Prediction of lung malignancy progression and survival with machine learning based on pre-treatment FDG-PET/CT

Prognostic nomogram combining 18F-FDG PET/CT radiomics and clinical data for stage III NSCLC survival prediction

The aim of this study was to establish and validate the precision of a novel radiomics approach that integrates 18Fluorine-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET)-computed tomography (CT) scan data with clinical information to improve the prognostication of survival rates in patients diagnosed with stage III Non-Small Cell Lung Cancer (NSCLC) who are not candidates for surgery. We evaluated pretreatment 18F-FDG PET-CT scans from 156 individuals diagnosed with stage III inoperable NSCLC at Shandong Cancer Hospital. These individuals were divided into two groups: a training set comprising 110 patients and an internal validation set consisting of 46 patients. By employing random forest classifier and cox proportional hazards model , we identified and utilized relevant features to create predictive models and a nomogram. The effectiveness of these models was assessed through the use of the receiver operating characteristics(ROC) curves, Kaplan-Meier (KM) curves, and the application of the nomogram. Our findings showed that the combined model, which integrates both clinical and radiomic data, outperformed those based solely on clinical or radiomic features in predicting 3-year overall survival(OS). Furthermore, calibration plots revealed a high level of agreement between predicted and actual survival times. The research successfully established a predictive radiomics model that integrates 18F-FDG PET/CT imaging with clinical indicators to enhance survival predictions for patients with stage III inoperable NSCLC.

Peri- and intra-nodular radiomic features based on 18F-FDG PET/CT to distinguish lung adenocarcinomas from pulmonary granulomas

Objective

To compare the effectiveness of radiomic features based on 18F-FDG PET/CT images within (intranodular) and around (perinodular) lung nodules/masses in distinguishing between lung adenocarcinoma and pulmonary granulomas.

Methods

For this retrospective study, 18F-FDG PET/CT images were collected for 228 patients. Patients diagnosed with lung adenocarcinoma (n = 156) or granulomas (n = 72) were randomly assigned to a training (n = 159) and validation (n = 69) groups. The volume of interest (VOI) of intranodular, perinodular (1-5 voxels, termed Lesion_margin1 to Lesion_margin5) and total area (intra- plus perinodular region, termed Lesion_total1 to Lesion_total5) on PET/CT images were delineated using PETtumor and Marge tool of segmentation editor. A total of 1,037 radiomic features were extracted separately from PET and CT images, and the optimal features were selected to develop radiomic models. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC).

Results

Good and acceptable performance was, respectively, observed in the training (AUC = 0.868, p < 0.001) and validation (AUC = 0.715, p = 0.004) sets for the intranodular radiomic model. Among the perinodular models, the Lesion_margin2 model demonstrated the highest AUC in both sets (0.883 and 0.616, p < 0.001 and p = 0.122). Similarly, in terms of total models, Lesion_total2 model was found to outperform others in the training (AUC = 0.879, p < 0.001) and validation (AUC = 0.742, p = 0.001) sets, slightly surpassing the intranodular model.

Conclusion

When intra- and perinodular radiomic features extracted from the immediate vicinity of the nodule/mass up to 2 voxels distance on 18F-FDG PET/CT imaging are combined, improved differential diagnostic performance in distinguishing between lung adenocarcinomas and granulomas is achieved compared to the intra- and perinodular radiomic features alone.

Development and validation of a machine learning-based 18F-fluorodeoxyglucose PET/CT radiomics signature for predicting gastric cancer survival

BACKGROUND

Survival prognosis of patients with gastric cancer (GC) often influences physicians' choice of their follow-up treatment. This study aimed to develop a positron emission tomography (PET)-based radiomics model combined with clinical tumor-node-metastasis (TNM) staging to predict overall survival (OS) in patients with GC.

METHODS

We reviewed the clinical information of a total of 327 patients with pathological confirmation of GC undergoing 18 F-fluorodeoxyglucose (18 F-FDG) PET scans. The patients were randomly classified into training (n = 229) and validation (n = 98) cohorts. We extracted 171 PET radiomics features from the PET images and determined the PET radiomics scores (RS) using the least absolute shrinkage and selection operator (LASSO) and random survival forest (RSF). A radiomics model, including PET RS and clinical TNM staging, was constructed to predict the OS of patients with GC. This model was evaluated for discrimination, calibration, and clinical usefulness.

RESULTS

On multivariate COX regression analysis, the difference between age, carcinoembryonic antigen (CEA), clinical TNM, and PET RS in GC patients was statistically significant (p < 0.05). A radiomics model was developed based on the results of COX regression. The model had the Harrell's concordance index (C-index) of 0.817 in the training cohort and 0.707 in the validation cohort and performed better than a single clinical model and a model with clinical features combined with clinical TNM staging. Further analyses showed higher PET RS in patients who were older (p < 0.001) and those who had elevated CEA (p < 0.001) and higher clinical TNM (p < 0.001). At different clinical TNM stages, a higher PET RS was associated with a worse survival prognosis.

CONCLUSIONS

Radiomics models based on PET RS, clinical TNM, and clinical features may provide new tools for predicting OS in patients with GC.

SH3GL2 and MMP17 as lung adenocarcinoma biomarkers: a machine-learning based approach

Objective

Using bioinformatics machine learning methods, our research aims to identify the potential key genes associated with Lung adenocarcinoma (LUAD).

Methods

We obtained two gene expression profiling microarrays (GSE68571 and GSE74706) from the public Gene Expression Omnibus (GEO) database at the National Centre for Biotechnology Information (NCBI). The purpose was to identify Differentially Expressed Genes (DEGs) between the lung adenocarcinoma group and the healthy control group. The limma R package in R was utilized for this analysis. For the differential gene diagnosis of lung adenocarcinoma, we employed the least absolute shrinkage and selection operator (LASSO) regression and SVM-RFE screening crossover. To evaluate the performance, ROC curves were plotted. We performed immuno-infiltration analysis using CIBERSORT. Finally, we validated the key genes through qRT-PCR and Western-blot verification, then downregulated MMP17 gene expression, upregulated SH3GL2 gene expression, and performed CCK8 experiments.

