Development and validation of a deep learning signature for predicting lymph node metastasis in lung adenocarcinoma: comparison with radiomics signature and clinical-semantic model

Multimodal Deep Learning Fusing Clinical and Radiomics Scores for Prediction of Early-Stage Lung Adenocarcinoma Lymph Node Metastasis

RATIONALE AND OBJECTIVES

To develop and validate a multimodal deep learning (DL) model based on computed tomography (CT) images and clinical knowledge to predict lymph node metastasis (LNM) in early lung adenocarcinoma.

MATERIALS AND METHODS

A total of 724 pathologically confirmed early invasive lung adenocarcinoma patients were retrospectively included from two centers. Clinical and CT semantic features of the patients were collected, and 3D radiomics features were extracted from nonenhanced CT images. We proposed a multimodal feature fusion DL network based on the InceptionResNetV2 architecture, which can effectively extract and integrate image and clinical knowledge to predict LNM.

RESULTS

A total of 524 lung adenocarcinoma patients from Center 1 were randomly divided into training (n=418) and internal validation (n=106) sets in a 4:1 ratio, while 200 lung adenocarcinoma patients from Center 2 served as the independent test set. Among the 16 collected clinical and imaging features, 8 were selected: gender, serum carcinoembryonic antigen, cytokeratin 19 fragment antigen 21-1, neuron-specific enolase, tumor size, location, density, and centrality. From the 1595 extracted radiomics features, six key features were identified. The CS-RS-DL fusion model achieved the highest area under the receiver operating characteristic curve in both the internal validation set (0.877) and the independent test set (0.906) compared to other models. The Delong test results for the independent test set indicated that the CS-RS-DL model significantly outperformed the clinical model (0.844), radiomics model (0.850), CS-RS model (0.872), single DL model (0.848), and the CS-DL model (0.875) (all P<0.05). Additionally, the CS-RS-DL model exhibited the highest sensitivity (0.941) and average precision (0.642).

CONCLUSION

The knowledge derived from clinical, radiomics, and DL is complementary in predicting LNM in lung adenocarcinoma. The integration of clinical and radiomics scores through DL can significantly improve the accuracy of lymph node status assessment.

Diagnostic accuracy of CT and PET/CT radiomics in predicting lymph node metastasis in non-small cell lung cancer

OBJECTIVES

This study evaluates the accuracy of radiomics in predicting lymph node metastasis in non-small cell lung cancer, which is crucial for patient management and prognosis.

METHODS

Adhering to PRISMA and AMSTAR guidelines, we systematically reviewed literature from March 2012 to December 2023 using databases including PubMed, Web of Science, and Embase. Radiomics studies utilizing computed tomography (CT) and positron emission tomography (PET)/CT imaging were included. The quality of studies was appraised with QUADAS-2 and RQS tools, and the TRIPOD checklist assessed model transparency. Sensitivity, specificity, and AUC values were synthesized to determine diagnostic performance, with subgroup and sensitivity analyses probing heterogeneity and a Fagan plot evaluating clinical applicability.

RESULTS

Our analysis incorporated 42 cohorts from 22 studies. CT-based radiomics demonstrated a sensitivity of 0.84 (95% CI: 0.79-0.88, p < 0.01) and specificity of 0.82 (95% CI: 0.75-0.87, p < 0.01), with an AUC of 0.90 (95% CI: 0.87-0.92), indicating no publication bias (p-value = 0.54 > 0.05). PET/CT radiomics showed a sensitivity of 0.82 (95% CI: 0.76-0.86, p < 0.01) and specificity of 0.86 (95% CI: 0.81-0.90, p < 0.01), with an AUC of 0.90 (95% CI: 0.87-0.93), with a slight publication bias (p-value = 0.03 < 0.05). Despite high clinical utility, subgroup analysis did not clarify heterogeneity sources, suggesting influences from possible factors like lymph node location and small subgroup sizes.

CONCLUSIONS

Radiomics models show accuracy in predicting lung cancer lymph node metastasis, yet further validation with larger, multi-center studies is necessary.

CLINICAL RELEVANCE STATEMENT

Radiomics models using CT and PET/CT imaging may improve the prediction of lung cancer lymph node metastasis, aiding personalized treatment strategies.

RESEARCH REGISTRATION UNIQUE IDENTIFYING NUMBER (UIN)

International Prospective Register of Systematic Reviews (PROSPERO), CRD42023494701. This study has been registered on the PROSPERO platform with a registration date of 18 December 2023. https://www.crd.york.ac.uk/prospero/ KEY POINTS: The study explores radiomics for lung cancer lymph node metastasis detection, impacting surgery and prognosis. Radiomics improves the accuracy of lymph node metastasis prediction in lung cancer. Radiomics can aid in the prediction of lymph node metastasis in lung cancer and personalized treatment.

Development and validation of MRI-derived deep learning score for non-invasive prediction of PD-L1 expression and prognostic stratification in head and neck squamous cell carcinoma

BACKGROUND

Immunotherapy has revolutionized the treatment landscape for head and neck squamous cell carcinoma (HNSCC) and PD-L1 combined positivity score (CPS) scoring is recommended as a biomarker for immunotherapy. Therefore, this study aimed to develop an MRI-based deep learning score (DLS) to non-invasively assess PD-L1 expression status in HNSCC patients and evaluate its potential effeciency in predicting prognostic stratification following treatment with immune checkpoint inhibitors (ICI).

