Identifying the histologic subtypes of non-small cell lung cancer with computed tomography imaging: a comparative study of capsule net, convolutional neural network, and radiomics

A radiomics-based brain network in T1 images: construction, attributes, and applications

T1 image is a widely collected imaging sequence in various neuroimaging datasets, but it is rarely used to construct an individual-level brain network. In this study, a novel individualized radiomics-based structural similarity network was proposed from T1 images. In detail, it used voxel-based morphometry to obtain the preprocessed gray matter images, and radiomic features were then extracted on each region of interest in Brainnetome atlas, and an individualized radiomics-based structural similarity network was finally built using the correlational values of radiomic features between any pair of regions of interest. After that, the network characteristics of individualized radiomics-based structural similarity network were assessed, including graph theory attributes, test-retest reliability, and individual identification ability (fingerprinting). At last, two representative applications for individualized radiomics-based structural similarity network, namely mild cognitive impairment subtype discrimination and fluid intelligence prediction, were exemplified and compared with some other networks on large open-source datasets. The results revealed that the individualized radiomics-based structural similarity network displays remarkable network characteristics and exhibits advantageous performances in mild cognitive impairment subtype discrimination and fluid intelligence prediction. In summary, the individualized radiomics-based structural similarity network provides a distinctive, reliable, and informative individualized structural brain network, which can be combined with other networks such as resting-state functional connectivity for various phenotypic and clinical applications.

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.

Predicting the pathological invasiveness in patients with a solitary pulmonary nodule via Shapley additive explanations interpretation of a tree-based machine learning radiomics model: a multicenter study

Background

Radiomics models could help assess the benign and malignant invasiveness and prognosis of pulmonary nodules. However, the lack of interpretability limits application of these models. We thus aimed to construct and validate an interpretable and generalized computed tomography (CT) radiomics model to evaluate the pathological invasiveness in patients with a solitary pulmonary nodule in order to improve the management of these patients.

Methods

We retrospectively enrolled 248 patients with CT-diagnosed solitary pulmonary nodules. Radiomic features were extracted from nodular region and perinodular regions of 3 and 5 mm. After coarse-to-fine feature selection, the radiomics score (radscore) was calculated using the least absolute shrinkage and selection operator logistic method. Univariate and multivariate logistic regression analyses were performed to determine the invasiveness-related clinicoradiological factors. The clinical-radiomics model was then constructed using the logistic and extreme gradient boosting (XGBoost) algorithms. The Shapley additive explanations (SHAP) method was then used to explain the contributions of the features. After removing batch effects with the ComBat algorithm, we assessed the generalization of the explainable clinical-radiomics model in two independent external validation cohorts (n=147 and n=149).

Results

The clinical-radiomic XGBoost model integrating the radscore, CT value, nodule length, and crescent sign demonstrated better predictive performance than did the clinical-radiomics logistic model in assessing pulmonary nodule invasiveness, with an area under the receiver operating characteristic (ROC) curve (AUC) of 0.889 [95% confidence interval (CI), 0.848-0.927] in the training cohort. The SHAP algorithm illustrates the contribution of each feature in the final model. The specific model decision process was visualized using a tree-based decision heatmap. Satisfactory generalization performance was shown with AUCs of 0.889 (95% CI, 0.823-0.942) and 0.915 (95% CI, 0.851-0.963) in the two external validation cohorts.

Conclusions

An interpretable and generalized clinical-radiomics model for predicting pulmonary nodule invasibility was constructed to help clinicians determine the invasiveness of pulmonary nodules and devise assessment strategies in an easily understandable manner.

On-cloud decision-support system for non-small cell lung cancer histology characterization from thorax computed tomography scans