Results

A total of 32 Differentially Expressed Genes (DEGs) were identified. Two diagnostic marker genes, SH3GL2 and MMP17, were selected by employing LASSO and SVM-RFE machine learning methods. In Lung adenocarcinoma cells, the expression of MMP17 was observed to be elevated compared to normal lung epithelial cells in the control group (P < 0.05). In contrast, a down-regulation of SH3GL2 was found in Lung adenocarcinoma cells (P < 0.05). Finally, we downregulated MMP17 and upregulated SH3GL2 gene expression, then the CCK8 showed that the proliferation of both lung cancer cells was inhibited.

Conclusion

SH3GL2 and MMP17 are expected to be potential biomarkers for Lung adenocarcinoma.

Integrating Machine Learning in Clinical Practice for Characterizing the Malignancy of Solitary Pulmonary Nodules in PET/CT Screening

The study investigates the efficiency of integrating Machine Learning (ML) in clinical practice for diagnosing solitary pulmonary nodules' (SPN) malignancy. Patient data had been recorded in the Department of Nuclear Medicine, University Hospital of Patras, in Greece. A dataset comprising 456 SPN characteristics extracted from CT scans, the SUVmax score from the PET examination, and the ultimate outcome (benign/malignant), determined by patient follow-up or biopsy, was used to build the ML classifier. Two medical experts provided their malignancy likelihood scores, taking into account the patient's clinical condition and without prior knowledge of the true label of the SPN. Incorporating human assessments into ML model training improved diagnostic efficiency by approximately 3%, highlighting the synergistic role of human judgment alongside ML. Under the latter setup, the ML model had an accuracy score of 95.39% (CI 95%: 95.29-95.49%). While ML exhibited swings in probability scores, human readers excelled in discerning ambiguous cases. ML outperformed the best human reader in challenging instances, particularly in SPNs with ambiguous probability grades, showcasing its utility in diagnostic grey zones. The best human reader reached an accuracy of 80% in the grey zone, whilst ML exhibited 89%. The findings underline the collaborative potential of ML and human expertise in enhancing SPN characterization accuracy and confidence, especially in cases where diagnostic certainty is elusive. This study contributes to understanding how integrating ML and human judgement can optimize SPN diagnostic outcomes, ultimately advancing clinical decision-making in PET/CT screenings.

Imaging of Solid Pulmonary Nodules

Early detection with accurate classification of solid pulmonary nodules is critical in reducing lung cancer morbidity and mortality. Computed tomography (CT) remains the most widely used imaging examination for pulmonary nodule evaluation; however, other imaging modalities, such as PET/CT and MRI, are increasingly used for nodule characterization. Current advances in solid nodule imaging are largely due to developments in machine learning, including automated nodule segmentation and computer-aided detection. This review explores current multi-modality solid pulmonary nodule detection and characterization with discussion of radiomics and risk prediction models.

Deep Learning Features and Metabolic Tumor Volume Based on PET/CT to Construct Risk Stratification in Non-small Cell Lung Cancer

RATIONALE AND OBJECTIVES

To build a risk stratification by incorporating PET/CT-based deep learning features and whole-body metabolic tumor volume (MTVwb), which was to make predictions about overall survival (OS) and progression-free survival (PFS) for those with non-small cell lung cancer (NSCLC) as a complement to the TNM staging.

MATERIALS AND METHODS

The study enrolled 590 patients with NSCLC (413 for training and 177 for testing). Features were extracted by employing a convolutional neural network. The combined risk stratification (CRS) was constructed by the selected features and MTVwb, which were contrasted and integrated with TNM staging. In the testing set, those were verified.

RESULTS

Multivariate analysis revealed that CRS was an independent predictor of OS and PFS. C-indexes of the CRS demonstrated statistically significant increases in comparison to TNM staging, excepting predicting OS in the testing set (for OS, C-index=0.71 vs. 0.691 in the training set and 0.73 vs. 0.736 in the testing set; for PFS, C-index=0.702 vs. 0.686 in the training set and 0.732 vs. 0.71 in the testing set). The nomogram that combined CRS with TNM staging demonstrated the most superior model performance in the training and testing sets (C-index=0.741 and 0.771).

CONCLUSION

The addition of CRS improves TNM staging's predictive power and shows potential as a useful tool to support physicians in making treatment decisions.

Prognostication of lung adenocarcinomas using CT-based deep learning of morphological and histopathological features: a retrospective dual-institutional study

OBJECTIVES

To develop and validate CT-based deep learning (DL) models that learn morphological and histopathological features for lung adenocarcinoma prognostication, and to compare them with a previously developed DL discrete-time survival model.

METHODS

DL models were trained to simultaneously predict five morphological and histopathological features using preoperative chest CT scans from patients with resected lung adenocarcinomas. The DL score was validated in temporal and external test sets, with freedom from recurrence (FFR) and overall survival (OS) as outcomes. Discrimination was evaluated using the time-dependent area under the receiver operating characteristic curve (TD-AUC) and compared with the DL discrete-time survival model. Additionally, we performed multivariable Cox regression analysis.

RESULTS

In the temporal test set (640 patients; median age, 64 years), the TD-AUC was 0.79 for 5-year FFR and 0.73 for 5-year OS. In the external test set (846 patients; median age, 65 years), the TD-AUC was 0.71 for 5-year OS, equivalent to the pathologic stage (0.71 vs. 0.71 [p = 0.74]). The prognostic value of the DL score was independent of clinical factors (adjusted per-percentage hazard ratio for FFR (temporal test), 1.02 [95% CI: 1.01-1.03; p < 0.001]; OS (temporal test), 1.01 [95% CI: 1.002-1.02; p = 0.01]; OS (external test), 1.01 [95% CI: 1.005-1.02; p < 0.001]). Our model showed a higher TD-AUC than the DL discrete-time survival model, but without statistical significance (2.5-year OS: 0.73 vs. 0.68; p = 0.13).

CONCLUSION

The CT-based prognostic score from collective deep learning of morphological and histopathological features showed potential in predicting survival in lung adenocarcinomas.