METHODS

In this study, we collected data from four patient cohorts comprising a total of 610 HNSCC patients from two separate institutions. We developed deep learning models based on the ResNet-101 convolutional neural network to analyze three MRI sequences (T1WI, T2WI, and contrast-enhanced T1WI). Tumor regions were manually segmented, and features extracted from different MRI sequences were fused using a transformer-based model incorporating attention mechanisms. The model's performance in predicting PD-L1 expression was evaluated using the area under the curve (AUC), sensitivity, specificity, and calibration metrics. Survival analyses were conducted using Kaplan-Meier survival curves and log-rank tests to evaluate the prognostic significance of the DLS.

RESULTS

The DLS demonstrated high predictive accuracy for PD-L1 expression, achieving an AUC of 0.981, 0.860 and 0.803 in the training, internal and external validation cohort. Patients with higher DLS scores demonstrated significantly improved progression-free survival (PFS) in both the internal validation cohort (hazard ratio: 0.491; 95% CI, 0.270-0.892; P = 0.005) and the external validation cohort (hazard ratio: 0.617; 95% CI, 0.391-0.973; P = 0.040). In the ICI-treated cohort, the DLS achieved an AUC of 0.739 for predicting durable clinical benefit (DCB).

CONCLUSIONS

The proposed DLS offered a non-invasive and accurate approach for assessing PD-L1 expression in patients with HNSCC and effectively stratified HNSCC patients to benefit from immunotherapy based on PFS.

Deep learning radiomics nomogram for preoperatively identifying moderate-to-severe chronic cholangitis in children with pancreaticobiliary maljunction: a multicenter study

BACKGROUND

Long-term severe cholangitis can lead to dense adhesions and increased fragility of the bile duct, complicating surgical procedures and elevating operative risk in children with pancreaticobiliary maljunction (PBM). Consequently, preoperative diagnosis of moderate-to-severe chronic cholangitis is essential for guiding treatment strategies and surgical planning. This study aimed to develop and validate a deep learning radiomics nomogram (DLRN) based on contrast-enhanced CT images and clinical characteristics to preoperatively identify moderate-to-severe chronic cholangitis in children with PBM.

METHODS

A total of 323 pediatric patients with PBM who underwent surgery were retrospectively enrolled from three centers, and divided into a training cohort (n = 153), an internal validation cohort (IVC, n = 67), and two external test cohorts (ETC1, n = 58; ETC2, n = 45). Chronic cholangitis severity was determined by postoperative pathology. Handcrafted radiomics features and deep learning (DL) radiomics features, extracted using transfer learning with the ResNet50 architecture, were obtained from portal venous-phase CT images. Multivariable logistic regression was used to establish the DLRN, integrating significant clinical factors with handcrafted and DL radiomics signatures. The diagnostic performances were evaluated in terms of discrimination, calibration, and clinical usefulness.

RESULTS

Biliary stones and peribiliary fluid collection were selected as important clinical factors. 5 handcrafted and 5 DL features were retained to build the two radiomics signatures, respectively. The integrated DLRN achieved satisfactory performance, achieving area under the curve (AUC) values of 0.913 (95% CI, 0.834-0.993), 0.916 (95% CI, 0.845-0.987), and 0.895 (95% CI, 0.801-0.989) in the IVC, and two ETCs, respectively. In comparison, the clinical model, handcrafted signature, and DL signature had AUC ranges of 0.654-0.705, 0.823-0.857, and 0.840-0.872 across the same cohorts. The DLRN outperformed single-modality clinical, handcrafted radiomics, and DL radiomics models, with all integrated discrimination improvement values > 0 and P < 0.05. The Hosmer-Lemeshow test and calibration curves showed good consistency of the DLRN (P > 0.05), and the decision curve analysis and clinical impact curve further confirmed its clinical utility.

CONCLUSIONS

The integrated DLRN can be a useful and non-invasive tool for preoperatively identifying moderate-to-severe chronic cholangitis in children with PBM, potentially enhancing clinical decision-making and personalized management strategies.

A CT-based deep learning model for preoperative prediction of spread through air spaces in clinical stage I lung adenocarcinoma

Objective

To develop and validate a deep learning signature for noninvasive prediction of spread through air spaces (STAS) in clinical stage I lung adenocarcinoma and compare its predictive performance with conventional clinical-semantic model.

Methods

A total of 513 patients with pathologically-confirmed stage I lung adenocarcinoma were retrospectively enrolled and were divided into training cohort (n = 386) and independent validation cohort (n = 127) according to different center. Clinicopathological data were collected and CT semantic features were evaluated. Multivariate logistic regression analyses were conducted to construct a clinical-semantic model predictive of STAS. The Swin Transformer architecture was adopted to develop a deep learning signature predictive of STAS. Model performance was assessed using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive and negative predictive value, and calibration curve. AUC comparisons were performed by the DeLong test.

Results

The proposed deep learning signature achieved an AUC of 0.869 (95% CI: 0.831, 0.901) in training cohort and 0.837 (95% CI: 0.831, 0.901) in validation cohort, surpassing clinical-semantic model both in training and validation cohort (all P<0.01). Calibration curves demonstrated good agreement between STAS predicted probabilities using deep learning signature and actual observed probabilities in both cohorts. The inclusion of all clinical-semantic risk predictors failed to show an incremental value with respect to deep learning signature.

Conclusions

The proposed deep learning signature based on Swin Transformer achieved a promising performance in predicting STAS in clinical stage I lung adenocarcinoma, thereby offering information in directing surgical strategy and facilitating adjuvant therapeutic scheduling.

