Toward automatic prediction of EGFR mutation status in pulmonary adenocarcinoma with 3D deep learning

Prediction of EGFR Mutations in Lung Adenocarcinoma via CT Images: A Comparative Study of Intratumoral and Peritumoral Radiomics, Deep Learning, and Fusion Models

RATIONALE AND OBJECTIVES

This study aims to analyze the intratumoral and peritumoral characteristics of lung adenocarcinoma patients on the basis of chest CT images via radiomic and deep learning methods and to develop and validate a multimodel fusion strategy for predicting epidermal growth factor receptor (EGFR) mutation statuses.

MATERIALS AND METHODS

Retrospective data from 826 lung adenocarcinoma patients across two hospitals were collected. Data from center1 were used for model training and internal validation, while data from center2 were reserved for external validation. Tumor segmentation was performed using the nnUNet network, and volumes of interest (VOIs) for the tumor and its peritumoral regions (2 mm, 4 mm, 6 mm, 8 mm, 10 mm) were subsequently derived.Radiomics features were extracted from various VOIs using PyRadiomics, and radiomics models were developed using Lasso and multiple machine learning algorithms.Using 2D, 2.5D, and 3D images derived from different VOIs as inputs, multiple deep learning models were trained and their performances compared.The radiomics and deep learning models demonstrating the best predictive performance were selected and integrated with clinical models for model fusion.Multi-model fusion of clinical, radiomics, and deep learning features was achieved using feature-level fusion and various decision-level fusion strategies, including hard voting, soft voting, and stacking ensemble.The predictive performances of various fusion models were evaluated and compared systematically.

RESULTS

Among the available radiomic models, the model based on intratumoral and peritumoral 2-mm regions (VOI_Comb2) achieved the best performance on the internal and external validation sets (AUC=0.843 and 0.803, respectively). Compared with 2D and 2.5D deep learning models, the 3D deep learning model demonstrated superior predictive performance. The 3D deep model based on the VOI_Comb2 region achieved the highest AUC among all the deep learning models on the internal and external validation sets (AUC=0.839 and 0.814, respectively). Among the fusion models, the soft voting strategy achieved the highest AUC on the internal and external validation sets, reaching 0.925 and 0.889, respectively. On the external validation set, the AUC of the soft voting model was significantly greater than that of the hard voting model, early fusion model, or any single modality model.

CONCLUSION

This study demonstrates that combining radiomic and deep learning models based on intratumoral and peritumoral regions is an effective method for capturing comprehensive imaging features in lung adenocarcinoma. The multimodal fusion approach using soft voting leverages the strengths of each modality and provides a robust framework for advanced image feature extraction to support personalized treatment.

Habitat Radiomics and Deep Learning Features Based on CT for Predicting Lymphovascular Invasion in T1-stage Lung Adenocarcinoma: A Multicenter Study

RATIONALE AND OBJECTIVES

The research aims to examine how CT-derived habitat radiomics can be used to predict lymphovascular invasion (LVI) in patients with T1-stage lung adenocarcinoma (LUAD), and compare its effectiveness to traditional radiomics and deep learning (DL) models.

MATERIALS AND METHODS

We retrospectively analyzed 349 T1-stage LUAD patients from three centers from January 2021 to March 2024. The K-means algorithm was utilized to cluster CT images and apparent diffusion coefficient maps. Following features selection, we constructed three types of models, namely radiomics, habitat, and DL to identify patients with LVI. The evaluation of all models was conducted by employing the area under the receiver operating characteristic curve (AUC), calibration curves and decision curve analysis.

RESULTS

349 eligible patients were divided into an internal training set of 210 and an external test set of 139. We identified four distinct habitats, with the AUC for the overall habitat area outperforming that of the four sub-areas. Within the test set, the habitat model reached a higher AUC of 0.941 in contrast to the radiomics model at 0.918 and the deep learning model at 0.896.

CONCLUSION

CT-based habitat radiomics shows promise in predicting LVI in T1-stage LUAD patients, with the habitat signature demonstrating superior performance and significant advantages in identifying patients who are LVI-positive.

Deep learning radiomics for the prediction of epidermal growth factor receptor mutation status based on MRI in brain metastasis from lung adenocarcinoma patients

BACKGROUND

Early and accurate identification of epidermal growth factor receptor (EGFR) mutation status in non-small cell lung cancer (NSCLC) patients with brain metastases is critical for guiding targeted therapy. This study aimed to develop a deep learning radiomics model utilizing multi-sequence magnetic resonance imaging (MRI) to differentiate between EGFR mutant type (MT) and wild type (WT).

METHODS

In this retrospective study, 288 NSCLC patients with confirmed brain metastases were enrolled, including 106 with EGFR MT and 182 with EGFR WT. All patients were randomly divided into a training dataset (75%) and a validation dataset (25%). Radiomics and deep learning features were extracted from the brain metastatic lesions using contrast-enhanced T1-weighted (T1CE) and T2-weighted (T2W) MRI images. Features extraction and selection were performed using the least absolute shrinkage and selection operator (LASSO) and ResNet34. The predictive performance of the signatures for EGFR mutation status was assessed using receiver operating characteristic (ROC) curves and area under the curve (AUC) analyses.

RESULTS

No significant differences were found between the training and validation datasets. A four-feature radiomics signature (RS) demonstrated excellent predictive accuracy for EGFR MT, with α-binormal-based and empirical AUCs of 0.931 (95% CI: 0.880-0.940) and 0.926 (95% CI: 0.877-0.933), respectively. Incorporating deep learning signature (DLS) further enhanced the model's performance, achieving α-binormal-based and empirical AUCs of 0.943 (95% CI: 0.921-0.965) and 0.938 (95% CI: 0.914-0.962) in the training dataset. These findings were confirmed in the validation dataset, with AUCs of 0.936 (95% CI: 0.917-0.955) and 0.921 (95% CI: 0.901-0.941), demonstrating robust and consistent predictive performance.

CONCLUSIONS

The multi-sequence MRI-based deep learning radiomics model exhibited high efficacy in predicting EGFR mutation status in NSCLC patients with brain metastases. This approach, which integrates advanced radiological features with deep learning techniques, offers a non-invasive and accurate method for determining EGFR mutation status, potentially guiding personalized treatment decisions in clinical practice.

Preoperative assessment of tertiary lymphoid structures in stage I lung adenocarcinoma using CT radiomics: a multicenter retrospective cohort study

OBJECTIVE

To develop a multimodal predictive model, Radiomics Integrated TLSs System (RAITS), based on preoperative CT radiomic features for the identification of TLSs in stage I lung adenocarcinoma patients and to evaluate its potential in prognosis stratification and guiding personalized treatment.

METHODS

The most recent preoperative chest CT thin-slice scans and postoperative hematoxylin and eosin-stained pathology sections of patients diagnosed with stage I LUAD were retrospectively collected. Tumor segmentation was achieved using an automatic virtual adversarial training segmentation algorithm based on a three-dimensional U-shape convolutional neural network (3D U-Net). Radiomic features were extracted from the tumor and peritumoral areas, with extensions of 2 mm, 4 mm, 6 mm, and 8 mm, respectively, and deep learning image features were extracted through a convolutional neural network. Subsequently, the RAITS was constructed. The performance of RAITS was then evaluated in both the train and validation cohorts.

RESULTS

RAITS demonstrated superior AUC, sensitivity, and specificity in both the training and external validation cohorts, outperforming traditional unimodal models. In the validation cohort, RAITS achieved an AUC of 0.78 (95% CI, 0.69-0.88) and showed higher net benefits across most threshold ranges. RAITS exhibited strong discriminative ability in risk stratification, with p < 0.01 in the training cohort and p = 0.02 in the validation cohort, consistent with the actual predictive performance of TLSs, where TLS-positive patients had significantly higher recurrence-free survival (RFS) compared to TLS-negative patients (p = 0.04 in the training cohort, p = 0.02 in the validation cohort).

CONCLUSION

As a multimodal predictive model based on preoperative CT radiomic features, RAITS demonstrated excellent performance in identifying TLSs in stage I LUAD and holds potential value in clinical decision-making.

Non-invasive, fast, and high-performance EGFR gene mutation prediction method based on deep transfer learning and model stacking for patients with Non-Small Cell Lung Cancer

Purpose

To propose an intelligent, non-invasive, highly precise, and rapid method to predict the mutation status of the Epidermal Growth Factor Receptor (EGFR) to accelerate treatment with Tyrosine Kinase Inhibitor (TKI) for patients with untreated adenocarcinoma Non-Small Cell Lung Cancer.

Materials and methods

Real-world data from 521 patients with adenocarcinoma NSCLC who performed a CT scan and underwent surgery or pathological biopsy to determine EGFR gene mutation between January 2021 and July 2022, is collected. Solutions to the problems that prevent the model from achieving very high precision, namely: human errors made during the annotation of the database and the low precision of the output decision of the model, are proposed. Thus, among the 521 analyzed cases, only 40 were selected as patients with EGFR gene mutation and 98 cases with wild-type EGFR.