CLINICAL RELEVANCE STATEMENT

Collective CT-based deep learning of morphological and histopathological features presents potential for enhancing lung adenocarcinoma prognostication and optimizing pre-/postoperative management.

KEY POINTS

• A CT-based prognostic model was developed using collective deep learning of morphological and histopathological features from preoperative CT scans of 3181 patients with resected lung adenocarcinoma. • The prognostic performance of the model was comparable-to-higher performance than the pathologic T category or stage. • Our approach yielded a higher discrimination performance than the direct survival prediction model, but without statistical significance (0.73 vs. 0.68; p=0.13).

A CT-based deep learning model predicts overall survival in patients with muscle invasive bladder cancer after radical cystectomy: a multicenter retrospective cohort study

BACKGROUND

Muscle invasive bladder cancer (MIBC) has a poor prognosis even after radical cystectomy (RC). Postoperative survival stratification based on radiomics and deep learning (DL) algorithms may be useful for treatment decision-making and follow-up management. This study was aimed to develop and validate a DL model based on preoperative computed tomography (CT) for predicting postcystectomy overall survival (OS) in patients with MIBC.

METHODS

MIBC patients who underwent RC were retrospectively included from four centers, and divided into the training, internal validation, and external validation sets. A DL model incorporated the convolutional block attention module (CBAM) was built for predicting OS using preoperative CT images. The authors assessed the prognostic accuracy of the DL model and compared it with classic handcrafted radiomics model and clinical model. Then, a deep learning radiomics nomogram (DLRN) was developed by combining clinicopathological factors, radiomics score (Rad-score) and deep learning score (DL-score). Model performance was assessed by C-index, KM curve, and time-dependent ROC curve.

RESULTS

A total of 405 patients with MIBC were included in this study. The DL-score achieved a much higher C-index than Rad-score and clinical model (0.690 vs. 0.652 vs. 0.618 in the internal validation set, and 0.658 vs. 0.601 vs. 0.610 in the external validation set). After adjusting for clinicopathologic variables, the DL-score was identified as a significantly independent risk factor for OS by the multivariate Cox regression analysis in all sets (all P <0.01). The DLRN further improved the performance, with a C-index of 0.713 (95% CI: 0.627-0.798) in the internal validation set and 0.685 (95% CI: 0.586-0.765) in external validation set, respectively.

CONCLUSIONS

A DL model based on preoperative CT can predict survival outcome of patients with MIBC, which may help in risk stratification and guide treatment decision-making and follow-up management.

Predicting Microwave Ablation Early Efficacy in Pulmonary Malignancies via Δ Radiomics Models

OBJECTIVE

This study aimed to explore the value of preoperative and postoperative computed tomography (CT)-based radiomic signatures and Δ radiomic signatures for evaluating the early efficacy of microwave ablation (MWA) for pulmonary malignancies.

METHODS

In total, 115 patients with pulmonary malignancies who underwent MWA treatment were categorized into response and nonresponse groups according to relevant guidelines and consensus. Quantitative image features of the largest pulmonary malignancies were extracted from CT noncontrast scan images preoperatively (time point 0, TP0) and immediately postoperatively (time point 1, TP1). Critical features were selected from TP0 and TP1 and as Δ radiomics signatures for building radiomics models. In addition, a combined radiomics model (C-RO) was developed by integrating radiomics parameters with clinical risk factors. Prediction performance was assessed using the area under the receiver operating characteristic curve (AUC) and decision curve analysis (DCA).

RESULTS

The radiomics model using Δ features outperformed the radiomics model using TP0 and TP1 features, with training and validation AUCs of 0.892, 0.808, and 0.787, and 0.705, 0.825, and 0.778, respectively. By combining the TP0, TP1, and Δ features, the logistic regression model exhibited the best performance, with training and validation AUCs of 0.945 and 0.744, respectively. The DCA confirmed the clinical utility of the Δ radiomics model.

CONCLUSIONS

A combined prediction model, including TP0, TP1, and Δ radiometric features, can be used to evaluate the early efficacy of MWA in pulmonary malignancies.

Machine Learning in Diagnosis and Prognosis of Lung Cancer by PET-CT

As a disease with high morbidity and high mortality, lung cancer has seriously harmed people's health. Therefore, early diagnosis and treatment are more important. PET/CT is usually used to obtain the early diagnosis, staging, and curative effect evaluation of tumors, especially lung cancer, due to the heterogeneity of tumors and the differences in artificial image interpretation and other reasons, it also fails to entirely reflect the real situation of tumors. Artificial intelligence (AI) has been applied to all aspects of life. Machine learning (ML) is one of the important ways to realize AI. With the help of the ML method used by PET/CT imaging technology, there are many studies in the diagnosis and treatment of lung cancer. This article summarizes the application progress of ML based on PET/CT in lung cancer, in order to better serve the clinical. In this study, we searched PubMed using machine learning, lung cancer, and PET/CT as keywords to find relevant articles in the past 5 years or more. We found that PET/CT-based ML approaches have achieved significant results in the detection, delineation, classification of pathology, molecular subtyping, staging, and response assessment with survival and prognosis of lung cancer, which can provide clinicians a powerful tool to support and assist in critical daily clinical decisions. However, ML has some shortcomings such as slightly poor repeatability and reliability.

The application of different machine learning models based on PET/CT images and EGFR in predicting brain metastasis of adenocarcinoma of the lung

OBJECTIVE

To explore the value of six machine learning models based on PET/CT radiomics combined with EGFR in predicting brain metastases of lung adenocarcinoma.

METHODS

Retrospectively collected 204 patients with lung adenocarcinoma who underwent PET/CT examination and EGFR gene detection before treatment from Cancer Hospital Affiliated to Shandong First Medical University in 2020. Using univariate analysis and multivariate logistic regression analysis to find the independent risk factors for brain metastasis. Based on PET/CT imaging combined with EGFR and PET metabolic indexes, established six machine learning models to predict brain metastases of lung adenocarcinoma. Finally, using ten-fold cross-validation to evaluate the predictive effectiveness.