Predicting lymph node metastasis of clinical T1 non-small cell lung cancer: a brief review of possible methodologies and controversies

Non-small cell lung cancer (NSCLC) is a major subtype of lung cancer and poses a serious threat to human health. Due to the advances in lung cancer screening, more and more clinical T1 NSCLC defined as a tumor with a maximum diameter of 3cm surrounded by lung tissue or visceral pleura have been detected and have achieved favorable treatment outcomes, greatly improving the prognosis of NSCLC patients. However, the preoperative lymph node staging and intraoperative lymph node dissection patterns of operable clinical T1 NSCLC are still subject to much disagreement, as well as the heterogeneity between primary tumors and metastatic lymph nodes poses a challenge in designing effective treatment strategies. This article comprehensively describes the clinical risk factors of clinical T1 NSCLC lymph node metastasis, and its invasive and non-invasive prediction, focusing on the genetic heterogeneity between the primary tumor and the metastatic lymph nodes, which is significant for a thoroughly understanding of the biological behavior of early-stage NSCLC.

Can computed tomography findings and radiomics analysis of mediastinal lymph nodes differentiate between sarcoidosis and lymphoma?

AIMS

To assess the ability of computed tomography (CT) findings and radiomics analysis to differentiate mediastinal lymphadenopathies as sarcoidosis versus lymphoma.

MATERIALS AND METHODS

94 patients with lymphoma and 97 patients with sarcoidosis, who had > 1cm mediastinal lymph node were included. Size, location of lymph nodes, and distribution of the largest lymph nodes in two groups were compared. A total of 636 lymphadenopathies in four different regions were segmented for radiomics. Lesion segmentation was semiautomatically performed with a dedicated commercial software package on chest CT images. 149 patients were grouped as a training cohort, while 42 patients who underwent CT in the oncology hospital were used for external validation. The least absolute shrinkage and selection operator (LASSO) analysis was used to perform feature selection. Using selected features, the classification performance of various data mining methods in separating groups of sarcoidosis and lymphoma was investigated.

RESULTS

Distribution and size of lymphadenopathies were significantly different in sarcoidosis and lymphoma groups (<0.05). Radiomics and data mining methods showed excellent performance in differentiating lymph nodes of sarcoidosis and lymphoma according to both the largest lymphadenopathy and lymphadenopathies in four different mediastinal regions (AUC >0,95).

CONCLUSIONS

Distribution and size of lymphadenopathies can help differential diagnosis in patients with sarcoidosis and lymphoma. CT radiomics analysis can discriminate the lymph nodes of sarcoidosis and lymphoma with great performance regardless of lymph node size and location and it can be used safely in the diagnosis of these diseases, which can sometimes be challenging to distinguish from each other.

A machine learning based radiomics approach for predicting No. 14v station lymph node metastasis in gastric cancer

Purpose

To evaluate the potential of radiomics approach for predicting No. 14v station lymph node metastasis (14vM) in gastric cancer (GC).

Methods

The contrast enhanced CT (CECT) images with corresponding clinical information of 288 GC patients were retrospectively collected. Patients were separated into training set (n = 202) and testing set (n = 86). A total of 1,316 radiomics feature were extracted from portal venous phase images of CECT. Seven machine learning (ML) algorithms including naïve Bayes (NB), k-nearest neighbor (KNN), decision tree (DT), logistic regression (LR), random forest (RF), eXtreme gradient boosting (XGBoost) and support vector machine (SVM) were trained for development of optimal radiomics signature. A combined model was established by combining radiomics with important clinicopathological factors. The diagnostic ability of the signature and model were evaluated.

Results

LR algorithm was chosen for signature construction. The radiomics signature exhibited good discrimination accuracy of 14vM with AUCs of 0.83 in the training and 0.77 in the testing set. The risk of 14vM showed significant association with higher radiomics score. A combined model exhibited increased predictive ability and good agreement in the training (AUC = 0.87) and testing (AUC = 0.85) sets.

Conclusion

The ML-based radiomics model provided a promising image biomarker for preoperative detection of 14vM and may help the surgeon to decide whether to add 14v dissection to lymphadenectomy.

Development of a multi-modal learning-based lymph node metastasis prediction model for lung cancer

PURPOSE

This study proposed a three-dimensional (3D) multi-modal learning-based model for the automated prediction and classification of lymph node metastasis in patients with non-small cell lung cancer (NSCLC) using computed tomography (CT) images and clinical information.

METHODS

We utilized clinical information and CT image data from 4239 patients with NSCLC across multiple institutions. Four deep learning algorithm-based multi-modal models were constructed and evaluated for lymph node classification. To further enhance classification performance, a soft-voting ensemble technique was applied to integrate the outcomes of multiple multi-modal models.

RESULTS

A comparison of the classification performance revealed that the multi-modal model, which integrated CT images and clinical information, outperformed the single-modal models. Among the four multi-modal models, the Xception model demonstrated the highest classification performance, with an area under the curve (AUC) of 0.756 for the internal test dataset and 0.736 for the external validation dataset. The ensemble model (SEResNet50_DenseNet121_Xception) exhibited even better performance, with an AUC of 0.762 for the internal test dataset and 0.751 for the external validation dataset, surpassing the multi-modal model's performance.

