Machine learning-based diagnostic method of pre-therapeutic 18F-FDG PET/CT for evaluating mediastinal lymph nodes in non-small cell lung cancer

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.

Construction of a predictive model for bone metastasis from first primary lung adenocarcinoma within 3 cm based on machine learning algorithm: a retrospective study

Background

Adenocarcinoma, the most prevalent histological subtype of non-small cell lung cancer, is associated with a significantly higher likelihood of bone metastasis compared to other subtypes. The presence of bone metastasis has a profound adverse impact on patient prognosis. However, to date, there is a lack of accurate bone metastasis prediction models. As a result, this study aims to employ machine learning algorithms for predicting the risk of bone metastasis in patients.

Method

We collected a dataset comprising 19,454 cases of solitary, primary lung adenocarcinoma with pulmonary nodules measuring less than 3 cm. These cases were diagnosed between 2010 and 2015 and were sourced from the Surveillance, Epidemiology, and End Results (SEER) database. Utilizing clinical feature indicators, we developed predictive models using seven machine learning algorithms, namely extreme gradient boosting (XGBoost), logistic regression (LR), light gradient boosting machine (LightGBM), Adaptive Boosting (AdaBoost), Gaussian Naive Bayes (GNB), multilayer perceptron (MLP) and support vector machine (SVM).

Results

The results demonstrated that XGBoost exhibited superior performance among the four algorithms (training set: AUC: 0.913; test set: AUC: 0.853). Furthermore, for convenient application, we created an online scoring system accessible at the following URL: https://www.xsmartanalysis.com/model/predict/?mid=731symbol=7Fr16wX56AR9Mk233917, which is based on the highest performing model.

Conclusion

XGBoost proves to be an effective algorithm for predicting the occurrence of bone metastasis in patients with solitary, primary lung adenocarcinoma featuring pulmonary nodules below 3 cm in size. Moreover, its robust clinical applicability enhances its potential utility.

LNAS: a clinically applicable deep-learning system for mediastinal enlarged lymph nodes segmentation and station mapping without regard to the pathogenesis using unenhanced CT images

BACKGROUND

The accurate identification and evaluation of lymph nodes by CT images is of great significance for disease diagnosis, treatment, and prognosis.

PURPOSE

To assess the lymph nodes' segmentation, size, and station by artificial intelligence (AI) for unenhanced chest CT images and evaluate its value in clinical scenarios.

MATERIAL AND METHODS

This retrospective study proposed an end-to-end Lymph Nodes Analysis System (LNAS) consisting of three models: the Lymph Node Segmentation model (LNS), the Mediastinal Organ Segmentation model (MOS), and the Lymph Node Station Registration model (LNR). We selected a healthy chest CT image as the template image and annotated 14 lymph node station masks according to the IASLC to build the lymph node station mapping template. The exact contours and stations of the lymph nodes were annotated by two junior radiologists and reviewed by a senior radiologist. Patients aged 18 and above, who had undergone unenhanced chest CT and had at least one suspicious enlarged mediastinal lymph node in imaging reports, were included. Exclusions were patients who had thoracic surgeries in the past 2 weeks or artifacts on CT images affecting lymph node observation by radiologists. The system was trained on 6725 consecutive chest CTs that from Tianjin Medical University General Hospital, among which 6249 patients had suspicious enlarged mediastinal lymph nodes. A total of 519 consecutive chest CTs from Qilu Hospital of Shandong University (Qingdao) were used for external validation. The gold standard for each CT was determined by two radiologists and reviewed by one senior radiologist.

RESULTS

The patient-level sensitivity of the LNAS system reached of 93.94% and 92.89% in internal and external test dataset, respectively. And the lesion-level sensitivity (recall) reached 89.48% and 85.97% in internal and external test dataset. For man-machine comparison, AI significantly apparently shortened the average reading time (p < 0.001) and had better lesion-level and patient-level sensitivities.

CONCLUSION

AI improved the sensitivity lymph node segmentation by radiologists with an advantage in reading time.