Results

The proposed model is trained, validated, and tested on 12,040 2D images extracted from the 138 CT scans images where patients were randomly partitioned into training (80 %) and test (20 %) sets. The performance obtained for EGFR gene mutation prediction was 95.22 % for accuracy, 960.2 for F1_score, 95.89 % for precision, 96.92 % for sensitivity, 94.01 % for Cohen kappa, and 98 % for AUC.

Conclusion

An EGFR gene mutation status prediction method, with high-performance thanks to an intelligent prediction model entrained by highly accurate annotated data is proposed. The outcome of this project will facilitate rapid decision-making when applying a TKI as an initial treatment.

PET/CT radiomics and deep learning in the diagnosis of benign and malignant pulmonary nodules: progress and challenges

Lung cancer is currently the leading cause of cancer-related deaths, and early diagnosis and screening can significantly reduce its mortality rate. Since some early-stage lung cancers lack obvious clinical symptoms and only present as pulmonary nodules (PNs) in imaging examinations, accurately determining the benign or malignant nature of PNs is crucial for improving patient survival rates. 18F-FDG PET/CT is important in diagnosing PNs, but its specificity needs improvement. Radiomics can provide information beyond traditional visual assessment, overcoming its limitations by extracting high-throughput quantitative features from medical images. Radiomics features based on 18F-FDG PET/CT and deep learning methods have shown great potential in the noninvasive diagnosis of PNs. This paper reviews the latest advancements in these methods and discusses their contributions to improving diagnostic accuracy and the challenges they face.

Artificial intelligence-driven multiomics predictive model for abdominal aortic aneurysm subtypes to identify heterogeneous immune cell infiltration and predict disease progression

BACKGROUND

Abdominal aortic aneurysm (AAA) poses a significant health risk and is influenced by various compositional features. This study aimed to develop an artificial intelligence-driven multiomics predictive model for AAA subtypes to identify heterogeneous immune cell infiltration and predict disease progression. Additionally, we investigated neutrophil heterogeneity in patients with different AAA subtypes to elucidate the relationship between the immune microenvironment and AAA pathogenesis.

METHODS

This study enrolled 517 patients with AAA, who were clustered using k-means algorithm to identify AAA subtypes and stratify the risk. We utilized residual convolutional neural network 200 to annotate and extract contrast-enhanced computed tomography angiography images of AAA. A precise predictive model for AAA subtypes was established using clinical, imaging, and immunological data. We performed a comparative analysis of neutrophil levels in the different subgroups and immune cell infiltration analysis to explore the associations between neutrophil levels and AAA. Quantitative polymerase chain reaction, Western blotting, and enzyme-linked immunosorbent assay were performed to elucidate the interplay between CXCL1, neutrophil activation, and the nuclear factor (NF)-κB pathway in AAA pathogenesis. Furthermore, the effect of CXCL1 silencing with small interfering RNA was investigated.

RESULTS

Two distinct AAA subtypes were identified, one clinically more severe and more likely to require surgical intervention. The CNN effectively detected AAA-associated lesion regions on computed tomography angiography, and the predictive model demonstrated excellent ability to discriminate between patients with the two identified AAA subtypes (area under the curve, 0.927). Neutrophil activation, AAA pathology, CXCL1 expression, and the NF-κB pathway were significantly correlated. CXCL1, NF-κB, IL-1β, and IL-8 were upregulated in AAA. CXCL1 silencing downregulated NF-κB, interleukin-1β, and interleukin-8.

CONCLUSION

The predictive model for AAA subtypes demonstrated accurate and reliable risk stratification and clinical management. CXCL1 overexpression activated neutrophils through the NF-κB pathway, contributing to AAA development. This pathway may, therefore, be a therapeutic target in AAA.

Radiogenomics: bridging the gap between imaging and genomics for precision oncology

Genomics allows the tracing of origin and evolution of cancer at molecular scale and underpin modern cancer diagnosis and treatment systems. Yet, molecular biomarker-guided clinical decision-making encounters major challenges in the realm of individualized medicine, consisting of the invasiveness of procedures and the sampling errors due to high tumor heterogeneity. By contrast, medical imaging enables noninvasive and global characterization of tumors at a low cost. In recent years, radiomics has overcomes the limitations of human visual evaluation by high-throughput quantitative analysis, enabling the comprehensive utilization of the vast amount of information underlying radiological images. The cross-scale integration of radiomics and genomics (hereafter radiogenomics) has the enormous potential to enhance cancer decoding and act as a catalyst for digital precision medicine. Herein, we provide a comprehensive overview of the current framework and potential clinical applications of radiogenomics in patient care. We also highlight recent research advances to illustrate how radiogenomics can address common clinical problems in solid tumors such as breast cancer, lung cancer, and glioma. Finally, we analyze existing literature to outline challenges and propose solutions, while also identifying future research pathways. We believe that the perspectives shared in this survey will provide a valuable guide for researchers in the realm of radiogenomics aiming to advance precision oncology.

Prediction of epidermal growth factor receptor (EGFR) mutation status in lung adenocarcinoma patients on computed tomography (CT) images using 3-dimensional (3D) convolutional neural network

Background

Noninvasively detecting epidermal growth factor receptor (EGFR) mutation status in lung adenocarcinoma patients before targeted therapy remains a challenge. This study aimed to develop a 3-dimensional (3D) convolutional neural network (CNN)-based deep learning model to predict EGFR mutation status using computed tomography (CT) images.

Methods

We retrospectively collected 660 patients from 2 large medical centers. The patients were divided into training (n=528) and external test (n=132) sets according to hospital source. The CNN model was trained in a supervised end-to-end manner, and its performance was evaluated using an external test set. To compare the performance of the CNN model, we constructed 1 clinical and 3 radiomics models. Furthermore, we constructed a comprehensive model combining the highest-performing radiomics and CNN models. The receiver operating characteristic (ROC) curves were used as primary measures of performance for each model. Delong test was used to compare performance differences between different models.

Results

Compared with the clinical [training set, area under the curve (AUC) =69.6%, 95% confidence interval (CI), 0.661-0.732; test set, AUC =68.4%, 95% CI, 0.609-0.752] and the highest-performing radiomics models (training set, AUC =84.3%, 95% CI, 0.812-0.873; test set, AUC =72.4%, 95% CI, 0.653-0.794) models, the CNN model (training set, AUC =94.3%, 95% CI, 0.920-0.961; test set, AUC =94.7%, 95% CI, 0.894-0.978) had significantly better predictive performance for predicting EGFR mutation status. In addition, compared with the comprehensive model (training set, AUC =95.7%, 95% CI, 0.942-0.971; test set, AUC =87.4%, 95% CI, 0.820-0.924), the CNN model had better stability.

Conclusions

The CNN model has excellent performance in non-invasively predicting EGFR mutation status in patients with lung adenocarcinoma and is expected to become an auxiliary tool for clinicians.

The digital revolution in pathology: Towards a smarter approach to research and treatment

Artificial intelligence (AI) applications in oncology are at the forefront of transforming healthcare during the Fourth Industrial Revolution, driven by the digital data explosion. This review provides an accessible introduction to the field of AI, presenting a concise yet structured overview of the foundations of AI, including expert systems, classical machine learning, and deep learning, along with their contextual application in clinical research and healthcare. We delve into the current applications of AI in oncology, with a particular focus on diagnostic imaging and pathology. Numerous AI tools have already received regulatory approval, and more are under active development, bringing clear benefits but not without challenges. We discuss the importance of data security, the need for transparent and interpretable models, and the ethical considerations that must guide AI development in healthcare. By providing a perspective on the opportunities and challenges, this review aims to inform and guide researchers, clinicians, and policymakers in the adoption of AI in oncology.

Brain metastasis magnetic resonance imaging-based deep learning for predicting epidermal growth factor receptor (EGFR) mutation and subtypes in metastatic non-small cell lung cancer

Background

The preoperative identification of epidermal growth factor receptor (EGFR) mutations and subtypes based on magnetic resonance imaging (MRI) of brain metastases (BM) is necessary to facilitate individualized therapy. This study aimed to develop a deep learning model to preoperatively detect EGFR mutations and identify the location of EGFR mutations in patients with non-small cell lung cancer (NSCLC) and BM.

Methods

We included 160 and 72 patients who underwent contrast-enhanced T1-weighted (T1w-CE) and T2-weighted (T2W) MRI at Liaoning Cancer Hospital and Institute (center 1) and Shengjing Hospital of China Medical University (center 2) to form a training cohort and an external validation cohort, respectively. A multiscale feature fusion network (MSF-Net) was developed by adaptively integrating features based on different stages of residual network (ResNet) 50 and by introducing channel and spatial attention modules. The external validation set from center 2 was used to assess the performance of MSF-Net and to compare it with that of handcrafted radiomics features. Receiver operating characteristic (ROC) curves, accuracy, precision, recall, and F1-score were used to evaluate the effectiveness of the models. Gradient-weighted class activation mapping (Grad-CAM) was used to demonstrate the attention of the MSF-Net model.

Results

The developed MSF-Net generated a better diagnostic performance than did the handcrafted radiomics in terms of the microaveraged area under the curve (AUC) (MSF-Net: 0.91; radiomics: 0.80) and macroaveraged AUC (MSF-Net: 0.90; radiomics: 0.81) for predicting EGFR mutations and subtypes.