Non-Small Cell Lung Cancer (NSCLC) accounts for about 85% of all lung cancers. Developing non-invasive techniques for NSCLC histology characterization may not only help clinicians to make targeted therapeutic treatments but also prevent subjects from undergoing lung biopsy, which is challenging and could lead to clinical implications. The motivation behind the study presented here is to develop an advanced on-cloud decision-support system, named LUCY, for non-small cell LUng Cancer histologY characterization directly from thorax Computed Tomography (CT) scans. This aim was pursued by selecting thorax CT scans of 182 LUng ADenocarcinoma (LUAD) and 186 LUng Squamous Cell carcinoma (LUSC) subjects from four openly accessible data collections (NSCLC-Radiomics, NSCLC-Radiogenomics, NSCLC-Radiomics-Genomics and TCGA-LUAD), in addition to the implementation and comparison of two end-to-end neural networks (the core layer of whom is a convolutional long short-term memory layer), the performance evaluation on test dataset (NSCLC-Radiomics-Genomics) from a subject-level perspective in relation to NSCLC histological subtype location and grade, and the dynamic visual interpretation of the achieved results by producing and analyzing one heatmap video for each scan. LUCY reached test Area Under the receiver operating characteristic Curve (AUC) values above 77% in all NSCLC histological subtype location and grade groups, and a best AUC value of 97% on the entire dataset reserved for testing, proving high generalizability to heterogeneous data and robustness. Thus, LUCY is a clinically-useful decision-support system able to timely, non-invasively and reliably provide visually-understandable predictions on LUAD and LUSC subjects in relation to clinically-relevant information.

Multi-task learning-based histologic subtype classification of non-small cell lung cancer

PURPOSE

In clinical applications, accurate histologic subtype classification of lung cancer is important for determining appropriate treatment plans. The purpose of this paper is to evaluate the role of multi-task learning in the classification of adenocarcinoma and squamous cell carcinoma.

MATERIAL AND METHODS

In this paper, we propose a novel multi-task learning model for histologic subtype classification of non-small cell lung cancer based on computed tomography (CT) images. The model consists of a histologic subtype classification branch and a staging branch, which share a part of the feature extraction layers and are simultaneously trained. By optimizing on the two tasks simultaneously, our model could achieve high accuracy in histologic subtype classification of non-small cell lung cancer without relying on physician's precise labeling of tumor areas. In this study, 402 cases from The Cancer Imaging Archive (TCIA) were used in total, and they were split into training set (n = 258), internal test set (n = 66) and external test set (n = 78).

RESULTS

Compared with the radiomics method and single-task networks, our multi-task model could reach an AUC of 0.843 and 0.732 on internal and external test set, respectively. In addition, multi-task network can achieve higher accuracy and specificity than single-task network.

CONCLUSION

Compared with the radiomics methods and single-task networks, our multi-task learning model could improve the accuracy of histologic subtype classification of non-small cell lung cancer by sharing network layers, which no longer relies on the physician's precise labeling of lesion regions and could further reduce the manual workload of physicians.

Feature selection methods and predictive models in CT lung cancer radiomics

Radiomics is a technique that extracts quantitative features from medical images using data-characterization algorithms. Radiomic features can be used to identify tissue characteristics and radiologic phenotyping that is not observable by clinicians. A typical workflow for a radiomics study includes cohort selection, radiomic feature extraction, feature and predictive model selection, and model training and validation. While there has been increasing attention given to radiomic feature extraction, standardization, and reproducibility, currently, there is a lack of rigorous evaluation of feature selection methods and predictive models. Herein, we review the published radiomics investigations in CT lung cancer and provide an overview of the commonly used radiomic feature selection methods and predictive models. We also compare limitations of various methods in clinical applications and present sources of uncertainty associated with those methods. This review is expected to help raise awareness of the impact of radiomic feature and model selection methods on the integrity of radiomics studies.

Lung nodule diagnosis and cancer histology classification from computed tomography data by convolutional neural networks: A survey

Lung cancer is among the deadliest cancers. Besides lung nodule classification and diagnosis, developing non-invasive systems to classify lung cancer histological types/subtypes may help clinicians to make targeted treatment decisions timely, having a positive impact on patients' comfort and survival rate. As convolutional neural networks have proven to be responsible for the significant improvement of the accuracy in lung cancer diagnosis, with this survey we intend to: show the contribution of convolutional neural networks not only in identifying malignant lung nodules but also in classifying lung cancer histological types/subtypes directly from computed tomography data; point out the strengths and weaknesses of slice-based and scan-based approaches employing convolutional neural networks; and highlight the challenges and prospective solutions to successfully apply convolutional neural networks for such classification tasks. To this aim, we conducted a comprehensive analysis of relevant Scopus-indexed studies involved in lung nodule diagnosis and cancer histology classification up to January 2022, dividing the investigation in convolutional neural network-based approaches fed with planar or volumetric computed tomography data. Despite the application of convolutional neural networks in lung nodule diagnosis and cancer histology classification is a valid strategy, some challenges raised, mainly including the lack of publicly-accessible annotated data, together with the lack of reproducibility and clinical interpretability. We believe that this survey will be helpful for future studies involved in lung nodule diagnosis and cancer histology classification prior to lung biopsy by means of convolutional neural networks.