A machine learning tool to improve prediction of mediastinal lymph node metastases in non-small cell lung cancer using routinely obtainable [18F]FDG-PET/CT parameters

BACKGROUND

In patients with non-small cell lung cancer (NSCLC), accuracy of [18F]FDG-PET/CT for pretherapeutic lymph node (LN) staging is limited by false positive findings. Our aim was to evaluate machine learning with routinely obtainable variables to improve accuracy over standard visual image assessment.

METHODS

Monocentric retrospective analysis of pretherapeutic [18F]FDG-PET/CT in 491 consecutive patients with NSCLC using an analog PET/CT scanner (training + test cohort, n = 385) or digital scanner (validation, n = 106). Forty clinical variables, tumor characteristics, and image variables (e.g., primary tumor and LN SUVmax and size) were collected. Different combinations of machine learning methods for feature selection and classification of N0/1 vs. N2/3 disease were compared. Ten-fold nested cross-validation was used to derive the mean area under the ROC curve of the ten test folds ("test AUC") and AUC in the validation cohort. Reference standard was the final N stage from interdisciplinary consensus (histological results for N2/3 LNs in 96%).

RESULTS

N2/3 disease was present in 190 patients (39%; training + test, 37%; validation, 46%; p = 0.09). A gradient boosting classifier (GBM) with 10 features was selected as the final model based on test AUC of 0.91 (95% confidence interval, 0.87-0.94). Validation AUC was 0.94 (0.89-0.98). At a target sensitivity of approx. 90%, test/validation accuracy of the GBM was 0.78/0.87. This was significantly higher than the accuracy based on "mediastinal LN uptake > mediastinum" (0.7/0.75; each p < 0.05) or combined PET/CT criteria (PET positive and/or LN short axis diameter > 10 mm; 0.68/0.75; each p < 0.001). Harmonization of PET images between the two scanners affected SUVmax and visual assessment of the LNs but did not diminish the AUC of the GBM.

CONCLUSIONS

A machine learning model based on routinely available variables from [18F]FDG-PET/CT improved accuracy in mediastinal LN staging compared to established visual assessment criteria. A web application implementing this model was made available.

A machine-learning-based combination of criteria to detect bladder cancer lymph node metastasis on [18F]FDG PET/CT: a pathology-controlled study

OBJECTIVES

Initial pelvic lymph node (LN) staging is pivotal for treatment planification in patients with muscle-invasive bladder cancer (MIBC), but [¹⁸F]FDG PET/CT provides insufficient and variable diagnostic performance. We aimed to develop and validate a machine-learning-based combination of criteria on [¹⁸F]FDG PET/CT to accurately identify pelvic LN involvement in bladder cancer patients.

METHODS

Consecutive patients with localized MIBC who performed preoperative [¹⁸F]FDG PET/CT between 2010 and 2017 were retrospectively assigned to training (n = 129) and validation (n = 44) sets. The reference standard was the pathological status after extended pelvic LN dissection. In the training set, a random forest algorithm identified the combination of criteria that best predicted LN status. The diagnostic performances (AUC) and interrater agreement of this combination of criteria were compared to a consensus of experts.

RESULTS

The overall prevalence of pelvic LN involvement was 24% (n = 41/173). In the training set, the top 3 features were derived from pelvic LNs (SUVmax of the most intense LN, and product of diameters of the largest LN) and primary bladder tumor (product of diameters). In the validation set, diagnostic performance did not differ significantly between the combination of criteria (AUC = 0.59 95%CI [0.43-0.73]) and the consensus of experts (AUC = 0.64 95%CI [0.48-0.78], p = 0.54). The interrater agreement was equally good with Κ = 0.66 for both.

CONCLUSION

The developed machine-learning-based combination of criteria performs as well as a consensus of experts to detect pelvic LN involvement on [¹⁸F]FDG PET/CT in patients with MIBC.