Conclusions

This study provides an end-to-end and noninvasive imaging tool for the preoperative prediction of EGFR mutation status and subtypes based on BM, which may be helpful for facilitating individualized clinical treatment plans.

Using Vision Transformer for high robustness and generalization in predicting EGFR mutation status in lung adenocarcinoma

BACKGROUND

Lung adenocarcinoma is a common cause of cancer-related deaths worldwide, and accurate EGFR genotyping is crucial for optimal treatment outcomes. Conventional methods for identifying the EGFR genotype have several limitations. Therefore, we proposed a deep learning model using non-invasive CT images to predict EGFR mutation status with robustness and generalizability.

METHODS

A total of 525 patients were enrolled at the local hospital to serve as the internal data set for model training and validation. In addition, a cohort of 30 patients from the publicly available Cancer Imaging Archive Data Set was selected for external testing. All patients underwent plain chest CT, and their EGFR mutation status labels were categorized as either mutant or wild type. The CT images were analyzed using a self-attention-based ViT-B/16 model to predict the EGFR mutation status, and the model's performance was evaluated. To produce an attention map indicating the suspicious locations of EGFR mutations, Grad-CAM was utilized.

RESULTS

The ViT deep learning model achieved impressive results, with an accuracy of 0.848, an AUC of 0.868, a sensitivity of 0.924, and a specificity of 0.718 on the validation cohort. Furthermore, in the external test cohort, the model achieved comparable performances, with an accuracy of 0.833, an AUC of 0.885, a sensitivity of 0.900, and a specificity of 0.800.

CONCLUSIONS

The ViT model demonstrates a high level of accuracy in predicting the EGFR mutation status of lung adenocarcinoma patients. Moreover, with the aid of attention maps, the model can assist clinicians in making informed clinical decisions.

GMILT: A Novel Transformer Network That Can Noninvasively Predict EGFR Mutation Status

Noninvasively and accurately predicting the epidermal growth factor receptor (EGFR) mutation status is a clinically vital problem. Moreover, further identifying the most suspicious area related to the EGFR mutation status can guide the biopsy to avoid false negatives. Deep learning methods based on computed tomography (CT) images may improve the noninvasive prediction of EGFR mutation status and potentially help clinicians guide biopsies by visual methods. Inspired by the potential inherent links between EGFR mutation status and invasiveness information, we hypothesized that the predictive performance of a deep learning network can be improved through extra utilization of the invasiveness information. Here, we created a novel explainable transformer network for EGFR classification named gated multiple instance learning transformer (GMILT) by integrating multi-instance learning and discriminative weakly supervised feature learning. Pathological invasiveness information was first introduced into the multitask model as embeddings. GMILT was trained and validated on a total of 512 patients with adenocarcinoma and tested on three datasets (the internal test dataset, the external test dataset, and The Cancer Imaging Archive (TCIA) public dataset). The performance (area under the curve (AUC) =0.772 on the internal test dataset) of GMILT exceeded that of previously published methods and radiomics-based methods (i.e., random forest and support vector machine) and attained a preferable generalization ability (AUC =0.856 in the TCIA test dataset and AUC =0.756 in the external dataset). A diameter-based subgroup analysis further verified the efficiency of our model (most of the AUCs exceeded 0.772) to noninvasively predict EGFR mutation status from computed tomography (CT) images. In addition, because our method also identified the "core area" of the most suspicious area related to the EGFR mutation status, it has the potential ability to guide biopsies.

[The contribution of artificial intelligence (AI) subsequent to the processing of thoracic imaging]

The contribution of artificial intelligence (AI) to medical imaging is currently the object of widespread experimentation. The development of deep learning (DL) methods, particularly convolution neural networks (CNNs), has led to performance gains often superior to those achieved by conventional methods such as machine learning. Radiomics is an approach aimed at extracting quantitative data not accessible to the human eye from images expressing a disease. The data subsequently feed machine learning models and produce diagnostic or prognostic probabilities. As for the multiple applications of AI methods in thoracic imaging, they are undergoing evaluation. Chest radiography is a practically ideal field for the development of DL algorithms able to automatically interpret X-rays. Current algorithms can detect up to 14 different abnormalities present either in isolation or in combination. Chest CT is another area offering numerous AI applications. Various algorithms have been specifically formed and validated for the detection and characterization of pulmonary nodules and pulmonary embolism, as well as segmentation and quantitative analysis of the extent of diffuse lung diseases (emphysema, infectious pneumonias, interstitial lung disease). In addition, the analysis of medical images can be associated with clinical, biological, and functional data (multi-omics analysis), the objective being to construct predictive approaches regarding disease prognosis and response to treatment.

Predicting EGFR Mutation Status in Non-Small Cell Lung Cancer Using Artificial Intelligence: A Systematic Review and Meta-Analysis

RATIONALE AND OBJECTIVES

Recent advancements in artificial intelligence (AI) render a substantial promise for epidermal growth factor receptor (EGFR) mutation status prediction in non-small cell lung cancer (NSCLC). We aimed to evaluate the performance and quality of AI algorithms that use radiomics features in predicting EGFR mutation status in patient with NSCLC.

MATERIALS AND METHODS

We searched PubMed (Medline), EMBASE, Web of Science, and IEEExplore for studies published up to February 28, 2022. Studies utilizing an AI algorithm (either conventional machine learning [cML] and deep learning [DL]) for predicting EGFR mutations in patients with NSLCL were included. We extracted binary diagnostic accuracy data and constructed a bivariate random-effects model to obtain pooled sensitivity, specificity, and 95% confidence interval. This study is registered with PROSPERO, CRD42021278738.

RESULTS

Our search identified 460 studies, of which 42 were included. Thirty-five studies were included in the meta-analysis. The AI algorithms exhibited an overall area under the curve (AUC) value of 0.789 and pooled sensitivity and specificity levels of 72.2% and 73.3%, respectively. The DL algorithms outperformed cML in terms of AUC (0.822 vs. 0.775) and sensitivity (80.1% vs. 71.1%), but had lower specificity (70.0% vs. 73.8%, p-value < 0.001) compared to cML. Subgroup analysis revealed that the use of positron-emission tomography/computed tomography, additional clinical information, deep feature extraction, and manual segmentation can improve diagnostic performance.

CONCLUSION

DL algorithms can serve as a novel method for increasing predictive accuracy and thus have considerable potential for use in predicting EGFR mutation status in patient with NSCLC. We also suggest that guidelines on using AI algorithms in medical image analysis should be developed with a focus on oncologic radiomics.

Deep learning-radiomics integrated noninvasive detection of epidermal growth factor receptor mutations in non-small cell lung cancer patients

This study focused on a novel strategy that combines deep learning and radiomics to predict epidermal growth factor receptor (EGFR) mutations in patients with non-small cell lung cancer (NSCLC) using computed tomography (CT). A total of 1280 patients with NSCLC who underwent contrast-enhanced CT scans and EGFR mutation testing before treatment were selected for the final study. Regions of interest were segmented from the CT images to extract radiomics features and obtain tumor images. These tumor images were input into a convolutional neural network model to extract 512 image features, which were combined with radiographic features and clinical data to predict the EGFR mutation. The generalization performance of the model was evaluated using external institutional data. The internal and external datasets contained 324 and 130 EGFR mutants, respectively. Sex, height, weight, smoking history, and clinical stage were significantly different between the EGFR-mutant patient groups. The EGFR mutations were predicted by combining the radiomics and clinical features, and an external validation dataset yielded an area under the curve (AUC) value of 0.7038. The model utilized 1280 tumor images, radiomics features, and clinical characteristics as input data and exhibited an AUC of approximately 0.81 and 0.78 during the primary cohort and external validation, respectively. These results indicate the feasibility of integrating radiomics analysis with deep learning for predicting EGFR mutations. CT-image-based genetic testing is a simple EGFR mutation prediction method, which can improve the prognosis of NSCLC patients and help establish personalized treatment strategies.

Non-invasively Discriminating the Pathological Subtypes of Non-small Cell Lung Cancer with Pretreatment 18F-FDG PET/CT Using Deep Learning

RATIONALE AND OBJECTIVES

To develop an end-to-end deep learning (DL) model for non-invasively predicting non-small cell lung cancer (NSCLC) pathological subtypes based on 18F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) images, and to explore the potential value of DL technology.

MATERIALS AND METHODS

Preoperative 18F-FDG PET/CT images of 189 patients with NSCLC were retrospectively collected. The whole cohort was randomly divided into a training cohort, a validation cohort, and an internal/extended test cohort at the ratio of 6:2:2 after preprocessing the images. In the training and validation cohorts, seven DL models-Shufflenet, VGG16, Googlenet, Inception v3, Resnet50, Densenet201, and Mobilenet v2-were trained and optimized. The generalization ability and clinical utility of the optimal model were evaluated in the internal and extended test cohorts. Moreover, Spearman's correlation analysis was used to evaluate the correlation between DL features and traditional radiological features such as tumor size and maximum standardized uptake values (SUVmax).