RESULTS

In univariate analysis, patients with N2-3, EGFR mutation-positive, LYM%≤20, and elevated tumor markers(P<0.05) were more likely to develop brain metastases. In multivariate Logistic regression analysis, PET metabolic indices revealed that SUVmax, SUVpeak, Volume, and TLG were risk factors for lung adenocarcinoma brain metastasis(P<0.05). The SVM model was the most efficient predictor of brain metastasis with an AUC of 0.82 (PET/CT group),0.70 (CT group),0.76 (PET group).

CONCLUSIONS

Radiomics combined with EGFR machine learning model as a new method have higher accuracy than EGFR mutation alone. SVM model is the most effective method for predicting brain metastases of lung adenocarcinoma, and the prediction efficiency of PET/CT group is better than PET group and CT group.

Clinical Utility of a CT-based AI Prognostic Model for Segmentectomy in Non-Small Cell Lung Cancer

Background Currently, no tool exists for risk stratification in patients undergoing segmentectomy for non-small cell lung cancer (NSCLC). Purpose To develop and validate a deep learning (DL) prognostic model using preoperative CT scans and clinical and radiologic information for risk stratification in patients with clinical stage IA NSCLC undergoing segmentectomy. Materials and Methods In this single-center retrospective study, transfer learning of a pretrained model was performed for survival prediction in patients with clinical stage IA NSCLC who underwent lobectomy from January 2008 to March 2017. The internal set was divided into training, validation, and testing sets based on the assignments from the pretraining set. The model was tested on an independent test set of patients with clinical stage IA NSCLC who underwent segmentectomy from January 2010 to December 2017. Its prognostic performance was analyzed using the time-dependent area under the receiver operating characteristic curve (AUC), sensitivity, and specificity for freedom from recurrence (FFR) at 2 and 4 years and lung cancer-specific survival and overall survival at 4 and 6 years. The model sensitivity and specificity were compared with those of the Japan Clinical Oncology Group (JCOG) eligibility criteria for sublobar resection. Results The pretraining set included 1756 patients. Transfer learning was performed in an internal set of 730 patients (median age, 63 years [IQR, 56-70 years]; 366 male), and the segmentectomy test set included 222 patients (median age, 65 years [IQR, 58-71 years]; 114 male). The model performance for 2-year FFR was as follows: AUC, 0.86 (95% CI: 0.76, 0.96); sensitivity, 87.4% (7.17 of 8.21 patients; 95% CI: 59.4, 100); and specificity, 66.7% (136 of 204 patients; 95% CI: 60.2, 72.8). The model showed higher sensitivity for FFR than the JCOG criteria (87.4% vs 37.6% [3.08 of 8.21 patients], P = .02), with similar specificity. Conclusion The CT-based DL model identified patients at high risk among those with clinical stage IA NSCLC who underwent segmentectomy, outperforming the JCOG criteria. © RSNA, 2024 Supplemental material is available for this article.

Machine learning based on SPECT/CT to differentiate bone metastasis and benign bone lesions in lung malignancy patients

BACKGROUND

Bone metastasis is a common event in lung cancer progression. Early diagnosis of lung malignant tumor with bone metastasis is crucial for selecting effective treatment strategies. However, 14.3% of patients are still difficult to diagnose after SPECT/CT examination.

PURPOSE

Machine learning analysis of [99mTc]-methylene diphosphate (99mTc-MDP) SPECT/CT scans to distinguish bone metastases from benign bone lesions in patients with lung cancer.

METHODS

One hundred forty-one patients (69 with bone metastases and 72 with benign bone lesions) were randomly assigned to the training group or testing group in a 7:3 ratio. Lesions were manually delineated using ITK-SNAP, and 944 radiomics features were extracted from SPECT and CT images. The least absolute shrinkage and selection operator (LASSO) regression was used to select the radiomics features in the training set, and the single/bimodal radiomics models were established based on support vector machine (SVM). To further optimize the model, the best bimodal radiomics features were combined with clinical features to establish an integrated Radiomics-clinical model. The diagnostic performance of models was evaluated using receiver operating characteristic (ROC) curve and confusion matrix, and performance differences between models were evaluated using the Delong test.

RESULTS

The optimal radiomics model comprised of structural modality (CT) and metabolic modality (SPECT), with an area under curve (AUC) of 0.919 and 0.907 for the training and testing set, respectively. The integrated model, which combined SPECT, CT, and two clinical features, exhibited satisfactory differentiation in the training and testing set, with AUC of 0.939 and 0.925, respectively.

CONCLUSIONS

The machine learning can effectively differentiate between bone metastases and benign bone lesions. The Radiomics-clinical integrated model demonstrated the best performance.

Development and validation of a multimodal model in predicting severe acute pancreatitis based on radiomics and deep learning

OBJECTIVE

Aim to establish a multimodal model for predicting severe acute pancreatitis (SAP) using machine learning (ML) and deep learning (DL).

METHODS

In this multicentre retrospective study, patients diagnosed with acute pancreatitis at admission were enrolled from January 2017 to December 2021. Clinical information within 24 h and CT scans within 72 h of admission were collected. First, we trained Model α based on clinical features selected by least absolute shrinkage and selection operator analysis. Second, radiomics features were extracted from 3D-CT scans and Model β was developed on the features after dimensionality reduction using principal component analysis. Third, Model γ was trained on 2D-CT images. Lastly, a multimodal model, namely PrismSAP, was constructed based on aforementioned features in the training set. The predictive accuracy of PrismSAP was verified in the validation and internal test sets and further validated in the external test set. Model performance was evaluated using area under the curve (AUC), accuracy, sensitivity, specificity, recall, precision and F1-score.

RESULTS

A total of 1,221 eligible patients were randomly split into a training set (n = 864), a validation set (n = 209) and an internal test set (n = 148). Data of 266 patients were for external testing. In the external test set, PrismSAP performed best with the highest AUC of 0.916 (0.873-0.960) among all models [Model α: 0.709 (0.618-0.800); Model β: 0.749 (0.675-0.824); Model γ: 0.687 (0.592-0.782); MCTSI: 0.778 (0.698-0.857); RANSON: 0.642 (0.559-0.725); BISAP: 0.751 (0.668-0.833); SABP: 0.710 (0.621-0.798)].