CONCLUSIONS

Integrating CT images and clinical information improved the performance of the lymph node metastasis prediction models in patients with NSCLC. The proposed 3D multi-modal lymph node prediction model can serve as an auxiliary tool for evaluating lymph node metastasis in patients with non-pretreated NSCLC, aiding in patient screening and treatment planning.

A transformer-based deep learning model for early prediction of lymph node metastasis in locally advanced gastric cancer after neoadjuvant chemotherapy using pretreatment CT images

Background

Early prediction of lymph node status after neoadjuvant chemotherapy (NAC) facilitates promptly optimization of treatment strategies. This study aimed to develop and validate a deep learning network (DLN) using baseline computed tomography images to predict lymph node metastasis (LNM) after NAC in patients with locally advanced gastric cancer (LAGC).

Methods

A total of 1205 LAGC patients were retrospectively recruited from three hospitals between January 2013 and March 2023, constituting a training cohort, an internal validation cohort, and two external validation cohorts. A transformer-based DLN was developed using 3D tumor images to predict LNM after NAC. A clinical model was constructed through multivariate logistic regression analysis as a baseline for subsequent comparisons. The performance of the models was evaluated through discrimination, calibration, and clinical applicability. Furthermore, Kaplan-Meier survival analysis was conducted to assess overall survival (OS) of LAGC patients at two follow-up centers.

Findings

The DLN outperformed the clinical model and demonstrated a robust performance for predicting LNM in the training and validation cohorts, with areas under the curve (AUCs) of 0.804 (95% confidence interval [CI], 0.752-0.849), 0.748 (95% CI, 0.660-0.830), 0.788 (95% CI, 0.735-0.835), and 0.766 (95% CI, 0.717-0.814), respectively. Decision curve analysis exhibited a high net clinical benefit of the DLN. Moreover, the DLN was significantly associated with the OS of LAGC patients [Center 1: hazard ratio (HR), 1.789, P < 0.001; Center 2:HR, 1.776, P = 0.013].

Interpretation

The transformer-based DLN provides early and effective prediction of LNM and survival outcomes in LAGC patients receiving NAC, with promise to guide individualized therapy. Future prospective multicenter studies are warranted to further validate our model.

Funding

National Natural Science Foundation of China (NO. 82373432, 82171923, 82202142), Project Funded by China Postdoctoral Science Foundation (NO. 2022M720857), Regional Innovation and Development Joint Fund of National Natural Science Foundation of China (NO. U22A20345), National Science Fund for Distinguished Young Scholars of China (NO. 81925023), Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application (NO. 2022B1212010011), High-level Hospital Construction Project (NO. DFJHBF202105), Natural Science Foundation of Guangdong Province for Distinguished Young Scholars (NO. 2024B1515020091).

Lymph node metastasis in early invasive lung adenocarcinoma: Prediction model establishment and validation based on genomic profiling and clinicopathologic characteristics

BACKGROUND

The presence of lymph node (LN) metastasis directly affects the treatment strategy for lung adenocarcinoma (LUAD). Next-generation sequencing (NGS) has been widely used in patients with advanced LUAD to identify targeted genes, while early detection of pathologic LN metastasis using NGS has not been assessed.

METHODS

Clinicopathologic features and molecular characteristics of 224 patients from Ruijin Hospital were analyzed to detect factors associated with LN metastases. Another 140 patients from Huashan Hospital were set as a test cohort.

RESULTS

Twenty-four out of 224 patients were found to have lymph node metastases (10.7%). Pathologic LN-positive tumors showed higher mutant allele tumor heterogeneity (p < 0.05), higher tumor mutation burden (p < 0.001), as well as more frequent KEAP1 (p = 0.001), STK11 (p = 0.004), KRAS (p = 0.007), CTNNB1 (p = 0.017), TP53, and ARID2 mutations (both p = 0.02); whereas low frequency of EGFR mutation (p = 0.005). A predictive nomogram involving male sex, solid tumor morphology, higher T stage, EGFR wild-type, and TP53, STK11, CDKN2A, KEAP1, ARID2, KRAS, SDHA, SPEN, CTNNB1, DICER1 mutations showed outstanding efficiency in both the training cohort (AUC = 0.819) and the test cohort (AUC = 0.780).

CONCLUSION

This study suggests that the integration of genomic profiling and clinical features identifies early-invasive LUAD patients at higher risk of LN metastasis. Improved identification of LN metastasis is beneficial for the optimization of the patient's therapy decisions.

Automatic image segmentation and online survival prediction model of medulloblastoma based on machine learning

OBJECTIVES

To develop a dynamic nomogram containing radiomics signature and clinical features for estimating the overall survival (OS) of patients with medulloblastoma (MB) and design an automatic image segmentation model to reduce labor and time costs.

METHODS

Data from 217 medulloblastoma (MB) patients over the past 4 years were collected and separated into a training set and a test set. Intraclass correlation coefficient (ICC), random survival forest (RSF), and least absolute shrinkage and selection operator (LASSO) regression methods were employed to select variables in the training set. Univariate and multivariate Cox proportional hazard models, as well as Kaplan-Meier analysis, were utilized to determine the relationship among the radiomics signature, clinical features, and overall survival. A dynamic nomogram was developed. Additionally, a 3D-Unet deep learning model was used to train the automatic tumor delineation model.