RESULTS

Some DL features were significantly correlated with SUVmax and tumor size (P < 0.05). The Mobilenet v2 model achieved the best performance during the model development and validation phases. In the internal test group (area under the receiver operating characteristic curve [AUC]: 0.744, area under the precision-recall curve [AP]: 0.759) and extended test group (AUC: 0.767, AP: 0.768), the Mobilenet v2 model showed good generalization ability and reproducibility. Meanwhile, the decision curve analysis revealed that patients can benefit from the decisions made based on the Mobilenet v2 model.

CONCLUSION

DL models offer great potential for classifying NSCLC pathological subtypes. Specifically, the Mobilenet v2 model performs well at end-to-end non-invasive pathological subtype stratification of NSCLC.

Lung-PNet: An Automated Deep Learning Model for the Diagnosis of Invasive Adenocarcinoma in Pure Ground-Glass Nodules on Chest CT

BACKGROUND. Pure ground-glass nodules (pGGNs) on chest CT representing invasive adenocarcinoma (IAC) warrant lobectomy with lymph node resection. For pGGNs representing other entities, close follow-up or sublobar resection without node dissection may be appropriate. OBJECTIVE. The purpose of this study was to develop and validate an automated deep learning model for differentiation of pGGNs on chest CT representing IAC from those representing atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), and minimally invasive adenocarcinoma (MIA). METHODS. This retrospective study included 402 patients (283 women, 119 men; mean age, 53.2 years) with a total of 448 pGGNs on noncontrast chest CT that were resected from January 2019 to June 2022 and were histologically diagnosed as AAH (n = 29), AIS (n = 83), MIA (n = 235), or IAC (n = 101). Lung-PNet, a 3D deep learning model, was developed for automatic segmentation and classification (probability of IAC vs other entities) of pGGNs on CT. Nodules resected from January 2019 to December 2021 were randomly allocated to training (n = 327) and internal test (n = 82) sets. Nodules resected from January 2022 to June 2022 formed a holdout test set (n = 39). Segmentation performance was assessed with Dice coefficients with radiologists' manual segmentations as reference. Classification performance was assessed by ROC AUC and precision-recall AUC (PR AUC) and compared with that of four readers (three radiologists, one surgeon). The code used is publicly available (https://github.com/XiaodongZhang-PKUFH/Lung-PNet.git). RESULTS. In the holdout test set, Dice coefficients for segmentation of IACs and of other lesions were 0.860 and 0.838, and ROC AUC and PR AUC for classification as IAC were 0.911 and 0.842. At threshold probability of 50.0% or greater for prediction of IAC, Lung-PNet had sensitivity, specificity, accuracy, and F1 score of 50.0%, 92.0%, 76.9%, and 60.9% in the holdout test set. In the holdout test set, accuracy and F1 score (p values vs Lung-PNet) for individual readers were as follows: reader 1, 51.3% (p = .02) and 48.6% (p = .008); reader 2, 79.5% (p = .75) and 75.0% (p = .10); reader 3, 66.7% (p = .35) and 68.3% (p < .001); reader 4, 71.8% (p = .48) and 42.1% (p = .18). CONCLUSION. Lung-PNet had robust performance for segmenting and classifying (IAC vs other entities) pGGNs on chest CT. CLINICAL IMPACT. This automated deep learning tool may help guide selection of surgical strategies for pGGN management.

AI/ML advances in non-small cell lung cancer biomarker discovery

Lung cancer is the leading cause of cancer deaths among both men and women, representing approximately 25% of cancer fatalities each year. The treatment landscape for non-small cell lung cancer (NSCLC) is rapidly evolving due to the progress made in biomarker-driven targeted therapies. While advancements in targeted treatments have improved survival rates for NSCLC patients with actionable biomarkers, long-term survival remains low, with an overall 5-year relative survival rate below 20%. Artificial intelligence/machine learning (AI/ML) algorithms have shown promise in biomarker discovery, yet NSCLC-specific studies capturing the clinical challenges targeted and emerging patterns identified using AI/ML approaches are lacking. Here, we employed a text-mining approach and identified 215 studies that reported potential biomarkers of NSCLC using AI/ML algorithms. We catalogued these studies with respect to BEST (Biomarkers, EndpointS, and other Tools) biomarker sub-types and summarized emerging patterns and trends in AI/ML-driven NSCLC biomarker discovery. We anticipate that our comprehensive review will contribute to the current understanding of AI/ML advances in NSCLC biomarker research and provide an important catalogue that may facilitate clinical adoption of AI/ML-derived biomarkers.

Parameterized Gompertz-Guided Morphological AutoEncoder for Predicting Pulmonary Nodule Growth

The growth rate of pulmonary nodules is a critical clue to the cancerous diagnosis. It is essential to monitor their dynamic progressions during pulmonary nodule management. To facilitate the prosperity of research on nodule growth prediction, we organized and published a temporal dataset called NLSTt with consecutive computed tomography (CT) scans. Based on the self-built dataset, we develop a visual learner to predict the growth for the following CT scan qualitatively and further propose a model to predict the growth rate of pulmonary nodules quantitatively, so that better diagnosis can be achieved with the help of our predicted results. To this end, in this work, we propose a parameterized Gempertz-guided morphological autoencoder (GM-AE) to generate any future-time-span high-quality visual appearances of pulmonary nodules from the baseline CT scan. Specifically, we parameterize a popular mathematical model for tumor growth kinetics, Gompertz, to predict future masses and volumes of pulmonary nodules. Then, we exploit the expected growth rate on the mass and volume to guide decoders generating future shape and texture of pulmonary nodules. We introduce two branches in an autoencoder to encourage shape-aware and textural-aware representation learning and integrate the generated shape into the textural-aware branch to simulate the future morphology of pulmonary nodules. We conduct extensive experiments on the self-built NLSTt dataset to demonstrate the superiority of our GM-AE to its competitive counterparts. Experiment results also reveal the learnable Gompertz function enjoys promising descriptive power in accounting for inter-subject variability of the growth rate for pulmonary nodules. Besides, we evaluate our GM-AE model on an in-house dataset to validate its generalizability and practicality. We make its code publicly available along with the published NLSTt dataset.

New research progress on 18F-FDG PET/CT radiomics for EGFR mutation prediction in lung adenocarcinoma: a review

Lung cancer, the most frequently diagnosed cancer worldwide, is the leading cause of cancer-associated deaths. In recent years, significant progress has been achieved in basic and clinical research concerning the epidermal growth factor receptor (EGFR), and the treatment of lung adenocarcinoma has also entered a new era of individualized, targeted therapies. However, the detection of lung adenocarcinoma is usually invasive. 18F-FDG PET/CT can be used as a noninvasive molecular imaging approach, and radiomics can acquire high-throughput data from standard images. These methods play an increasingly prominent role in diagnosing and treating cancers. Herein, we reviewed the progress in applying 18F-FDG PET/CT and radiomics in lung adenocarcinoma clinical research and how these data are analyzed via traditional statistics, machine learning, and deep learning to predict EGFR mutation status, all of which achieved satisfactory results. Traditional statistics extract features effectively, machine learning achieves higher accuracy with complex algorithms, and deep learning obtains significant results through end-to-end methods. Future research should combine these methods to achieve more accurate predictions, providing reliable evidence for the precision treatment of lung adenocarcinoma. At the same time, facing challenges such as data insufficiency and high algorithm complexity, future researchers must continuously explore and optimize to better apply to clinical practice.

Virtual clinical trial based on outcome modeling with iteratively redistributed extrapolation data

Virtual clinical trials (VCTs) can potentially simulate clinical trials on a computer, but their application with a limited number of past clinical cases is challenging due to the biased estimation of the statistical population. In this study, we developed ExMixup, a novel training technique based on machine learning, using iteratively redistributed extrapolated data. Information obtained from 100 patients with prostate cancer and 385 patients with oropharyngeal cancer was used to predict the recurrence after radiotherapy. Model performance was evaluated by developing outcome prediction models based on three types of training methods: training with original data (baseline), interpolation data (Mixup), and interpolation + extrapolation data (ExMixup). Two types of VCTs were conducted to predict the treatment response of patients with distinct characteristics compared to the training data obtained from patient cohorts categorized under risk classification or cancer stage. The prediction models developed with ExMixup yielded concordance indices (95% confidence intervals) of 0.751 (0.719-0.818) and 0.752 (0.734-0.785) for VCTs on the prostate and oropharyngeal cancer datasets, respectively, which significantly outperformed the baseline and Mixup models (P < 0.01). The proposed approach could enhance the ability of VCTs to predict treatment results in patients excluded from past clinical trials.

Predictive value of peritumour radiomics in the diagnosis of benign and malignant pulmonary nodules with halo sign

AIM

To evaluate peritumour radiomics in predicting benign and malignant pulmonary nodules with halo sign.

MATERIALS AND METHODS

In this retrospective study, 305 pulmonary nodules with halo sign (benign, 120; adenocarcinoma, 185) were collected. Manual segmentation was used to mark the gross tumour volume (GTV) and the peritumour volume (PTV) was established by uniform dilation (1 cm) of the tumour area in three dimensions. The GTV and PTV radiomic features were combined to produce the gross tumour and peritumour volume (GPTV). The minimum-redundancy maximum-relevance (mRMR) feature ranking method and least absolute shrinkage and selection operator (LASSO) algorithm were used to eliminate redundant radiomic features. Predictive models combined with clinical features and radiomic signatures were established. Multivarible logistic regression analysis was used to establish the combined model and develop a nomogram. Receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive performance of the model.