KEY POINTS

• The developed machine-learning-based combination of criteria performs as well as experts to detect pelvic LN involvement on [¹⁸F]FDG PET/CT in patients with muscle-invasive bladder cancer. • The top 3 features to predict LN involvement were the SUVmax of the most intense LN, the product of diameters of the largest LN, and the product of diameters of the primary bladder tumor.

Radiomics Analysis of 18F-FDG PET/CT for Prognosis Prediction in Patients with Stage III Non-Small Cell Lung Cancer Undergoing Neoadjuvant Chemoradiation Therapy Followed by Surgery

We investigated the prognostic significance of radiomic features from 18F-FDG PET/CT to predict overall survival (OS) in patients with stage III NSCLC undergoing neoadjuvant chemoradiation therapy followed by surgery. We enrolled 300 patients with stage III NSCLC who underwent PET/CT at the initial work-up (PET1) and after neoadjuvant concurrent chemoradiotherapy (PET2). Radiomic primary tumor features were subjected to LASSO regression to select the most useful prognostic features of OS. The prognostic significance of the LASSO score and conventional PET parameters was assessed by Cox proportional hazards regression analysis. In conventional PET parameters, metabolic tumor volume (MTV) and total lesion glycolysis (TLG) of each PET1 and PET2 were significantly associated with OS. In addition, both the PET1-LASSO score and the PET2-LASSO score were significantly associated with OS. In multivariate Cox regression analysis, only the PET2-LASSO score was an independently significant factor for OS. The LASSO score showed better predictive performance for OS regarding the time-dependent receiver operating characteristic curve and decision curve analysis than conventional PET parameters. Radiomic features from PET/CT were an independent prognostic factor for the estimation of OS in stage III NSCLC. The newly developed LASSO score using radiomic features showed better prognostic results for individualized OS estimation than conventional PET parameters.

Predicting N2 lymph node metastasis in presurgical stage I-II non-small cell lung cancer using multiview radiomics and deep learning method

BACKGROUND

Accurate diagnosis of N2 lymph node status of the resectable stage I-II non-small cell lung cancer (NSCLC) before surgery is crucial, while there is lack of corresponding method clinically.

PURPOSE

To develop and validate a model to quantitively predict the N2 lymph node metastasis in presurgical clinical stage I-II NSCLC using multiview radiomics and deep learning method.

METHODS

In this study, 140 NSCLC patients were enrolled and randomly divided into training and test sets. Univariate and multiple analysis method were used step by step to establish the clinical model; Then a multiview radiomics modeling scheme was designed, in which the optimal input feature set was determined by subcategorizing radiomics features (C1: original; C2: LoG and C3: wavelet) and comparison of corresponding radiomics model. The minimum-redundancy maximum-relevance (mRMR) selection and the least absolute shrinkage and selection operator (LASSO) algorithm were used for the feature selection and construction of each radiomics model (Rad). Next, an end-to-end ResNet18 architecture and transfer learning techniques were designed to construct a deep learning model (DL). Subsequently, the screened clinical risk factors and constructed Rad and DL models were combined and compared and a nomogram was constructed. Finally, the diagnostic performance of all constructed models were evaluated and compared using receiver operating characteristic curve (ROC) analysis, Delong test, Calibration analysis, Hosmer-Lemeshow test, and decision curves, respectively.

RESULTS

Carcinoma embryonic antigen (CEA) level and spiculation were screened to make up the Clinical model, while seven radiomics features in the optimal input feature set C2 + C3 were selected to construct the Rad. DL was constructed by training on 1.8 million natural images and small sample data of our N2 lymph node volume of interest (VOI) images. Except for the Clinical model, all other models showed good predictive accuracy and consistency in both training set and test set. DL (area under curve (AUC): 0.83) was better than Rad (AUC: 0.76) in predictive accuracy, but their difference was not significant (p = 0.45). The combined models showed better diagnostic performance than the model only clinical or image risk factors were used (AUC for Clinical, Rad + DL, Rad + Clinical, DL + Clinical, and Rad + DL + Clinical were respectively 0.66, 0.86, 0.82, 0.86, and 0.88). Finally, the Rad + DL + Clinical model with the best diagnostic performance was selected to draw the final nomogram for clinical use.