RESULTS

Higher Rad-scores were significantly associated with worse OS in both the training and validation sets (p < 0.001 and p = 0.047, respectively). The Cox model combined clinical and radiomics signatures ([IBS = 0.079], [C-index = 0.747, SE = 0.045]) outperformed either radiomics signatures alone ([IBS = 0.081], [C-index = 0.738, SE = 0.041]) or clinical features alone ([IBS = 0.085], [C-index = 0.565, SE = 0.041]). The segmentation model had mean Dice coefficients of 0.80, 0.82, and 0.78 in the training, validation, and test sets respectively. A deep learning-based tumor segmentation model was built with Dice coefficients of 0.8372, 0.8017, and 0.7673 on the training set, validation set, and test set, respectively.

CONCLUSIONS

A combination of radiomics features and clinical characteristics enhances the accuracy of OS prediction in medulloblastoma patients. Additionally, building an MRI image automatic segmentation model reduces labor and time costs.

CLINICAL RELEVANCE STATEMENT

A survival prognosis model based on radiomics and clinical characteristics could improve the accuracy of prognosis estimation for medulloblastoma patients, and an MRI-based automatic tumor segmentation model could reduce the cost of time.

KEY POINTS

• A model that combines radiomics and clinical features can predict the survival prognosis of patients with medulloblastoma. • Online nomogram and image automatic segmentation model can help doctors better judge the prognosis of medulloblastoma and save working time. • The developed AI system can help doctors judge the prognosis of diseases and promote the development of precision medicine.

Predictive value of radiomic features extracted from primary lung adenocarcinoma in forecasting thoracic lymph node metastasis: a systematic review and meta-analysis

BACKGROUND

The application of radiomics in thoracic lymph node metastasis (LNM) of lung adenocarcinoma is increasing, but diagnostic performance of radiomics from primary tumor to predict LNM has not been systematically reviewed. Therefore, this study sought to provide a general overview regarding the methodological quality and diagnostic performance of using radiomic approaches to predict the likelihood of LNM in lung adenocarcinoma.

METHODS

Studies were gathered from literature databases such as PubMed, Embase, the Web of Science Core Collection, and the Cochrane library. The Radiomic Quality Score (RQS) and the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) were both used to assess the quality of each study. The pooled sensitivity, specificity, and area under the curve (AUC) of the best radiomics models in the training and validation cohorts were calculated. Subgroup and meta-regression analyses were also conducted.

RESULTS

Seventeen studies with 159 to 1202 patients each were enrolled between the years of 2018 to 2022, of which ten studies had sufficient data for the quantitative evaluation. The percentage of RQS was between 11.1% and 44.4% and most of the studies were considered to have a low risk of bias and few applicability concerns in QUADAS-2. Pyradiomics and logistic regression analysis were the most commonly used software and methods for radiomics feature extraction and selection, respectively. In addition, the best prediction models in seventeen studies were mainly based on radiomics features combined with non-radiomics features (semantic features and/or clinical features). The pooled sensitivity, specificity, and AUC of the training cohorts were 0.84 (95% confidence interval (CI) [0.73-0.91]), 0.88 (95% CI [0.81-0.93]), and 0.93(95% CI [0.90-0.95]), respectively. For the validation cohorts, the pooled sensitivity, specificity, and AUC were 0.89 (95% CI [0.82-0.94]), 0.86 (95% CI [0.74-0.93]) and 0.94 (95% CI [0.91-0.96]), respectively.

CONCLUSIONS

Radiomic features based on the primary tumor have the potential to predict preoperative LNM of lung adenocarcinoma. However, radiomics workflow needs to be standardized to better promote the applicability of radiomics.

TRIAL REGISTRATION

CRD42022375712.

Reproducibility of radiomics quality score: an intra- and inter-rater reliability study

OBJECTIVES

To investigate the intra- and inter-rater reliability of the total radiomics quality score (RQS) and the reproducibility of individual RQS items' score in a large multireader study.

METHODS

Nine raters with different backgrounds were randomly assigned to three groups based on their proficiency with RQS utilization: Groups 1 and 2 represented the inter-rater reliability groups with or without prior training in RQS, respectively; group 3 represented the intra-rater reliability group. Thirty-three original research papers on radiomics were evaluated by raters of groups 1 and 2. Of the 33 papers, 17 were evaluated twice with an interval of 1 month by raters of group 3. Intraclass coefficient (ICC) for continuous variables, and Fleiss' and Cohen's kappa (k) statistics for categorical variables were used.

RESULTS

The inter-rater reliability was poor to moderate for total RQS (ICC 0.30-055, p < 0.001) and very low to good for item's reproducibility (k - 0.12 to 0.75) within groups 1 and 2 for both inexperienced and experienced raters. The intra-rater reliability for total RQS was moderate for the less experienced rater (ICC 0.522, p = 0.009), whereas experienced raters showed excellent intra-rater reliability (ICC 0.91-0.99, p < 0.001) between the first and second read. Intra-rater reliability on RQS items' score reproducibility was higher and most of the items had moderate to good intra-rater reliability (k - 0.40 to 1).

CONCLUSIONS

Reproducibility of the total RQS and the score of individual RQS items is low. There is a need for a robust and reproducible assessment method to assess the quality of radiomics research.

CLINICAL RELEVANCE STATEMENT

There is a need for reproducible scoring systems to improve quality of radiomics research and consecutively close the translational gap between research and clinical implementation.

KEY POINTS

• Radiomics quality score has been widely used for the evaluation of radiomics studies. • Although the intra-rater reliability was moderate to excellent, intra- and inter-rater reliability of total score and point-by-point scores were low with radiomics quality score. • A robust, easy-to-use scoring system is needed for the evaluation of radiomics research.