RESULTS

In the testing cohort, the area under the ROC curve (AUC) of the GTV, PTV, and GPTV radiomic models was 0.701 (95% CI: 0.589-0.814), 0.674 (95% CI: 0.557-0.791) and 0.755 (95% CI: 0.643-0.867), respectively. The AUC of the nomogram model based on clinical and GPTV radiomic signatures was 0.804 (95% CI: 0.707-0.901).

CONCLUSION

The nomogram model based on clinical and GPTV radiomic signatures can better predict benign and malignant pulmonary nodules with halo signs, demonstrating that the model has potential as a convenient and effective auxiliary diagnostic tool for radiologists.

Identifying Solitary Granulomatous Nodules from Solid Lung Adenocarcinoma: Exploring Robust Image Features with Cross-Domain Transfer Learning

PURPOSE

This study aimed to find suitable source domain data in cross-domain transfer learning to extract robust image features. Then, a model was built to preoperatively distinguish lung granulomatous nodules (LGNs) from lung adenocarcinoma (LAC) in solitary pulmonary solid nodules (SPSNs).

METHODS

Data from 841 patients with SPSNs from five centres were collected retrospectively. First, adaptive cross-domain transfer learning was used to construct transfer learning signatures (TLS) under different source domain data and conduct a comparative analysis. The Wasserstein distance was used to assess the similarity between the source domain and target domain data in cross-domain transfer learning. Second, a cross-domain transfer learning radiomics model (TLRM) combining the best performing TLS, clinical factors and subjective CT findings was constructed. Finally, the performance of the model was validated through multicentre validation cohorts.

RESULTS

Relative to other source domain data, TLS based on lung whole slide images as source domain data (TLS-LW) had the best performance in all validation cohorts (AUC range: 0.8228-0.8984). Meanwhile, the Wasserstein distance of TLS-LW was 1.7108, which was minimal. Finally, TLS-LW, age, spiculated sign and lobulated shape were used to build the TLRM. In all validation cohorts, The AUC ranges were 0.9074-0.9442. Compared with other models, decision curve analysis and integrated discrimination improvement showed that TLRM had better performance.

CONCLUSIONS

The TLRM could assist physicians in preoperatively differentiating LGN from LAC in SPSNs. Furthermore, compared with other images, cross-domain transfer learning can extract robust image features when using lung whole slide images as source domain data and has a better effect.

Data augmentation for medical imaging: A systematic literature review

Recent advances in Deep Learning have largely benefited from larger and more diverse training sets. However, collecting large datasets for medical imaging is still a challenge due to privacy concerns and labeling costs. Data augmentation makes it possible to greatly expand the amount and variety of data available for training without actually collecting new samples. Data augmentation techniques range from simple yet surprisingly effective transformations such as cropping, padding, and flipping, to complex generative models. Depending on the nature of the input and the visual task, different data augmentation strategies are likely to perform differently. For this reason, it is conceivable that medical imaging requires specific augmentation strategies that generate plausible data samples and enable effective regularization of deep neural networks. Data augmentation can also be used to augment specific classes that are underrepresented in the training set, e.g., to generate artificial lesions. The goal of this systematic literature review is to investigate which data augmentation strategies are used in the medical domain and how they affect the performance of clinical tasks such as classification, segmentation, and lesion detection. To this end, a comprehensive analysis of more than 300 articles published in recent years (2018-2022) was conducted. The results highlight the effectiveness of data augmentation across organs, modalities, tasks, and dataset sizes, and suggest potential avenues for future research.

Deep learning magnetic resonance imaging predicts platinum sensitivity in patients with epithelial ovarian cancer

OBJECTIVE

The aim of the study is to develop and validate a deep learning model to predict the platinum sensitivity of patients with epithelial ovarian cancer (EOC) based on contrast-enhanced magnetic resonance imaging (MRI).

METHODS

In this retrospective study, 93 patients with EOC who received platinum-based chemotherapy (≥4 cycles) and debulking surgery at the Sun Yat-sen Memorial Hospital from January 2011 to January 2020 were enrolled and randomly assigned to the training and validation cohorts (2:1). Two different models were built based on either the primary tumor or whole volume of the abdomen as the volume of interest (VOI) within the same cohorts, and then a pre-trained convolutional neural network Med3D (Resnet 10 version) was transferred to automatically extract 1,024 features from two MRI sequences (CE-T1WI and T2WI) of each patient to predict platinum sensitivity. The performance of the two models was compared.

RESULTS

A total of 93 women (mean age, 50.5 years ± 10.5 [standard deviation]) were evaluated (62 in the training cohort and 31 in the validation cohort). The AUCs of the whole abdomen model were 0.97 and 0.98 for the training and validation cohorts, respectively, which was better than the primary tumor model (AUCs of 0.88 and 0.81 in the training and validation cohorts, respectively). In k-fold cross-validation and stratified analysis, the whole abdomen model maintained a stable performance, and the decision function value generated by the model was a prognostic indicator that successfully discriminates high- and low-risk recurrence patients.

CONCLUSION

The non-manually segmented whole-abdomen deep learning model based on MRI exhibited satisfactory predictive performance for platinum sensitivity and may assist gynecologists in making optimal treatment decisions.

Using stacked deep learning models based on PET/CT images and clinical data to predict EGFR mutations in lung cancer

Purpose

To determine whether stacked deep learning models based on PET/CT images and clinical data can help to predict epidermal growth factor receptor (EGFR) mutations in lung cancer.

Methods

We analyzed data from two public datasets of patients who underwent ¹⁸F-FDG PET/CT. Three PET deep learning ResNet models and one CT deep learning ResNet model were trained as low-level predictors based on PET and CT images, respectively. A high-level Support Vector Machine model (Stack PET/CT and Clinical model) was trained using the prediction results of the low-level predictors and clinical data. The clinical data included sex, age, smoking history, SUVmax and SUVmean of the lesion. Fivefold cross-validation was used in this study to validate the prediction performance of the models. The predictive performance of the models was evaluated by receiver operator characteristic (ROC) curves. The area under the curve (AUC) was calculated.

Results

One hundred forty-seven patients were included in this study. Among them, 37/147 cases were EGFR mutations, and 110/147 cases were EGFR wild-type. The ROC analysis showed that the Stack PET/CT & Clinical model had the best performance (AUC = 0.85 ± 0.09), with 0.76, 0.85 and 0.83 in sensitivity, specificity and accuracy, respectively. Three ResNet PET models had relatively higher AUCs (0.82 ± 0.07, 0.80 ± 0.08 and 0.79 ± 0.07) and outperformed the CT model (AUC = 0.58 ± 0.12).

Conclusion

Using stack generalization, the deep learning model was able to efficiently combine the anatomic and biological imaging information gathered from PET/CT images with clinical data. This stacked deep learning model showed a strong ability to predict EGFR mutations with high accuracy.

A Machine Learning-Based Predictive Model of Epidermal Growth Factor Mutations in Lung Adenocarcinomas

Data from 758 patients with lung adenocarcinoma were retrospectively collected. All patients had undergone computed tomography imaging and EGFR gene testing. Radiomic features were extracted using the medical imaging tool 3D-Slicer and were combined with the clinical features to build a machine learning prediction model. The high-dimensional feature set was screened for optimal feature subsets using principal component analysis (PCA) and the least absolute shrinkage and selection operator (LASSO). Model prediction of EGFR mutation status in the validation group was evaluated using multiple classifiers. We showed that six clinical features and 622 radiomic features were initially collected. Thirty-one radiomic features with non-zero correlation coefficients were obtained by LASSO regression, and 24 features correlated with label values were obtained by PCA. The shared radiomic features determined by these two methods were selected and combined with the clinical features of the respective patient to form a subset of features related to EGFR mutations. The full dataset was partitioned into training and test sets at a ratio of 7:3 using 10-fold cross-validation. The area under the curve (AUC) of the four classifiers with cross-validations was: (1) K-nearest neighbor (AUCmean = 0.83, Acc = 81%); (2) random forest (AUCmean = 0.91, Acc = 83%); (3) LGBM (AUCmean = 0.94, Acc = 88%); and (4) support vector machine (AUCmean = 0.79, Acc = 83%). In summary, the subset of radiographic and clinical features selected by feature engineering effectively predicted the EGFR mutation status of this NSCLC patient cohort.

Deep learning analysis to predict EGFR mutation status in lung adenocarcinoma manifesting as pure ground-glass opacity nodules on CT

BACKGROUND

Epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) showed potency as a non-invasive therapeutic approach in pure ground-glass opacity nodule (pGGN) lung adenocarcinoma. However, optimal methods of extracting information about EGFR mutation from pGGN lung adenocarcinoma images remain uncertain. We aimed to develop, validate, and evaluate the clinical utility of a deep learning model for predicting EGFR mutation status in lung adenocarcinoma manifesting as pGGN on computed tomography (CT).