CONCLUSION

This study proposes a nomogram based on multiview radiomics, deep learning, and clinical features that can be efficiently used to quantitively predict presurgical N2 diseases in patients with clinical stage I-II NSCLC.

Prediction of malignant lymph nodes in NSCLC by machine-learning classifiers using EBUS-TBNA and PET/CT

Accurate determination of lymph-node (LN) metastases is a prerequisite for high precision radiotherapy. The primary aim is to characterise the performance of PET/CT-based machine-learning classifiers to predict LN-involvement by endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in stage-III NSCLC. Prediction models for LN-positivity based on [¹⁸F]FDG-PET/CT features were built using logistic regression and machine-learning models random forest (RF) and multilayer perceptron neural network (MLP) for stage-III NSCLC before radiochemotherapy. A total of 675 LN-stations were sampled in 180 patients. The logistic and RF models identified SUV_max, the short-axis LN-diameter and the echelon of the considered LN among the most important parameters for EBUS-positivity. Adjusting the sensitivity of machine-learning classifiers to that of the expert-rater of 94.5%, MLP (P = 0.0061) and RF models (P = 0.038) showed lower misclassification rates (MCR) than the standard-report, weighting false positives and false negatives equally. Increasing the sensitivity of classifiers from 94.5 to 99.3% resulted in increase of MCR from 13.3/14.5 to 29.8/34.2% for MLP/RF, respectively. PET/CT-based machine-learning classifiers can achieve a high sensitivity (94.5%) to detect EBUS-positive LNs at a low misclassification rate. As the specificity decreases rapidly above that level, a combined test of a PET/CT-based MLP/RF classifier and EBUS-TBNA is recommended for radiation target volume definition.

Machine learning classification of mediastinal lymph node metastasis in NSCLC: a multicentre study in a Western European patient population

BACKGROUND

[¹⁸F] FDG PET-CT has an important role in the initial staging of lung cancer; however, accurate differentiation between activity in malignant and benign intrathoracic lymph nodes on PET-CT scans can be challenging. The purpose of the current study was to investigate the effect of incorporating primary tumour data and clinical features to differentiate between [¹⁸F] FDG-avid malignant and benign intrathoracic lymph nodes.

METHODS

We retrospectively selected lung cancer patients who underwent PET-CT for initial staging in two centres in the Netherlands. The primary tumour and suspected lymph node metastases were annotated and cross-referenced with pathology results. Lymph nodes were classified as malignant or benign. From the image data, we extracted radiomic features and trained the classifier model using the extreme gradient boost (XGB) algorithm. Various scenarios were defined by selecting different combinations of data input and clinical features. Data from centre 1 were used for training and validation of the models using the XGB algorithm. To determine the performance of the model in a different hospital, the XGB model was tested using data from centre 2.

RESULTS

Adding primary tumour data resulted in a significant gain in the performance of the trained classifier model. Adding the clinical information about distant metastases did not lead to significant improvement. The performance of the model in the test set (centre 2) was slightly but statistically significantly lower than in the validation set (centre 1).

CONCLUSIONS

Using the XGB algorithm potentially leads to an improved model for the classification of intrathoracic lymph nodes. The inclusion of primary tumour data improved the performance of the model, while additional knowledge of distant metastases did not. In patients in whom metastases are limited to lymph nodes in the thorax, this may reduce costly and invasive procedures such as endobronchial ultrasound or mediastinoscopy procedures.