CONCLUSION

The proposed multimodal model outperformed any single-modality models and traditional scoring systems.

Enhancing the prediction of symptomatic radiation pneumonitis for locally advanced non-small-cell lung cancer by combining 3D deep learning-derived imaging features with dose-volume metrics: a two-center study

OBJECTIVE

This study aims to examine the ability of deep learning (DL)-derived imaging features for the prediction of radiation pneumonitis (RP) in locally advanced non-small-cell lung cancer (LA-NSCLC) patients.

MATERIALS AND METHODS

The study cohort consisted of 90 patients from the Fudan University Shanghai Cancer Center and 59 patients from the Affiliated Hospital of Jiangnan University. Occurrences of RP were used as the endpoint event. A total of 512 3D DL-derived features were extracted from two regions of interest (lung-PTV and PTV-GTV) delineated on the pre-radiotherapy planning CT. Feature selection was done using LASSO regression, and the classification models were built using the multilayered perceptron method. Performances of the developed models were evaluated by receiver operating characteristic curve analysis. In addition, the developed models were supplemented with clinical variables and dose-volume metrics of relevance to search for increased predictive value.

RESULTS

The predictive model using DL features derived from lung-PTV outperformed the one based on features extracted from PTV-GTV, with AUCs of 0.921 and 0.892, respectively, in the internal test dataset. Furthermore, incorporating the dose-volume metric V30Gy into the predictive model using features from lung-PTV resulted in an improvement of AUCs from 0.835 to 0.881 for the training data and from 0.690 to 0.746 for the validation data, respectively (DeLong p < 0.05).

CONCLUSION

Imaging features extracted from pre-radiotherapy planning CT using 3D DL networks could predict radiation pneumonitis and may be of clinical value for risk stratification and toxicity management in LA-NSCLC patients.

CLINICAL RELEVANCE STATEMENT

Integrating DL-derived features with dose-volume metrics provides a promising noninvasive method to predict radiation pneumonitis in LA-NSCLC lung cancer radiotherapy, thus improving individualized treatment and patient outcomes.

Understand how machine learning impact lung cancer research from 2010 to 2021: A bibliometric analysis

Advances in lung cancer research applying machine learning (ML) technology have generated many relevant literature. However, there is absence of bibliometric analysis review that aids a comprehensive understanding of this field and its progress. Present article for the first time performed a bibliometric analysis to clarify research status and focus from 2010 to 2021. In the analysis, a total of 2,312 relevant literature were searched and retrieved from the Web of Science Core Collection database. We conducted a bibliometric analysis and further visualization. During that time, exponentially growing annual publication and our model have shown a flourishing research prospect. Annual citation reached the peak in 2017. Researchers from United States and China have produced most of the relevant literature and strongest partnership between them. Medical image analysis and Nature appeared to bring more attention to the public. The computer-aided diagnosis, precision medicine, and survival prediction were the focus of research, reflecting the development trend at that period. ML did make a big difference in lung cancer research in the past decade.

Machine Learning to Predict the Need for Postmastectomy Radiotherapy after Immediate Breast Reconstruction

Background

Post mastectomy radiotherapy (PMRT) is an independent predictor of reconstructive complications. PMRT may alter the timing and type of reconstruction recommended. This study aimed to create a machine learning model to predict the probability of requiring PMRT after immediate breast reconstruction (IBR).

Methods

In this retrospective study, breast cancer patients who underwent IBR from January 2017 to December 2020 were reviewed and data were collected on 81 preoperative characteristics. Primary outcome was recommendation for PMRT. Four algorithms were compared to maximize performance and clinical utility: logistic regression, elastic net (EN), logistic lasso, and random forest (RF). The cohort was split into a development dataset (75% of cohort for training-validation) and 25% used for the test set. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), precision-recall curves, and calibration plots.

Results

In a total of 800 patients, 325 (40.6%) patients were recommended to undergo PMRT. With the training-validation dataset (n = 600), model performance was logistic regression 0.73 AUC [95% confidence interval (CI) 0.65-0.80]; RF 0.77 AUC (95% CI, 0.74-0.81); EN 0.77 AUC (95% CI, 0.73-0.81); logistic lasso 0.76 AUC (95% CI, 0.72-0.80). Without significantly sacrificing performance, 81 predictive factors were reduced to 12 for prediction with the EN method. With the test dataset (n = 200), performance of the EN prediction model was confirmed [0.794 AUC (95% CI, 0.730-0.858)].

Conclusion

A parsimonious accurate machine learning model for predicting PMRT after IBR was developed, tested, and translated into a clinically applicable online calculator for providers and patients.

Prediction of positive pulmonary nodules based on machine learning algorithm combined with central carbon metabolism data

BACKGROUND

Lung cancer causes a huge disease burden, and early detection of positive pulmonary nodules (PPNs) as an early sign of lung cancer is extremely important for effective intervention. It is necessary to develop PPNs risk recognizer based on machine learning algorithm combined with central carbon metabolomics.

METHODS

The study included 2248 participants at high risk for lung cancer from the Ma'anshan Community Lung Cancer Screening cohort. The Least Absolute Shrinkage and Selection Operator (LASSO) was used to screen 18 central carbon-related metabolites in plasma, recursive feature elimination (RFE) was used to select all 42 features, followed by five machine learning algorithms for model development. The performance of the model was evaluated using area under the receiver operator characteristic curve (AUC), accuracy, precision, recall, and F1 scores. In addition, SHapley Additive exPlanations (SHAP) was performed to assess the interpretability of the final selected model and to gain insight into the impact of features on the predicted results.