METHODS

We included 185 resected pGGN lung adenocarcinomas in the primary cohort. The patients were divided into training (n = 125), validation (n = 23), and test sets (n = 37). A preoperative CT-based deep learning model with clinical factors as well as clinical and radiomics models was constructed and applied to the test set. We evaluated the clinical utility of the deep learning model by applying it to 83 GGNs that received EGFR-TKI from an independent cohort (clinical validation set), and treatment response was regarded as the reference standard.

RESULTS

The prediction efficiencies of each model were compared in terms of area under the curve (AUC). Among the 185 pGGN lung adenocarcinomas, 122 (65.9%) were EGFR-mutant and 63 (34.1%) were EGFR-wild type. The AUC of the clinical, radiomics, and deep learning with clinical models to predict EGFR mutations were 0.50, 0.64, and 0.85, respectively, for the test set. The AUC of deep learning with the clinical model in the validation set was 0.72.

CONCLUSIONS

Deep learning approach of CT images combined with clinical factors can predict EGFR mutations in patients with lung adenocarcinomas manifesting as pGGN, and its clinical utility was demonstrated in a real-world sample.

Artificial intelligence in lung cancer: Application and future thinking

The availability of medical big data and the rapid development of computer software and hardware have greatly promoted the advancement of intelligent medical healthcare. Artificial intelligence (AI) has been successfully applied in many fields of medicine, especially in lung cancer. The performance of AI in some specific tasks has surpassed that of humans. Several AI software has been deeply used in clinical practice to help decision-making, which is producing a profound influence on clinicians. At present, the application of AI in the field of lung cancer mainly includes detection, segmentation, classification, prognosis prediction, efficacy evaluation, and so on. AI faces certain challenges and opportunities in the era of big data in terms of data acquisition, annotation and interpretability. Researchers have conducted deep and extensive studies using AI in the field of lung cancer, and AI is expected to become a powerful assistant in the prevention and treatment of lung cancer. AI is bringing an unprecedented revolution to radiologists, but the role of radiologists is crucial in the development of AI.

3D convolutional neural network model from contrast-enhanced CT to predict spread through air spaces in non-small cell lung cancer

PURPOSE

The purpose of this study was to compare the efficacy of five non-invasive models, including three-dimensional (3D) convolutional neural network (CNN) model, to predict the spread through air spaces (STAS) status of non-small cell lung cancer (NSCLC), and to obtain the best prediction model to provide a basis for clinical surgery planning.

MATERIALS AND METHODS

A total of 203 patients (112 men, 91 women; mean age, 60 years; age range 22-80 years) with NSCLC were retrospectively included. Of these, 153 were used for training cohort and 50 for validation cohort. According to the image biomarker standardization initiative reference manual, the image processing and feature extraction were standardized using PyRadiomics. The logistic regression classifier was used to build the model. Five models (clinicopathological/CT model, conventional radiomics model, computer vision (CV) model, 3D CNN model and combined model) were constructed to predict STAS by NSCLC. Area under the receiver operating characteristic curves (AUC) were used to validate the capability of the five models to predict STAS.

RESULTS

For predicting STAS, the 3D CNN model was superior to the clinicopathological/CT model, conventional radiomics model, CV model and combined model and achieved satisfactory discrimination performance, with an AUC of 0.93 (95% CI: 0.70-0.82) in the training cohort and 0.80 (95% CI: 0.65-0.86) in the validation cohort. Decision curve analysis indicated that, when the probability of the threshold was over 10%, the 3D CNN model was beneficial for predicting STAS status compared to either treating all or treating none of the patients within certain ranges of risk threshold CONCLUSION: The 3D CNN model can be used for the preoperative prediction of STAS in patients with NSCLC, and was superior to the other four models in predicting patients' risk of developing STAS.

The value of AI in the Diagnosis, Treatment, and Prognosis of Malignant Lung Cancer

Malignant tumors is a serious public health threat. Among them, lung cancer, which has the highest fatality rate globally, has significantly endangered human health. With the development of artificial intelligence (AI) and its integration with medicine, AI research in malignant lung tumors has become critical. This article reviews the value of CAD, computer neural network deep learning, radiomics, molecular biomarkers, and digital pathology for the diagnosis, treatment, and prognosis of malignant lung tumors.

External Validation of Deep Learning Algorithms for Radiologic Diagnosis: A Systematic Review

Purpose

To assess generalizability of published deep learning (DL) algorithms for radiologic diagnosis.

Materials and Methods

In this systematic review, the PubMed database was searched for peer-reviewed studies of DL algorithms for image-based radiologic diagnosis that included external validation, published from January 1, 2015, through April 1, 2021. Studies using nonimaging features or incorporating non-DL methods for feature extraction or classification were excluded. Two reviewers independently evaluated studies for inclusion, and any discrepancies were resolved by consensus. Internal and external performance measures and pertinent study characteristics were extracted, and relationships among these data were examined using nonparametric statistics.

Results

Eighty-three studies reporting 86 algorithms were included. The vast majority (70 of 86, 81%) reported at least some decrease in external performance compared with internal performance, with nearly half (42 of 86, 49%) reporting at least a modest decrease (≥0.05 on the unit scale) and nearly a quarter (21 of 86, 24%) reporting a substantial decrease (≥0.10 on the unit scale). No study characteristics were found to be associated with the difference between internal and external performance.

Conclusion

Among published external validation studies of DL algorithms for image-based radiologic diagnosis, the vast majority demonstrated diminished algorithm performance on the external dataset, with some reporting a substantial performance decrease.Keywords: Meta-Analysis, Computer Applications-Detection/Diagnosis, Neural Networks, Computer Applications-General (Informatics), Epidemiology, Technology Assessment, Diagnosis, Informatics Supplemental material is available for this article. © RSNA, 2022.

Development and validation of novel radiomics-based nomograms for the prediction of EGFR mutations and Ki-67 proliferation index in non-small cell lung cancer

BACKGROUND

We developed and validated novel radiomics-based nomograms to identify epidermal growth factor receptor (EGFR) mutations and the Ki-67 proliferation index of non-small cell lung cancer.

METHODS

We enrolled 132 patients with histologically verified non-small cell lung cancer from four hospital institutions who underwent computed tomography (CT) scans. EGFR mutations and the Ki-67 proliferation index were measured from tumor tissues. A total of 1,287 radiomic features were extracted, and a three-stage feature selection method was implemented to acquire the most valuable radiomic features. Finally, the radiomic scores and nomograms of the two tasks were established and tested. Receiver operating characteristic curves, calibration curves, and decision curves were used to evaluate their prediction performance and clinical utility.

RESULTS

In task [1], smoking status and histological type were significantly associated with EGFR mutations. After feature selection, 10 features were used to establish radiomic score, which showed good performance [area under the curve (AUC) =0.800] in the validation cohort. The radiomic nomogram had an AUC of 0.798 (95% CI: 0.664 to 0.931) with a C-index of 0.798 in the validation cohort. In task [2], gender, smoking status, histological type, and stage showed a significant correlation with Ki-67 proliferation index expression. A total of 28 features were selected to develop a radiomic score, with an AUC of 0.820 in the validation cohort. The final nomogram showed an AUC of 0.828 (95% CI: 0.703 to 0.953) with a C-index of 0.828 in the validation cohort.

CONCLUSIONS

EGFR mutations and Ki-67 proliferation index in non-small cell lung cancer can be predicted efficiently by the novel radiomic scores and nomograms.

[Clinical Study of Artificial Intelligence-assisted Diagnosis System in Predicting the Invasive Subtypes of Early-stage Lung Adenocarcinoma Appearing as Pulmonary Nodules]

BACKGROUND

Lung cancer is the cancer with the highest mortality at home and abroad at present. The detection of lung nodules is a key step to reducing the mortality of lung cancer. Artificial intelligence-assisted diagnosis system presents as the state of the art in the area of nodule detection, differentiation between benign and malignant and diagnosis of invasive subtypes, however, a validation with clinical data is necessary for further application. Therefore, the aim of this study is to evaluate the effectiveness of artificial intelligence-assisted diagnosis system in predicting the invasive subtypes of early‑stage lung adenocarcinoma appearing as pulmonary nodules.

METHODS

Clinical data of 223 patients with early-stage lung adenocarcinoma appearing as pulmonary nodules admitted to the Lanzhou University Second Hospital from January 1st, 2016 to December 31th, 2021 were retrospectively analyzed, which were divided into invasive adenocarcinoma group (n=170) and non-invasive adenocarcinoma group (n=53), and the non-invasive adenocarcinoma group was subdivided into minimally invasive adenocarcinoma group (n=31) and preinvasive lesions group (n=22). The malignant probability and imaging characteristics of each group were compared to analyze their predictive ability for the invasive subtypes of early-stage lung adenocarcinoma. The concordance between qualitative diagnostic results of artificial intelligence-assisted diagnosis of the invasive subtypes of early-stage lung adenocarcinoma and postoperative pathology was then analyzed.