[18F]FDG-PET/CT radiomics and artificial intelligence in lung cancer: Technical aspects and potential clinical applications

Lung cancer is the second most common cancer and the leading cause of cancer-related death worldwide. Molecular imaging using [¹⁸F]fluorodeoxyglucose Positron Emission Tomography and/or Computed Tomography ([¹⁸F]FDG-PET/CT) plays an essential role in the diagnosis, evaluation of response to treatment, and prediction of outcomes. The images are evaluated using qualitative and conventional quantitative indices. However, there is far more information embedded in the images, which can be extracted by sophisticated algorithms. Recently, the concept of uncovering and analyzing the invisible data extracted from medical images, called radiomics, is gaining more attention. Currently, [¹⁸F]FDG-PET/CT radiomics is growingly evaluated in lung cancer to discover if it enhances the diagnostic performance or implication of [¹⁸F]FDG-PET/CT in the management of lung cancer. In this review, we provide a short overview of the technical aspects, as they are discussed in different articles of this special issue. We mainly focus on the diagnostic performance of the [¹⁸F]FDG-PET/CT-based radiomics and the role of artificial intelligence in non-small cell lung cancer, impacting the early detection, staging, prediction of tumor subtypes, biomarkers, and patient's outcomes.

Using Natural Language Processing and Machine Learning to Preoperatively Predict Lymph Node Metastasis for Non-Small Cell Lung Cancer With Electronic Medical Records: Development and Validation Study

Background

Lymph node metastasis (LNM) is critical for treatment decision making of patients with resectable non–small cell lung cancer, but it is difficult to precisely diagnose preoperatively. Electronic medical records (EMRs) contain a large volume of valuable information about LNM, but some key information is recorded in free text, which hinders its secondary use.

Objective

This study aims to develop LNM prediction models based on EMRs using natural language processing (NLP) and machine learning algorithms.

Methods

We developed a multiturn question answering NLP model to extract features about the primary tumor and lymph nodes from computed tomography (CT) reports. We then combined these features with other structured clinical characteristics to develop LNM prediction models using machine learning algorithms. We conducted extensive experiments to explore the effectiveness of the predictive models and compared them with size criteria based on CT image findings (the maximum short axis diameter of lymph node >10 mm was regarded as a metastatic node) and clinician’s evaluation. Since the NLP model may extract features with mistakes, we also calculated the concordance correlation between the predicted probabilities of models using NLP-extracted features and gold standard features to explore the influence of NLP-driven automatic extraction.

Results

Experimental results show that the random forest models achieved the best performances with 0.792 area under the receiver operating characteristic curve (AUC) value and 0.456 average precision (AP) value for pN2 LNM prediction and 0.768 AUC value and 0.524 AP value for pN1&N2 LNM prediction. And all machine learning models outperformed the size criteria and clinician’s evaluation. The concordance correlation between the random forest models using NLP-extracted features and gold standard features is 0.950 and improved to 0.984 when the top 5 important NLP-extracted features were replaced with gold standard features.

Conclusions

The LNM models developed can achieve competitive performance using only limited EMR data such as CT reports and tumor markers in comparison with the clinician’s evaluation. The multiturn question answering NLP model can extract features effectively to support the development of LNM prediction models, which may facilitate the clinical application of predictive models.

Predictive Value of 18F-FDG PET/CT Using Machine Learning for Pathological Response to Neoadjuvant Concurrent Chemoradiotherapy in Patients with Stage III Non-Small Cell Lung Cancer

We investigated predictions from 18F-FDG PET/CT using machine learning (ML) to assess the neoadjuvant CCRT response of patients with stage III non-small cell lung cancer (NSCLC) and compared them with predictions from conventional PET parameters and from physicians. A retrospective study was conducted of 430 patients. They underwent 18F-FDG PET/CT before initial treatment and after neoadjuvant CCRT followed by curative surgery. We analyzed texture features from segmented tumors and reviewed the pathologic response. The ML model employed a random forest and was used to classify the binary outcome of the pathological complete response (pCR). The predictive accuracy of the ML model for the pCR was 93.4%. The accuracy of predicting pCR using the conventional PET parameters was up to 70.9%, and the accuracy of the physicians’ assessment was 80.5%. The accuracy of the prediction from the ML model was significantly higher than those derived from conventional PET parameters and provided by physicians (p < 0.05). The ML model is useful for predicting pCR after neoadjuvant CCRT, which showed a higher predictive accuracy than those achieved from conventional PET parameters and from physicians.