen, Gaoxiang; Li, Xiang; Shi, Yan; Li, Shaolei; Yang, Yue; Kang, Yue; Li, Quanzheng; Wu, Nan

doi:10.1016/j.ejrad.2020.109150

Cited by 30 publications

(24 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, a fusion of intranodular (solid and ground-glass) and perinodular radiomic features can be more predictive than the full gross tumor alone [67,68]. In response to the specific types of invasive adenocarcinoma, both handcrafted and deep radiomics have shown efficacy in predicting higher invasive levels of solid/micropapillary adenocarcinoma [69][70][71].…”

Section: Applications Of Structural Radiomic Features In Lung Cancermentioning

confidence: 99%

Structural and functional radiomics for lung cancer

Jochems

Ibrahim

et al. 2021

Eur J Nucl Med Mol Imaging

View full text Add to dashboard Cite

Introduction Lung cancer ranks second in new cancer cases and first in cancer-related deaths worldwide. Precision medicine is working on altering treatment approaches and improving outcomes in this patient population. Radiological images are a powerful non-invasive tool in the screening and diagnosis of early-stage lung cancer, treatment strategy support, prognosis assessment, and follow-up for advanced-stage lung cancer. Recently, radiological features have evolved from solely semantic to include (handcrafted and deep) radiomic features. Radiomics entails the extraction and analysis of quantitative features from medical images using mathematical and machine learning methods to explore possible ties with biology and clinical outcomes. Methods Here, we outline the latest applications of both structural and functional radiomics in detection, diagnosis, and prediction of pathology, gene mutation, treatment strategy, follow-up, treatment response evaluation, and prognosis in the field of lung cancer. Conclusion The major drawbacks of radiomics are the lack of large datasets with high-quality data, standardization of methodology, the black-box nature of deep learning, and reproducibility. The prerequisite for the clinical implementation of radiomics is that these limitations are addressed. Future directions include a safer and more efficient model-training mode, merge multi-modality images, and combined multi-discipline or multi-omics to form “Medomics.”

show abstract

Section: Applications Of Structural Radiomic Features In Lung Cancermentioning

confidence: 99%

Structural and functional radiomics for lung cancer

Jochems

Ibrahim

et al. 2021

Eur J Nucl Med Mol Imaging

View full text Add to dashboard Cite

show abstract

“…Combining radiomic features and deep learning machine learning has also been studied showing an AUC of 0.966 when combining methods while demonstrating an AUC of 0.744 and 0.847 using deep learning and radiomic features respectively (75).…”

Section: Predicting Noninvasive and Invasive Adenocarcinomasmentioning

confidence: 99%

Narrative review of radiomics for classifying pulmonary nodules and potential impact on lung cancer screening

Stephens¹

2021

Curr Chall Thorac Surg

View full text Add to dashboard Cite

Lung cancer screening has proven to be a useful tool for identifying early stage lung cancers, however, the overall accuracy can sometimes lead to false positive and negatives that have potential adverse effects on patient outcomes. Advancement in computational methods have allowed for quantification of pulmonary nodule imaging features, referred to as radiomics, which have the potential to increase lung cancer screening accuracy and improve patient management. The initial part of this review covers common radiomic features and the challenges in deriving them. The second part of this review systematically evaluates literature relating to radiomics and lung cancer finding articles in areas that might have the potential to change management in lung cancer screening. Pertinent literature included initial nodule classification as benign or malignant, classifying subsolid nodules as invasive or noninvasive, and prediction of tumor recurrence after surgical resection. The reviewed articles evaluating use of radiomics are mostly limited due to small sample sizes and lack of a validation cohort. These studies show potential for radiomic features to improve pulmonary nodule classification and change the way patients are managed, however, comparison between studies is limited due to variabilities in the way these features are derived. To make these features useful will require further research and standardization of the workflows that derive these features.

show abstract

“…At present, in medical image processing, deep learning and radiomics are also showing a trend of mutual integration and collaborative development. A common kind of method is the feature-level fusion, which combines the deep features and radiomics features into a new feature vector for the subsequent disease classification and prediction [ 15 , 23 ]. This scheme has been applied to tasks such as detection and classification of lung nodules [ 24 , 25 ], image attribute analysis of tumors [ 26 ], and prediction of cancer survival rates [ 27 ].…”

Section: Introductionmentioning

confidence: 99%

Prenatal prediction and typing of placental invasion using MRI deep and radiomic features

Xuan

Wang

et al. 2021

BioMed Eng OnLine

View full text Add to dashboard Cite

Background To predict placental invasion (PI) and determine the subtype according to the degree of implantation, and to help physicians develop appropriate therapeutic measures, a prenatal prediction and typing of placental invasion method using MRI deep and radiomic features were proposed. Methods The placental tissue of abdominal magnetic resonance (MR) image was segmented to form the regions of interest (ROI) using U-net. The radiomic features were subsequently extracted from ROI. Simultaneously, a deep dynamic convolution neural network (DDCNN) with codec structure was established, which was trained by an autoencoder model to extract the deep features from ROI. Finally, combining the radiomic features and deep features, a classifier based on the multi-layer perceptron model was designed. The classifier was trained to predict prenatal placental invasion as well as determine the invasion subtype. Results The experimental results show that the average accuracy, sensitivity, and specificity of the proposed method are 0.877, 0.857, and 0.954 respectively, and the area under the ROC curve (AUC) is 0.904, which outperforms the traditional radiomic based auxiliary diagnostic methods. Conclusions This work not only labeled the placental tissue of MR image in pregnant women automatically but also realized the objective evaluation of placental invasion, thus providing a new approach for the prenatal diagnosis of placental invasion.

show abstract

Deep learning combined with radiomics may optimize the prediction in differentiating high-grade lung adenocarcinomas in ground glass opacity lesions on CT scans

Cited by 30 publications

References 34 publications

Structural and functional radiomics for lung cancer

Structural and functional radiomics for lung cancer

Narrative review of radiomics for classifying pulmonary nodules and potential impact on lung cancer screening

Prenatal prediction and typing of placental invasion using MRI deep and radiomic features

Contact Info

Product

Resources

About