RESULTS

In different invasive subtypes of early-stage lung adenocarcinoma, the mean CT value of pulmonary nodules (P<0.001), diameter (P<0.001), volume (P<0.001), malignant probability (P<0.001), pleural retraction sign (P<0.001), lobulation (P<0.001), spiculation (P<0.001) were significantly different. At the same time, it was also found that with the increased invasiveness of different invasive subtypes of early-stage lung adenocarcinoma, the proportion of dominant signs of each group gradually increased. On the issue of binary classification, the sensitivity, specificity, and area under the curve (AUC) values of the artificial intelligence-assisted diagnosis system for the qualitative diagnosis of invasive subtypes of early-stage lung adenocarcinoma were 81.76%, 92.45% and 0.871 respectively. On the issue of three classification, the accuracy, recall rate, F1 score, and AUC values of the artificial intelligence-assisted diagnosis system for the qualitative diagnosis of invasive subtypes of early-stage lung adenocarcinoma were 83.86%, 85.03%, 76.46% and 0.879 respectively.

CONCLUSIONS

Artificial intelligence-assisted diagnosis system could predict the invasive subtypes of early‑stage lung adenocarcinoma appearing as pulmonary nodules, and has a certain predictive value. With the optimization of algorithms and the improvement of data, it may provide guidance for individualized treatment of patients.

Towards Machine Learning-Aided Lung Cancer Clinical Routines: Approaches and Open Challenges

Advancements in the development of computer-aided decision (CAD) systems for clinical routines provide unquestionable benefits in connecting human medical expertise with machine intelligence, to achieve better quality healthcare. Considering the large number of incidences and mortality numbers associated with lung cancer, there is a need for the most accurate clinical procedures; thus, the possibility of using artificial intelligence (AI) tools for decision support is becoming a closer reality. At any stage of the lung cancer clinical pathway, specific obstacles are identified and "motivate" the application of innovative AI solutions. This work provides a comprehensive review of the most recent research dedicated toward the development of CAD tools using computed tomography images for lung cancer-related tasks. We discuss the major challenges and provide critical perspectives on future directions. Although we focus on lung cancer in this review, we also provide a more clear definition of the path used to integrate AI in healthcare, emphasizing fundamental research points that are crucial for overcoming current barriers.

Prediction of Glioma Grade Using Intratumoral and Peritumoral Radiomic Features From Multiparametric MRI Images

The accurate prediction of glioma grade before surgery is essential for treatment planning and prognosis. Since the gold standard (i.e., biopsy)for grading gliomas is both highly invasive and expensive, and there is a need for a noninvasive and accurate method. In this study, we proposed a novel radiomics-based pipeline by incorporating the intratumoral and peritumoral features extracted from preoperative mpMRI scans to accurately and noninvasively predict glioma grade. To address the unclear peritumoral boundary, we designed an algorithm to capture the peritumoral region with a specified radius. The mpMRI scans of 285 patients derived from a multi-institutional study were adopted. A total of 2153 radiomic features were calculated separately from intratumoral volumes (ITVs)and peritumoral volumes (PTVs)on mpMRI scans, and then refined using LASSO and mRMR feature ranking methods. The top-ranking radiomic features were entered into the classifiers to build radiomic signatures for predicting glioma grade. The prediction performance was evaluated with five-fold cross-validation on a patient-level split. The radiomic signatures utilizing the features of ITV and PTV both show a high accuracy in predicting glioma grade, with AUCs reaching 0.968. By incorporating the features of ITV and PTV, the AUC of IPTV radiomic signature can be increased to 0.975, which outperforms the state-of-the-art methods. Additionally, our proposed method was further demonstrated to have strong generalization performance in an external validation dataset with 65 patients. The source code of our implementation is made publicly available at https://github.com/chengjianhong/glioma_grading.git.

Three-Dimensional Convolutional Neural Network-Based Prediction of Epidermal Growth Factor Receptor Expression Status in Patients With Non-Small Cell Lung Cancer

Objectives

EGFR testing is a mandatory step before targeted therapy for non-small cell lung cancer patients. Combining some quantifiable features to establish a predictive model of EGFR expression status, break the limitations of tissue biopsy.

Materials and Methods

We retrospectively analyzed 1074 patients of non-small cell lung cancer with complete reports of EGFR gene testing. Then manually segmented VOI, captured the clinicopathological features, analyzed traditional radiology features, and extracted radiomic, and deep learning features. The cases were randomly divided into training and test set. We carried out feature screening; then applied the light GBM algorithm, Resnet-101 algorithm, logistic regression to develop sole models, and fused models to predict EGFR mutation conditions. The efficiency of models was evaluated by ROC and PRC curves.

Results

We successfully established Model_clinical, Model_radiomic, Model_CNN (based on clinical-radiology, radiomic and deep learning features respectively), Model_{radiomic+clinical} (combining clinical-radiology and radiomic features), and Model_{CNN+radiomic+clinical} (combining clinical-radiology, radiomic, and deep learning features). Among the prediction models, Model_{CNN+radiomic+clinical} showed the highest performance, followed by Model_CNN, and then Model_{radiomic+clinical}. All three models were able to accurately predict EGFR mutation with AUC values of 0.751, 0.738, and 0.684, respectively. There was no significant difference in the AUC values between Model_{CNN+radiomic+clinical} and Model_CNN. Further analysis showed that Model_{CNN+radiomic+clinical} effectively improved the efficacy of Model_{radiomic+clinical} and showed better efficacy than Model_CNN. The inclusion of clinical-radiology features did not effectively improve the efficacy of Model_radiomic.

Conclusions

Either deep learning or radiomic signature-based models can provide a fairly accurate non-invasive prediction of EGFR expression status. The model combined both features effectively enhanced the performance of radiomic models and provided marginal enhancement to deep learning models. Collectively, fusion models offer a novel and more reliable way of providing the efficacy of currently developed prediction models, and have far-reaching potential for the optimization of noninvasive EGFR mutation status prediction methods.

A narrative review of artificial intelligence-assisted histopathologic diagnosis and decision-making for non-small cell lung cancer: achievements and limitations

Objective

To summarize the current evidence regarding the applications, workflow, and limitations of artificial intelligence (AI) in the management of patients pathologically-diagnosed with lung cancer.

Background

Lung cancer is one of the most common cancers and the leading cause of cancer-related deaths worldwide. AI technologies have been applied to daily medical workflow and have achieved an excellent performance in predicting histopathologic subtypes, analyzing gene mutation profiles, and assisting in clinical decision-making for lung cancer treatment. More advanced deep learning for classifying pathologic images with minimal human interactions has been developed in addition to the conventional machine learning scheme.

Methods

Studies were identified by searching databases, including PubMed, EMBASE, Web of Science, and Cochrane Library, up to February 2021 without language restrictions.

Conclusions

A number of studies have evaluated AI pipelines and confirmed that AI is robust and efficacious in lung cancer diagnosis and decision-making, demonstrating that AI models are a useful tool for assisting oncologists in health management. Although several limitations that pose an obstacle for the widespread use of AI schemes persist, the unceasing refinement of AI techniques is poised to overcome such problems. Thus, AI technology is a promising tool for use in diagnosing and managing lung cancer.

Deep Learning to Predict EGFR Mutation and PD-L1 Expression Status in Non-Small-Cell Lung Cancer on Computed Tomography Images

OBJECTIVE

The detection of epidermal growth factor receptor (EGFR) mutation and programmed death ligand-1 (PD-L1) expression status is crucial to determine the treatment strategies for patients with non-small-cell lung cancer (NSCLC). Recently, the rapid development of radiomics including but not limited to deep learning techniques has indicated the potential role of medical images in the diagnosis and treatment of diseases.

METHODS

Eligible patients diagnosed/treated at the West China Hospital of Sichuan University from January 2013 to April 2019 were identified retrospectively. The preoperative CT images were obtained, as well as the gene status regarding EGFR mutation and PD-L1 expression. Tumor region of interest (ROI) was delineated manually by experienced respiratory specialists. We used 3D convolutional neural network (CNN) with ROI information as input to construct a classification model and established a prognostic model combining deep learning features and clinical features to stratify survival risk of lung cancer patients.

RESULTS

The whole cohort (N = 1262) was divided into a training set (N = 882, 70%), validation set (N = 125, 10%), and test set (N = 255, 20%). We used a 3D convolutional neural network (CNN) to construct a prediction model, with AUCs of 0.96 (95% CI: 0.94-0.98), 0.80 (95% CI: 0.72-0.88), and 0.73 (95% CI: 0.63-0.83) in the training, validation, and test cohorts, respectively. The combined prognostic model showed a good performance on survival prediction in NSCLC patients (C-index: 0.71).

CONCLUSION

In this study, a noninvasive and effective model was proposed to predict EGFR mutation and PD-L1 expression status as a clinical decision support tool. Additionally, the combination of deep learning features with clinical features demonstrated great stratification capabilities in the prognostic model. Our team would continue to explore the application of imaging markers for treatment selection of lung cancer patients.