A Transformer-Based microvascular invasion classifier enhances prognostic stratification in HCC following radiofrequency ablation

BACKGROUND & AIMS

We aimed to develop a Transformer-based deep learning (DL) network for prognostic stratification in hepatocellular carcinoma (HCC) patients undergoing RFA.

METHODS

A Swin Transformer DL network was trained to establish associations between magnetic resonance imaging (MRI) datasets and the ground truth of microvascular invasion (MVI) based on 696 surgical resection (SR) patients with solitary HCC ≤3 cm, and was validated in an external cohort (n = 180). The multiphase MRI-based DL risk outputs using an optimal threshold of .5 was employed as a MVI classifier for prognosis stratification in the RFA cohort (n = 180).

RESULTS

Over 90% of all enrolled patients exhibited hepatitis B virus infection. Liver cirrhosis was significantly more prevalent in the RFA cohort compared to the SR cohort (72.2% vs. 44.1%, p < .001). The MVI risk outputs exhibited good performance (area under the curve values = .938 and .883) for predicting MVI in the training and validation cohort, respectively. The RFA patients at high risk of MVI classified by the MVI classifier demonstrated significantly lower recurrence-free survival (RFS) and overall survival rates at 1, 3 and 5 years compared to those classified as low risk (p < .001). Multivariate cox regression modelling of a-fetoprotein > 20 ng/mL [hazard ratio (HR) = 1.53; 95% confidence interval (95% CI): 1.02-2.33, p = .047], high risk of MVI (HR = 3.76; 95% CI: 2.40-5.88, p < .001) and unfavourable tumour location (HR = 2.15; 95% CI: 1.40-3.29, p = .001) yielded a c-index of .731 (bootstrapped 95% CI: .667-.778) for evaluating RFS after RFA. Among the three risk factors, MVI was the most powerful predictor for intrahepatic distance recurrence.

CONCLUSIONS

The proposed MVI classifier can serve as a valuable imaging biomarker for prognostic stratification in early-stage HCC patients undergoing RFA.

Prediction of lymph node metastasis in operable cervical cancer using clinical parameters and deep learning with MRI data: a multicentre study

OBJECTIVES

To develop and validate a magnetic resonance imaging-based (MRI) deep multiple instance learning (D-MIL) model and combine it with clinical parameters for preoperative prediction of lymph node metastasis (LNM) in operable cervical cancer.

METHODS

A total of 392 patients with cervical cancer were retrospectively enrolled. Clinical parameters were analysed by logistical regression to construct a clinical model (M1). A ResNet50 structure is applied to extract features at the instance level without using manual annotations about the tumour region and then construct a D-MIL model (M2). A hybrid model (M3) was constructed by M1 and M2 scores. The diagnostic performance of each model was evaluated by the area under the receiver operating characteristic curve (AUC) and compared using the Delong method. Disease-free survival (DFS) was evaluated by the Kaplan‒Meier method.

RESULTS

SCC-Ag, maximum lymph node short diameter (LNmax), and tumour volume were found to be independent predictors of M1 model. For the diagnosis of LNM, the AUC of the training/internal/external cohort of M1 was 0.736/0.690/0.732, the AUC of the training/internal/external cohort of M2 was 0.757/0.714/0.765, and the AUC of the training/internal/external cohort of M3 was 0.838/0.764/0.835. M3 showed better performance than M1 and M2. Through the survival analysis, patients with higher hybrid model scores had a shorter time to reach DFS.

CONCLUSION

The proposed hybrid model could be used as a personalised non-invasive tool, which is helpful for predicting LNM in operable cervical cancer. The score of the hybrid model could also reflect the DFS of operable cervical cancer.

CRITICAL RELEVANCE STATEMENT

Lymph node metastasis is an important factor affecting the prognosis of cervical cancer. Preoperative prediction of lymph node status is helpful to make treatment decisions, improve prognosis, and prolong survival time.

KEY POINTS

• The MRI-based deep-learning model can predict the LNM in operable cervical cancer. • The hybrid model has the highest diagnostic efficiency for the LNM prediction. • The score of the hybrid model can reflect the DFS of operable cervical cancer.

Computer-aided diagnosis of distal metastasis in non-small cell lung cancer by low-dose CT based radiomics and deep learning signatures

BACKGROUND

This study aimed to develop and validate radiomics and deep learning (DL) signatures for predicting distal metastasis (DM) of non-small cell lung cancer (NSCLC) in low-dose computed tomography (LDCT).

METHODS

Images and clinical data were retrospectively collected for 381 NSCLC patients and prospectively collected for 114 patients at the Fifth Affiliated Hospital of Sun Yat-Sen University. Additionally, we enrolled 179 patients from the Jiangmen Central Hospital to externally validate the signatures. Machine-learning algorithms were employed to develop radiomics signature while the DL signature was developed using neural architecture search. The diagnostic efficiency was primarily quantified with the area under receiver operating characteristic curve (AUC). We interpreted the reasoning process of the radiomics signature and DL signature by radiomics voxel mapping and attention weight tracking.

RESULTS

A total of 674 patients with pathologically-confirmed NSCLC were included from two institutions, with 143 of them having DM. The radiomics signature achieved AUCs of 0.885, 0.854, and 0.733 in the internal validation, prospective validation, and external validation while those for DL signature were 0.893, 0.786, and 0.780. The proposed signatures achieved a promising performance in predicting the DM of NSCLC and outperformed the approaches proposed in previous studies. Interpretability analysis revealed that both radiomics and DL signatures could detect the variations among voxels inside tumors, which helped in identifying the DM of NSCLC.