RESULTS

Finally, the two prediction models based on the random forest (RF) algorithm performed best, with AUC values of 0.87 and 0.83, respectively, better than other models. We found that homogentisic acid, fumaric acid, maleic acid, hippuric acid, gluconic acid, and succinic acid played a significant role in both PPNs prediction model and NPNs vs PPNs model, while 2-oxadipic acid only played a role in the former model and phosphopyruvate only played a role in the NPNs vs PPNs model. This model demonstrates the potential of central carbon metabolism for PPNs risk prediction and identification.

CONCLUSION

We developed a series of predictive models for PPNs, which can help in the early detection of PPNs and thus reduce the risk of lung cancer.

PET/CT-based deep learning grading signature to optimize surgical decisions for clinical stage I invasive lung adenocarcinoma and biologic basis under its prediction: a multicenter study

PURPOSE

No consensus on a grading system for invasive lung adenocarcinoma had been built over a long period of time. Until October 2020, a novel grading system was proposed to quantify the whole landscape of histologic subtypes and proportions of pulmonary adenocarcinomas. This study aims to develop a deep learning grading signature (DLGS) based on positron emission tomography/computed tomography (PET/CT) to personalize surgical treatments for clinical stage I invasive lung adenocarcinoma and explore the biologic basis under its prediction.

METHODS

A total of 2638 patients with clinical stage I invasive lung adenocarcinoma from 4 medical centers were retrospectively included to construct and validate the DLGS. The predictive performance of the DLGS was evaluated by the area under the receiver operating characteristic curve (AUC), its potential to optimize surgical treatments was investigated via survival analyses in risk groups defined by the DLGS, and its biological basis was explored by comparing histologic patterns, genotypic alternations, genetic pathways, and infiltration of immune cells in microenvironments between risk groups.

RESULTS

The DLGS to predict grade 3 achieved AUCs of 0.862, 0.844, and 0.851 in the validation set (n = 497), external cohort (n = 382), and prospective cohort (n = 600), respectively, which were significantly better than 0.814, 0.810, and 0.806 of the PET model, 0.813, 0.795, and 0.824 of the CT model, and 0.762, 0.734, and 0.751 of the clinical model. Additionally, for DLGS-defined high-risk population, lobectomy yielded an improved prognosis compared to sublobectomy p = 0.085 for overall survival [OS] and p = 0.038 for recurrence-free survival [RFS]) and systematic nodal dissection conferred a superior prognosis to limited nodal dissection (p = 0.001 for OS and p = 0.041 for RFS).

CONCLUSION

The DLGS harbors the potential to predict the histologic grade and personalize the surgical treatments for clinical stage I invasive lung adenocarcinoma. Its applicability to other territories should be further validated by a larger international study.

Clinical application of 18F-fluorodeoxyglucose positron emission tomography/computed tomography radiomics-based machine learning analyses in the field of oncology

Machine learning (ML) analyses using 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET)/computed tomography (CT) radiomics features have been applied in the field of oncology. The current review aimed to summarize the current clinical articles about 18F-FDG PET/CT radiomics-based ML analyses to solve issues in classifying or constructing prediction models for several types of tumors. In these studies, lung and mediastinal tumors were the most commonly evaluated lesions, followed by lymphatic, abdominal, head and neck, breast, gynecological, and other types of tumors. Previous studies have commonly shown that 18F-FDG PET radiomics-based ML analysis has good performance in differentiating benign from malignant tumors, predicting tumor characteristics and stage, therapeutic response, and prognosis by examining significant differences in the area under the receiver operating characteristic curves, accuracies, or concordance indices (> 0.70). However, these studies have reported several ML algorithms. Moreover, different ML models have been applied for the same purpose. Thus, various procedures were used in 18F-FDG PET/CT radiomics-based ML analysis in oncology, and 18F-FDG PET/CT radiomics-based ML models, which are easy and universally applied in clinical practice, would be expected to be established.

Establishment of a prognosis prediction model for lung squamous cell carcinoma related to PET/CT: basing on immunogenic cell death-related lncRNA

BACKGROUND

Immunogenic cell death (ICD) stimulates adaptive immunity and holds significant promise in cancer therapy. Nevertheless, the influence of ICD-associated long non-coding RNAs (lncRNAs) on the prognosis of patients with lung squamous cell carcinoma (LUSC) remains unexplored.

METHODS

We employed data from the The Cancer Genome Atlas (TCGA)database to identify ICD-related lncRNAs associated with the prognosis of LUSC using univariate Cox regression analysis. Subsequently, we utilized the LOSS regression model to construct a predictive risk model for assessing the prognosis of LUSC patients based on ICD-related lncRNAs. Our study randomly allocated187 TCGA patients into a training group and 184 patients for testing the predictive model. Furthermore, we conducted quantitative polymerase chain reaction (qPCR) analysis on 43 tumor tissues from LUSC patients to evaluate lncRNA expression levelsPearson correlation analysis was utilized to analyze the correlation of risk scores with positron emission tomography/computed tomography (PET/CT) parameters among LUSC patients.

RESULTS

The findings from the univariate Cox regression revealed 16 ICD-associated lncRNAs linked to LUSC prognosis, with 12 of these lncRNAs integrated into our risk model utilizing the LOSS regression. Survival analysis indicated a markedly higher overall survival time among patients in the low-risk group compared to those in the high-risk group. The area under the Receiver operating characteristic (ROC) curve to differentiate high-risk and low-risk patients was 0.688. Additionally, the overall survival rate was superior in the low-risk group compared to the high-risk group. Correlation analysis demonstrated a positive association between the risk score calculated based on the ICD-lncRNA risk model and the maximum standard uptake value (SUVmax) (r = 0.427, P = 0.0043) as well as metabolic volume (MTV)of PET-CT (r = 0.360, P = 0.0177) in 43 LUSC patients.

CONCLUSION

We have successfully developed a risk model founded on ICD-related lncRNAs that proves effective in predicting the overall survival of LUSC patients.

Methodological evaluation of original articles on radiomics and machine learning for outcome prediction based on positron emission tomography (PET)

AIM

Despite a vast number of articles on radiomics and machine learning in positron emission tomography (PET) imaging, clinical applicability remains limited, partly owing to poor methodological quality. We therefore systematically investigated the methodology described in publications on radiomics and machine learning for PET-based outcome prediction.

