CT-based radiomics for prediction of histologic subtype and metastatic disease in primary malignant lung neoplasms

Ferreira, José Raniery; Koenigkam-Santos, Marcel; Tenório, Ariane Priscilla Magalhães; Faleiros, Matheus Calil; Cipriano, Federico Enrique Garcia; Fabro, Alexandre Todorovic; Näppi, Janne J.; Yoshida, Hiroyuki; Azevedo-Marques, Paulo Mazzoncini de

doi:10.1007/s11548-019-02093-y

Cited by 50 publications

(41 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Busyness is a parameter of NGLDM, which measures the spatial frequency of changes in intensity between nearby voxels of different grey-levels. The role of busyness in TETs has not been reported before, but it has been used to assess the tissue heterogeneity in glioma and lung cancer [31][32][33]. Heterogeneity is a recognized feature of tumors that considered to be positively correlated with the tumor malignancy, which is of great clinical significance for effective personalized therapies [34; 35].…”

Section: Discussionmentioning

confidence: 99%

Development and validation of a CT-TA nomogram for preoperatively differentiating thymic epithelial tumor histologic subtypes

Ren

Zhang

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: Thymic epithelial tumors (TETs) are the most common primary tumors in the anterior mediastinum with considerable histologic heterogeneity. This study aimed to develop and validate a nomogram based on computed tomography (CT) and texture analysis (TA) for preoperatively predicting pathological classification for TET patients.Methods: Totally 172 patients with pathologically confirmed TET after surgery were retrospectively analyzed and randomly divided into a training cohort (n=120) and a validation cohort (n=52). Preoperative clinical demographic, CT signs and texture features of each patient were analyzed and prediction models were developed using the least absolute shrinkage and selection operator (LASSO) regression. The performance of models was evaluated and compared by the area under receiver operating characteristic (ROC) curve (AUC) and DeLong test. The clinical application value of models was determined through the decision curve analysis (DCA). Then a nomogram was developed based on the model with the best predictive accuracy and clinical utility and validated using the calibration plots.Results: Totally 87 patients with low-risk TET (LTET) (types A, AB, B1) and 85 patients with high-risk TET (HTET) (types B2, B3, C) were enrolled in this study. We separately constructed 4 prediction models for differentiating LTET from HTET using clinical, CT, texture features and their combination. These 4 prediction models achieved AUC values of 0.66, 0.79, 0.82, 0.88 in the training cohort and 0.64, 0.82, 0.86, 0.94 in the validation cohort, respectively. DeLong test and DCA analysis showed that the Combined model consisting of 2 CT signs and 2 texture parameters held the highest predictive efficiency and clinical utility ( p <0.05). A prediction nomogram was subsequently developed using the combined model’s 4 independently risk factors. The calibration curves indicated a good consistency between the actual observation and nomogram prediction for differentiating TET classifications.Conclusion: A prediction nomogram incorporating both the CT and texture parameters was constructed and validated in our study, which was conveniently used to facilitate the preoperative individualized prediction of simplified histologic subtypes in TET patients, assisting in clinical treatment decision making and achieving precision treatment.

show abstract

Section: Discussionmentioning

confidence: 99%

Development and validation of a CT-TA nomogram for preoperatively differentiating thymic epithelial tumor histologic subtypes

Ren

Zhang

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…To improve modeling sensitivity, we employed a strategy to increase the weight of the abnormal class for the training [Ferreira Junior et al 2020b]. For this purpose, normal images weighted 0.25, and augmented images from patients with cardiomegaly weighted 0.75 (Table 1).…”

Section: Convolutional Neural Network Trainingmentioning

confidence: 99%

“…A better approach is to automate this measurement to improve medical care. Computer-based tools can analyze large volumes of medical images and accurately detect disease patterns automatically [Ferreira Junior et al 2020b]. Recently, the use of deep learning has been gaining considerable importance in medical image analysis as it has the potential to improve the efficiency of specialists [LeCun et al 2015].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multicenter Validation of Convolutional Neural Networks for Automated Detection of Cardiomegaly on Chest Radiographs

Cárdenas¹,

Ferreira²,

Moreno³

et al. 2020

Anais Principais Do Simpósio Brasileiro De Computação Aplicada À Saúde (SBCAS 2020)

View full text Add to dashboard Cite

This work focused on validating five convolutional neural network models to detect automatically cardiomegaly, a health complication that causes heart enlargement, which may lead to cardiac arrest. To do that, we trained the models with a customized multilayer perceptron. Radiographs from two public datasets were used in experiments, one of them only for external validation. Images were pre-processed to contain just the chest cavity. The EfficientNet model yielded the highest area under the curve (AUC) of 0.91 on the test set. However, the Inception-based model obtained the best generalization performance with AUC of 0.88 on the independent multicentric dataset. Therefore, this work accurately validated radiographic models to identify patients with cardiomegaly.

show abstract

“…In this era of personalized medicine, radiomics has markedly grown as a precision imaging tool for the improvement of diagnosis, assessment of prognosis, and prediction of therapy response [Gillies et al 2016]. Radiomics currently consists of modeling quantitative features from medical images into a machine-learning method to improve clinical decision support systems [Ferreira Junior et al 2020]. Once the model has been designed, it can be used as a radiomics-aided support system for precision medicine and SpA evaluation.…”

Section: Introductionmentioning

confidence: 99%

Radiomics Assessment of SPAIR and STIR MRI Sequences to Predict Axial and Peripheral Spondyloarthritis

Tenório¹,

Ferreira²,

Dalto³

et al. 2020

Anais Principais Do Simpósio Brasileiro De Computação Aplicada À Saúde (SBCAS 2020)

View full text Add to dashboard Cite

In an attempt to aid the subtyping of spondyloarthritis (SpA), this work assessed neural nets and magnetic resonance imaging (MRI) features to predict SpA. Patients underwent SPAIR and STIR MRI sequences. Radiologists manually segmented sacroiliac joints images for extracting MRI features. A neural net used these features to predict SpA. The STIR-based model yielded higher performance than SPAIR to diagnose SpA, although no statistical difference was found between them. The SPAIR model yielded an area under the curve of 0.83 to differentiate axial and peripheral subtypes, while STIR yielded 0.57 (p < 0.05 on curves difference). Therefore, neural nets modeled with SPAIR-extracted features distinguished SpA using a single MRI exam of the sacroiliac joints.

show abstract

CT-based radiomics for prediction of histologic subtype and metastatic disease in primary malignant lung neoplasms

Cited by 50 publications

References 41 publications

Development and validation of a CT-TA nomogram for preoperatively differentiating thymic epithelial tumor histologic subtypes

Development and validation of a CT-TA nomogram for preoperatively differentiating thymic epithelial tumor histologic subtypes

Multicenter Validation of Convolutional Neural Networks for Automated Detection of Cardiomegaly on Chest Radiographs

Radiomics Assessment of SPAIR and STIR MRI Sequences to Predict Axial and Peripheral Spondyloarthritis

Contact Info

Product

Resources

About