CONCLUSIONS

Our study demonstrates the potential of LDCT-based radiomics and DL signatures for predicting DM in NSCLC. These signatures could help improve lung cancer screening regarding further diagnostic tests and treatment strategies.

Exploring non-invasive precision treatment in non-small cell lung cancer patients through deep learning radiomics across imaging features and molecular phenotypes

BACKGROUND

Accurate prediction of tumor molecular alterations is vital for optimizing cancer treatment. Traditional tissue-based approaches encounter limitations due to invasiveness, heterogeneity, and molecular dynamic changes. We aim to develop and validate a deep learning radiomics framework to obtain imaging features that reflect various molecular changes, aiding first-line treatment decisions for cancer patients.

METHODS

We conducted a retrospective study involving 508 NSCLC patients from three institutions, incorporating CT images and clinicopathologic data. Two radiomic scores and a deep network feature were constructed on three data sources in the 3D tumor region. Using these features, we developed and validated the 'Deep-RadScore,' a deep learning radiomics model to predict prognostic factors, gene mutations, and immune molecule expression levels.

FINDINGS

The Deep-RadScore exhibits strong discrimination for tumor molecular features. In the independent test cohort, it achieved impressive AUCs: 0.889 for lymphovascular invasion, 0.903 for pleural invasion, 0.894 for T staging; 0.884 for EGFR and ALK, 0.896 for KRAS and PIK3CA, 0.889 for TP53, 0.895 for ROS1; and 0.893 for PD-1/PD-L1. Fusing features yielded optimal predictive power, surpassing any single imaging feature. Correlation and interpretability analyses confirmed the effectiveness of customized deep network features in capturing additional imaging phenotypes beyond known radiomic features.

INTERPRETATION

This proof-of-concept framework demonstrates that new biomarkers across imaging features and molecular phenotypes can be provided by fusing radiomic features and deep network features from multiple data sources. This holds the potential to offer valuable insights for radiological phenotyping in characterizing diverse tumor molecular alterations, thereby advancing the pursuit of non-invasive personalized treatment for NSCLC patients.

Predicting occult lymph node metastasis in solid-predominantly invasive lung adenocarcinoma across multiple centers using radiomics-deep learning fusion model

BACKGROUND

In solid-predominantly invasive lung adenocarcinoma (SPILAC), occult lymph node metastasis (OLNM) is pivotal for determining treatment strategies. This study seeks to develop and validate a fusion model combining radiomics and deep learning to predict OLNM preoperatively in SPILAC patients across multiple centers.

METHODS

In this study, 1325 cT1a-bN0M0 SPILAC patients from six hospitals were retrospectively analyzed and divided into pathological nodal positive (pN+) and negative (pN-) groups. Three predictive models for OLNM were developed: a radiomics model employing decision trees and support vector machines; a deep learning model using ResNet-18, ResNet-34, ResNet-50, DenseNet-121, and Swin Transformer, initialized randomly or pre-trained on large-scale medical data; and a fusion model integrating both approaches using addition and concatenation techniques. The model performance was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC).

RESULTS

All patients were assigned to four groups: training set (n = 470), internal validation set (n = 202), and independent test set 1 (n = 227) and 2 (n = 426). Among the 1325 patients, 478 (36%) had OLNM (pN+). The fusion model, combining radiomics with pre-trained ResNet-18 features via concatenation, outperformed others with an average AUC (aAUC) of 0.754 across validation and test sets, compared to aAUCs of 0.715 for the radiomics model and 0.676 for the deep learning model.

CONCLUSION

The radiomics-deep learning fusion model showed promising ability to generalize in predicting OLNM from CT scans, potentially aiding personalized treatment for SPILAC patients across multiple centers.

Application of radiomics in diagnosis and treatment of lung cancer

Radiomics has become a research field that involves the process of converting standard nursing images into quantitative image data, which can be combined with other data sources and subsequently analyzed using traditional biostatistics or artificial intelligence (Al) methods. Due to the capture of biological and pathophysiological information by radiomics features, these quantitative radiomics features have been proven to provide fast and accurate non-invasive biomarkers for lung cancer risk prediction, diagnosis, prognosis, treatment response monitoring, and tumor biology. In this review, radiomics has been emphasized and discussed in lung cancer research, including advantages, challenges, and drawbacks.

Clinical applications of radiomics in non-small cell lung cancer patients with immune checkpoint inhibitor-related pneumonitis

Immune checkpoint inhibitors (ICIs) modulate the body's immune function to treat tumors but may also induce pneumonitis. Immune checkpoint inhibitor-related pneumonitis (ICIP) is a serious immune-related adverse event (irAE). Immunotherapy is currently approved as a first-line treatment for non-small cell lung cancer (NSCLC), and the incidence of ICIP in NSCLC patients can be as high as 5%-19% in clinical practice. ICIP can be severe enough to lead to the death of NSCLC patients, but there is a lack of a gold standard for the diagnosis of ICIP. Radiomics is a method that uses computational techniques to analyze medical images (e.g., CT, MRI, PET) and extract important features from them, which can be used to solve classification and regression problems in the clinic. Radiomics has been applied to predict and identify ICIP in NSCLC patients in the hope of transforming clinical qualitative problems into quantitative ones, thus improving the diagnosis and treatment of ICIP. In this review, we summarize the pathogenesis of ICIP and the process of radiomics feature extraction, review the clinical application of radiomics in ICIP of NSCLC patients, and discuss its future application prospects.