METHODS

A systematic search for original articles was run on PubMed. All articles were rated according to 17 criteria proposed by the authors. Criteria with >2 rating categories were binarized into "adequate" or "inadequate". The association between the number of "adequate" criteria per article and the date of publication was examined.

RESULTS

One hundred articles were identified (published between 07/2017 and 09/2023). The median proportion of articles per criterion that were rated "adequate" was 65% (range: 23-98%). Nineteen articles (19%) mentioned neither a test cohort nor cross-validation to separate training from testing. The median number of criteria with an "adequate" rating per article was 12.5 out of 17 (range, 4-17), and this did not increase with later dates of publication (Spearman's rho, 0.094; p = 0.35). In 22 articles (22%), less than half of the items were rated "adequate". Only 8% of articles published the source code, and 10% made the dataset openly available.

CONCLUSION

Among the articles investigated, methodological weaknesses have been identified, and the degree of compliance with recommendations on methodological quality and reporting shows potential for improvement. Better adherence to established guidelines could increase the clinical significance of radiomics and machine learning for PET-based outcome prediction and finally lead to the widespread use in routine clinical practice.

The use of artificial intelligence tools in cancer detection compared to the traditional diagnostic imaging methods: An overview of the systematic reviews

BACKGROUND AND PURPOSE

In comparison to conventional medical imaging diagnostic modalities, the aim of this overview article is to analyze the accuracy of the application of Artificial Intelligence (AI) techniques in the identification and diagnosis of malignant tumors in adult patients.

DATA SOURCES

The acronym PIRDs was used and a comprehensive literature search was conducted on PubMed, Cochrane, Scopus, Web of Science, LILACS, Embase, Scielo, EBSCOhost, and grey literature through Proquest, Google Scholar, and JSTOR for systematic reviews of AI as a diagnostic model and/or detection tool for any cancer type in adult patients, compared to the traditional diagnostic radiographic imaging model. There were no limits on publishing status, publication time, or language. For study selection and risk of bias evaluation, pairs of reviewers worked separately.

RESULTS

In total, 382 records were retrieved in the databases, 364 after removing duplicates, 32 satisfied the full-text reading criterion, and 09 papers were considered for qualitative synthesis. Although there was heterogeneity in terms of methodological aspects, patient differences, and techniques used, the studies found that several AI approaches are promising in terms of specificity, sensitivity, and diagnostic accuracy in the detection and diagnosis of malignant tumors. When compared to other machine learning algorithms, the Super Vector Machine method performed better in cancer detection and diagnosis. Computer-assisted detection (CAD) has shown promising in terms of aiding cancer detection, when compared to the traditional method of diagnosis.

CONCLUSIONS

The detection and diagnosis of malignant tumors with the help of AI seems to be feasible and accurate with the use of different technologies, such as CAD systems, deep and machine learning algorithms and radiomic analysis when compared with the traditional model, although these technologies are not capable of to replace the professional radiologist in the analysis of medical images. Although there are limitations regarding the generalization for all types of cancer, these AI tools might aid professionals, serving as an auxiliary and teaching tool, especially for less trained professionals. Therefore, further longitudinal studies with a longer follow-up duration are required for a better understanding of the clinical application of these artificial intelligence systems.

TRIAL REGISTRATION

Systematic review registration. Prospero registration number: CRD42022307403.

Radiotherapy modification based on artificial intelligence and radiomics applied to (18F)-fluorodeoxyglucose positron emission tomography/computed tomography

Over the last decades, the refinement of radiation therapy techniques has been associated with an increasing interest for individualized radiation therapy with the aim of increasing or maintaining tumor control and reducing radiation toxicity. Developments in artificial intelligence (AI), particularly machine learning and deep learning, in imaging sciences, including nuclear medecine, have led to significant enthusiasm for the concept of "rapid learning health system". AI combined with radiomics applied to (18F)-fluorodeoxyglucose positron emission tomography/computed tomography ([18F]-FDG PET/CT) offers a unique opportunity for the development of predictive models that can help stratify each patient's risk and guide treatment decisions for optimal outcomes and quality of life of patients treated with radiation therapy. Here we present an overview of the current contribution of AI and radiomics-based machine learning models applied to (18F)-FDG PET/CT in the management of cancer treated by radiation therapy.

Novel tools for early diagnosis and precision treatment based on artificial intelligence

Lung cancer has the highest mortality rate among all cancers in the world. Hence, early diagnosis and personalized treatment plans are crucial to improving its 5-year survival rate. Chest computed tomography (CT) serves as an essential tool for lung cancer screening, and pathology images are the gold standard for lung cancer diagnosis. However, medical image evaluation relies on manual labor and suffers from missed diagnosis or misdiagnosis, and physician heterogeneity. The rapid development of artificial intelligence (AI) has brought a whole novel opportunity for medical task processing, demonstrating the potential for clinical application in lung cancer diagnosis and treatment. AI technologies, including machine learning and deep learning, have been deployed extensively for lung nodule detection, benign and malignant classification, and subtype identification based on CT images. Furthermore, AI plays a role in the non-invasive prediction of genetic mutations and molecular status to provide the optimal treatment regimen, and applies to the assessment of therapeutic efficacy and prognosis of lung cancer patients, enabling precision medicine to become a reality. Meanwhile, histology-based AI models assist pathologists in typing, molecular characterization, and prognosis prediction to enhance the efficiency of diagnosis and treatment. However, the leap to extensive clinical application still faces various challenges, such as data sharing, standardized label acquisition, clinical application regulation, and multimodal integration. Nevertheless, AI holds promising potential in the field of lung cancer to improve cancer care.