AIR-Net: A novel multi-task learning method with auxiliary image reconstruction for predicting EGFR mutation status on CT images of NSCLC patients

Automated and accurate EGFR mutation status prediction using computed tomography (CT) imagery is of great value for tailoring optimal treatments to non-small cell lung cancer (NSCLC) patients. However, existing deep learning based methods usually adopt a single task learning strategy to design and train EGFR mutation status prediction models with limited training data, which may be insufficient to learn distinguishable representations for promoting prediction performance. In this paper, a novel multi-task learning method named AIR-Net is proposed to precisely predict EGFR mutation status on CT images. First, an auxiliary image reconstruction task is effectively integrated with EGFR mutation status prediction, aiming at providing extra supervision at the training phase. Particularly, we adequately employ multi-level information in a shared encoder to generate more comprehensive representations of tumors. Second, a powerful feature consistency loss is further introduced to constrain semantic consistency of original and reconstructed images, which contributes to enhanced image reconstruction and offers more effective regularization to AIR-Net during training. Performance analysis of AIR-Net indicates that auxiliary image reconstruction plays an essential role in identifying EGFR mutation status. Furthermore, extensive experimental results demonstrate that our method achieves favorable performance against other competitive prediction methods. All the results executed in this study suggest that the effectiveness and superiority of AIR-Net in precisely predicting EGFR mutation status of NSCLC.

AI in spotting high-risk characteristics of medical imaging and molecular pathology

Medical imaging provides a comprehensive perspective and rich information for disease diagnosis. Combined with artificial intelligence technology, medical imaging can be further mined for detailed pathological information. Many studies have shown that the macroscopic imaging characteristics of tumors are closely related to microscopic gene, protein and molecular changes. In order to explore the function of artificial intelligence algorithms in in-depth analysis of medical imaging information, this paper reviews the articles published in recent years from three perspectives: medical imaging analysis method, clinical applications and the development of medical imaging in the direction of pathological molecular prediction. We believe that AI-aided medical imaging analysis will be extensively contributing to precise and efficient clinical decision.

Deep learning predicts epidermal growth factor receptor mutation subtypes in lung adenocarcinoma

PURPOSE

This study aimed to explore the predictive ability of deep learning (DL) for the common epidermal growth factor receptor (EGFR) mutation subtypes in patients with lung adenocarcinoma.

METHODS

A total of 665 patients with lung adenocarcinoma (528/137) were recruited from two different institutions. In the training set, an 18-layer convolutional neural network (CNN) and fivefold cross-validation strategy were used to establish a CNN model. Subsequently, an independent external validation cohort from the other institution was used to evaluate the predictive efficacy of the CNN model. Grad-weighted class activation mapping (Grad-CAM) technology was used for the visual interpretation of the CNN model. In addition, this study also compared the prediction abilities of the radiomics and CNN models. Receiver operating characteristic (ROC) curves, accuracy and precision values, and recall and F1-score were used to evaluate the effectiveness of the CNN model and compare its performance with that of the radiomics model.

RESULTS

In the validation set, the micro- and macroaverage values of the area under the ROC curve of the CNN model to identify the three EGFR subtypes were 0.78 and 0.79, respectively. All evaluation indicators of the CNN model were better than those of the radiomics model.

CONCLUSIONS

Our study confirmed the potential of DL for predicting the EGFR mutation status in lung adenocarcinoma. The imaging phenotypes of the three mutation subtypes were found to be different, which can provide a basis for choosing more accurate and personalized treatment in patients with lung adenocarcinoma.

Machine Learning-Based Radiomics Signatures for EGFR and KRAS Mutations Prediction in Non-Small-Cell Lung Cancer

Early identification of epidermal growth factor receptor (EGFR) and Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations is crucial for selecting a therapeutic strategy for patients with non-small-cell lung cancer (NSCLC). We proposed a machine learning-based model for feature selection and prediction of EGFR and KRAS mutations in patients with NSCLC by including the least number of the most semantic radiomics features. We included a cohort of 161 patients from 211 patients with NSCLC from The Cancer Imaging Archive (TCIA) and analyzed 161 low-dose computed tomography (LDCT) images for detecting EGFR and KRAS mutations. A total of 851 radiomics features, which were classified into 9 categories, were obtained through manual segmentation and radiomics feature extraction from LDCT. We evaluated our models using a validation set consisting of 18 patients derived from the same TCIA dataset. The results showed that the genetic algorithm plus XGBoost classifier exhibited the most favorable performance, with an accuracy of 0.836 and 0.86 for detecting EGFR and KRAS mutations, respectively. We demonstrated that a noninvasive machine learning-based model including the least number of the most semantic radiomics signatures could robustly predict EGFR and KRAS mutations in patients with NSCLC.

Radiomic Feature-Based Nomogram: A Novel Technique to Predict EGFR-Activating Mutations for EGFR Tyrosin Kinase Inhibitor Therapy

Objectives

To develop and validate a radiomic feature-based nomogram for preoperative discriminating the epidermal growth factor receptor (EGFR) activating mutation from wild-type EGFR in non-small cell lung cancer (NSCLC) patients.

Material

A group of 301 NSCLC patients were retrospectively reviewed. The EGFR mutation status was determined by ARMS PCR analysis. All patients underwent nonenhanced CT before surgery. Radiomic features were extracted (GE healthcare). The maximum relevance minimum redundancy (mRMR) and LASSO, were used to select features. We incorporated the independent clinical features into the radiomic feature model and formed a joint model (i.e., the radiomic feature-based nomogram). The performance of the joint model was compared with that of the other two models.

Results

In total, 396 radiomic features were extracted. A radiomic signature model comprising 9 selected features was established for discriminating patients with EGFR-activating mutations from wild-type EGFR. The radiomic score (Radscore) in the two groups was significantly different between patients with wild-type EGFR and EGFR-activating mutations (training cohort: P<0.0001; validation cohort: P=0.0061). Five clinical features were retained and contributed as the clinical feature model. Compared to the radiomic feature model alone, the nomogram incorporating the clinical features and Radscore exhibited improved sensitivity and discrimination for predicting EGFR-activating mutations (sensitivity: training cohort: 0.84, validation cohort: 0.76; AUC: training cohort: 0.81, validation cohort: 0.75). Decision curve analysis demonstrated that the nomogram was clinically useful and surpassed traditional clinical and radiomic features.

Conclusions

The joint model showed favorable performance in the individualized, noninvasive prediction of EGFR-activating mutations in NSCLC patients.

Prediction of EGFR Mutation Status Based on 18F-FDG PET/CT Imaging Using Deep Learning-Based Model in Lung Adenocarcinoma

Objective

The purpose of this study was to develop a deep learning-based system to automatically predict epidermal growth factor receptor (EGFR) mutant lung adenocarcinoma in ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT).

Methods

Three hundred and one lung adenocarcinoma patients with EGFR mutation status were enrolled in this study. Two deep learning models (SE_CT and SE_PET) were developed with Squeeze-and-Excitation Residual Network (SE-ResNet) module for the prediction of EGFR mutation with CT and PET images, respectively. The deep learning models were trained with a training data set of 198 patients and tested with a testing data set of 103 patients. Stacked generalization was used to integrate the results of SE_CT and SE_PET.

Results

The AUCs of the SE_CT and SE_PET were 0.72 (95% CI, 0.62–0.80) and 0.74 (95% CI, 0.65–0.82) in the testing data set, respectively. After integrating SE_CT and SE_PET with stacked generalization, the AUC was further improved to 0.84 (95% CI, 0.75–0.90), significantly higher than SE_CT (p<0.05).

Conclusion

The stacking model based on ¹⁸F-FDG PET/CT images is capable to predict EGFR mutation status of patients with lung adenocarcinoma automatically and non-invasively. The proposed model in this study showed the potential to help clinicians identify suitable advanced patients with lung adenocarcinoma for EGFR‐targeted therapy.

Vertebral MRI-based radiomics model to differentiate multiple myeloma from metastases: influence of features number on logistic regression model performance

OBJECTIVES

This study aimed to use the most frequent features to establish a vertebral MRI-based radiomics model that could differentiate multiple myeloma (MM) from metastases and compare the model performance with different features number.

METHODS

We retrospectively analyzed conventional MRI (T1WI and fat-suppression T2WI) of 103 MM patients and 138 patients with metastases. The feature selection process included four steps. The first three steps defined as conventional feature selection (CFS), carried out 50 times (ten times with 5-fold cross-validation), included variance threshold, SelectKBest, and least absolute shrinkage and selection operator. The most frequent fixed features were selected for modeling during the last step. The number of events per independent variable (EPV) is the number of patients in a smaller subgroup divided by the number of radiomics features considered in developing the prediction model. The EPV values considered were 5, 10, 15, and 20. Therefore, we constructed four models using the top 16, 8, 6, and 4 most frequent features, respectively. The models constructed with features selected by CFS were also compared.

RESULTS

The AUCs of 20EPV-Model, 15EPV-Model, and CSF-Model (AUC = 0.71, 0.81, and 0.78) were poor than 10EPV-Model (AUC = 0.84, p < 0.001). The AUC of 10EPV-Model was comparable with 5EPV-Model (AUC = 0.85, p = 0.480).

CONCLUSIONS

The radiomics model constructed with an appropriate small number of the most frequent features could well distinguish metastases from MM based on conventional vertebral MRI. Based on our results, we recommend following the 10 EPV as the rule of thumb for feature selection.

KEY POINTS

• The developed radiomics model could distinguish metastases from multiple myeloma based on conventional vertebral MRI. • An accurate model based on just a handful of the most frequent features could be constructed by utilizing multiple feature reduction techniques. • An event per independent variable value of 10 is recommended as a rule of thumb for modeling feature selection.