Improving diagnosis and prognosis of lung cancer using vision transformers: a scoping review

BACKGROUND

Vision transformer-based methods are advancing the field of medical artificial intelligence and cancer imaging, including lung cancer applications. Recently, many researchers have developed vision transformer-based AI methods for lung cancer diagnosis and prognosis.

OBJECTIVE

This scoping review aims to identify the recent developments on vision transformer-based AI methods for lung cancer imaging applications. It provides key insights into how vision transformers complemented the performance of AI and deep learning methods for lung cancer. Furthermore, the review also identifies the datasets that contributed to advancing the field.

METHODS

In this review, we searched Pubmed, Scopus, IEEEXplore, and Google Scholar online databases. The search terms included intervention terms (vision transformers) and the task (i.e., lung cancer, adenocarcinoma, etc.). Two reviewers independently screened the title and abstract to select relevant studies and performed the data extraction. A third reviewer was consulted to validate the inclusion and exclusion. Finally, the narrative approach was used to synthesize the data.

RESULTS

Of the 314 retrieved studies, this review included 34 studies published from 2020 to 2022. The most commonly addressed task in these studies was the classification of lung cancer types, such as lung squamous cell carcinoma versus lung adenocarcinoma, and identifying benign versus malignant pulmonary nodules. Other applications included survival prediction of lung cancer patients and segmentation of lungs. The studies lacked clear strategies for clinical transformation. SWIN transformer was a popular choice of the researchers; however, many other architectures were also reported where vision transformer was combined with convolutional neural networks or UNet model. Researchers have used the publicly available lung cancer datasets of the lung imaging database consortium and the cancer genome atlas. One study used a cluster of 48 GPUs, while other studies used one, two, or four GPUs.

CONCLUSION

It can be concluded that vision transformer-based models are increasingly in popularity for developing AI methods for lung cancer applications. However, their computational complexity and clinical relevance are important factors to be considered for future research work. This review provides valuable insights for researchers in the field of AI and healthcare to advance the state-of-the-art in lung cancer diagnosis and prognosis. We provide an interactive dashboard on lung-cancer.onrender.com/ .

Multi-view radiomics and deep learning modeling for prostate cancer detection based on multi-parametric MRI

Introduction

This study aims to develop an imaging model based on multi-parametric MR images for distinguishing between prostate cancer (PCa) and prostate hyperplasia.

Methods

A total of 236 subjects were enrolled and divided into training and test sets for model construction. Firstly, a multi-view radiomics modeling strategy was designed in which different combinations of radiomics feature categories (original, LoG, and wavelet) were compared to obtain the optimal input feature sets. Minimum-redundancy maximum-relevance (mRMR) selection and least absolute shrinkage selection operator (LASSO) were used for feature reduction, and the next logistic regression method was used for model construction. Then, a Swin Transformer architecture was designed and trained using transfer learning techniques to construct the deep learning models (DL). Finally, the constructed multi-view radiomics and DL models were combined and compared for model selection and nomogram construction. The prediction accuracy, consistency, and clinical benefit were comprehensively evaluated in the model comparison.

Results

The optimal input feature set was found when LoG and wavelet features were combined, while 22 and 17 radiomic features in this set were selected to construct the ADC and T2 multi-view radiomic models, respectively. ADC and T2 DL models were built by transferring learning from a large number of natural images to a relatively small sample of prostate images. All individual and combined models showed good predictive accuracy, consistency, and clinical benefit. Compared with using only an ADC-based model, adding a T2-based model to the combined model would reduce the model's predictive performance. The ADCCombinedScore model showed the best predictive performance among all and was transformed into a nomogram for better use in clinics.

Discussion

The constructed models in our study can be used as a predictor in differentiating PCa and BPH, thus helping clinicians make better clinical treatment decisions and reducing unnecessary prostate biopsies.

A Multi-Modal Heterogeneous Graph Forest to Predict Lymph Node Metastasis of Non-Small Cell Lung Cancer

Lymph node metastasis (LNM) is critical for treatment decision-making for cancer patients, but it is difficult to diagnose accurately before surgery. Machine learning can learn nontrivial knowledge from multi-modal data to support accurate diagnosis. In this paper, we proposed a Multi-modal Heterogeneous Graph Forest (MHGF) approach to extract the deep representations of LNM from multi-modal data. Specifically, we first extracted the deep image features from CT images to represent the pathological anatomic extent of the primary tumor (pathological T stage) using a ResNet-Trans network. And then, a heterogeneous graph with six vertices and seven bi-directional relations was defined by medical experts to describe the possible relations between the clinical and image features. After that, we proposed a graph forest approach to construct the sub-graphs by removing each vertex in the complete graph iteratively. Finally, we used graph neural networks to learn the representations of each sub-graph in the forest to predict LNM and averaged all the prediction results as final results. We conducted experiments on 681 patients' multi-modal data. The proposed MHGF achieves the best performances with a 0.806 AUC value and 0.513 AP value compared with state-of-art machine learning and deep learning methods. The results indicate that the graph method can explore the relations between different types of features to learn effective deep representations for LNM prediction. Moreover, we found that the deep image features about the pathological anatomic extent of the primary tumor are useful for LNM prediction. And the graph forest approach can further improve the generalization ability and stability of the LNM prediction model.