Identification of the molecular subtypes and construction of risk models in neuroblastoma

The heterogeneity of neuroblastoma directly affects the prognosis of patients. Individualization of patient treatment to improve prognosis is a clinical challenge at this stage and the aim of this study is to characterize different patient populations. To achieve this, immune-related cell cycle genes, identified in the GSE45547 dataset using WGCNA, were used to classify cases from multiple datasets (GSE45547, GSE49710, GSE73517, GES120559, E-MTAB-8248, and TARGET) into subgroups by consensus clustering. ESTIMATES, CIBERSORT and ssGSEA were used to assess the immune status of the patients. And a 7-gene risk model was constructed based on differentially expressed genes between subtypes using randomForestSRC and LASSO. Enrichment analysis was used to demonstrate the biological characteristics between different groups. Key genes were screened using randomForest to construct neural network and validated. Finally, drug sensitivity was assessed in the GSCA and CellMiner databases. We classified the 1811 patients into two subtypes based on immune-related cell cycle genes. The two subtypes (Cluster1 and Cluster2) exhibited distinct clinical features, immune levels, chromosomal instability and prognosis. The same significant differences were demonstrated between the high-risk and low-risk groups. Through our analysis, we identified neuroblastoma subtypes with unique characteristics and established risk models which will improve our understanding of neuroblastoma heterogeneity.

The Efficacy of Pretreatment 18F-FDG PET-CT-Based Deep Learning Network Structure to Predict Survival in Nasopharyngeal Carcinoma

Background

Previous studies have shown that the 5-year survival rates of patients with nasopharyngeal carcinoma (NPC) were still not ideal despite great improvement in NPC treatments. To achieve individualized treatment of NPC, we have been looking for novel models to predict the prognosis of patients with NPC. The objective of this study was to use a novel deep learning network structural model to predict the prognosis of patients with NPC and to compare it with the traditional PET-CT model combining metabolic parameters and clinical factors.

Methods

A total of 173 patients were admitted to 2 institutions between July 2014 and April 2020 for the retrospective study; each received a PET-CT scan before treatment. The least absolute shrinkage and selection operator (LASSO) was employed to select some features, including SUVpeak-P, T3, age, stage II, MTV-P, N1, stage III and pathological type, which were associated with overall survival (OS) of patients. We constructed 2 survival prediction models: an improved optimized adaptive multimodal task (a 3D Coordinate Attention Convolutional Autoencoder and an uncertainty-based jointly Optimizing Cox Model, CACA-UOCM for short) and a clinical model. The predictive power of these models was assessed using the Harrell Consistency Index (C index). Overall survival of patients with NPC was compared by Kaplan-Meier and Log-rank tests.

Results

The results showed that CACA-UOCM model could estimate OS (C index, 0.779 for training, 0.774 for validation, and 0.819 for testing) and divide patients into low and high mortality risk groups, which were significantly associated with OS (P < .001). However, the C-index of the model based only on clinical variables was only 0.42.

Conclusions

The deep learning network model based on ¹⁸F-FDG PET/CT can serve as a reliable and powerful predictive tool for NPC and provide therapeutic strategies for individual treatment.

Optimal 18F-FDG PET/CT radiomics model development for predicting EGFR mutation status and prognosis in lung adenocarcinoma: a multicentric study

Purpose

To develop and interpret optimal predictive models to identify epidermal growth factor receptor (EGFR) mutation status and subtypes in patients with lung adenocarcinoma based on multicentric ¹⁸F-FDG PET/CT data, and further construct a prognostic model to predict their clinical outcome.

Methods

The ¹⁸F-FDG PET/CT imaging and clinical characters of 767 patients with lung adenocarcinoma from 4 cohorts were collected. Seventy-six radiomics candidates using cross-combination method to identity EGFR mutation status and subtypes were built. Further, Shapley additive explanations and local interpretable model-agnostic explanations were used for optimal models' interpretation. Moreover, in order to predict the overall survival, a multivariate Cox proportional hazard model based on handcrafted radiomics features and clinical characteristics was constructed. The predictive performance and clinical net benefit of the models were evaluated via area under receiver operating characteristic (AUC), C-index and decision curve analysis.

Results

Among the 76 radiomics candidates, light gradient boosting machine classifier (LGBM) combined with recursive feature elimination wrapped LGBM feature selection method achieved best performance in predicting EGFR mutation status (AUC reached 0.80, 0.61, 0.71 in the internal test cohort and two external test cohorts, respectively). And extreme gradient boosting classifier combined with support vector machine feature selection method achieved best performance in predicting EGFR subtypes (AUC reached 0.76, 0.63, 0.61 in the internal test cohort and two external test cohorts, respectively). The C-index of the Cox proportional hazard model achieved 0.863.

Conclusions

The integration of cross-combination method and the external validation from multi-center data achieved a good prediction and generalization performance in predicting EGFR mutation status and its subtypes. The combination of handcrafted radiomics features and clinical factors achieved good performance in predicting prognosis. With the urgent needs of multicentric ¹⁸F-FDG PET/CT trails, robust and explainable radiomics models have great potential in decision making and prognosis prediction of lung adenocarcinoma.

The role of artificial intelligence based on PET/CT radiomics in NSCLC: Disease management, opportunities, and challenges

Objectives

Lung cancer has been widely characterized through radiomics and artificial intelligence (AI). This review aims to summarize the published studies of AI based on positron emission tomography/computed tomography (PET/CT) radiomics in non-small-cell lung cancer (NSCLC).

Materials and methods

A comprehensive search of literature published between 2012 and 2022 was conducted on the PubMed database. There were no language or publication status restrictions on the search. About 127 articles in the search results were screened and gradually excluded according to the exclusion criteria. Finally, this review included 39 articles for analysis.

Results

Classification is conducted according to purposes and several studies were identified at each stage of disease:1) Cancer detection (n=8), 2) histology and stage of cancer (n=11), 3) metastases (n=6), 4) genotype (n=6), 5) treatment outcome and survival (n=8). There is a wide range of heterogeneity among studies due to differences in patient sources, evaluation criteria and workflow of radiomics. On the whole, most models show diagnostic performance comparable to or even better than experts, and the common problems are repeatability and clinical transformability.

Conclusion

AI-based PET/CT Radiomics play potential roles in NSCLC clinical management. However, there is still a long way to go before being translated into clinical application. Large-scale, multi-center, prospective research is the direction of future efforts, while we need to face the risk of repeatability of radiomics features and the limitation of access to large databases.