TongueCaps: An Improved Capsule Network Model for Multi-Classification of Tongue Color

Tongue color is an important part of tongue diagnosis. The change of tongue color is affected by pathological state of body, blood rheology, and other factors. Therefore, physicians can understand a patient's condition by observing tongue color. Currently, most studies use machine learning, which is time consuming and labor intensive. Other studies use deep learning based on convolutional neural network (CNN), but the affine transformation of CNN is less robust and easily loses the spatial relationship between features. Recently, Capsule Networks (CapsNet) have been proposed to overcome these problems. In our work, CapsNet is used for tongue color research for the first time, and improved model TongueCaps is proposed, which combines the advantage of CapsNet and residual block structure to achieve end to end tongue color classification. We conduct experiments on 1371 tongue images; TongueCaps achieve accuracy is 0.8456, sensitivity is 0.8474, and specificity is 0.9586. In addition, the size of TongueCaps is 8.11 M, and FLOPs is 1,335,342, which are smaller than CNN in comparison models. Experiments have confirmed that the CapsNet can be used for tongue color research, and improved model TongueCaps, in this paper, is superior to other comparison models in terms of accuracy, specificity and sensitivity, computational complexity, and size of model.

Predicting the Ki-67 proliferation index in pulmonary adenocarcinoma patients presenting with subsolid nodules: construction of a nomogram based on CT images

BACKGROUND

The Ki-67 proliferation index (PI) reflects the proliferation of cells. However, the conventional methods for the acquisition of the Ki-67 PI, such as surgery and biopsy, are generally invasive. This study investigated a potential noninvasive method of predicting the Ki-67 PI in patients with lung adenocarcinoma presenting with subsolid nodules.

METHODS

This retrospective study enrolled 153 patients who presented with pulmonary adenocarcinoma appearing as subsolid nodules (SSNs) on computed tomography (CT) images between January 2015 and December 2018. Presence of LUAD with SSNs was confirmed by histopathology. Of these participants, 107 patients were from institution 1 and were divided into a training cohort and an internal validation cohort in a 7:3 ratio. The other 46 patients were from institution 2 and were enrolled as an external validation cohort. All patients underwent conventional CT scans with thin-slice (≤1.25 mm) reconstruction, and 1,316 quantitative radiomic features were extracted from the CT images for each nodule. The minimum redundancy maximum relevance and the least absolute shrinkage and selection operator were used for feature selection, and the radiomics signature was constructed based on these selected features. Clinical features were examined using univariate logistic regression analysis. The nomogram was developed based on the radiomics signature and the independent clinical risk factors. The Delong test and t test were employed for statistical analysis. The performance of different models was assessed by the receiver operating characteristic (ROC) curve.

RESULTS

The diameter of the nodules [odds ratio (OR) =1.17; P=0.003] was identified as an independent predictive parameter. Both the radiomics signature and the nomogram suggested a good predictive probability for Ki-67 expression. For the radiomics signature, the area under the ROC curve (AUC) for the training cohort, the internal validation cohort, and the external validation cohort was 0.86 [95% confidence interval (CI): 0.77 to 0.95], 0.81 (95% CI: 0.64 to 0.98), and 0.77 (95% CI: 0.62 to 0.91), respectively. For the nomogram, the AUC for the training cohort, the internal validation cohort, and the external validation cohort was 0.86 (95% CI: 0.77 to 0.95), 0.80 (95% CI: 0.64 to 0.97), and 0.79 (95% CI: 0.65 to 0.94), respectively. There were no statistical differences in the AUCs between the radiomics signature and the radiomic nomogram in the training cohort or the validation cohorts (all P>0.05).

CONCLUSIONS

The nomogram provides a novel strategy for determining the Ki-67 PI in predicting the proliferation of subsolid nodules, which may be beneficial for the management of patients with SSNs.

Analysis of EPID Transmission Fluence Maps Using Machine Learning Models and CNN for Identifying Position Errors in the Treatment of GO Patients

Purpose

To find a suitable method for analyzing electronic portal imaging device (EPID) transmission fluence maps for the identification of position errors in the in vivo dose monitoring of patients with Graves' ophthalmopathy (GO).

Methods

Position errors combining 0-, 2-, and 4-mm errors in the left-right (LR), anterior-posterior (AP), and superior-inferior (SI) directions in the delivery of 40 GO patient radiotherapy plans to a human head phantom were simulated and EPID transmission fluence maps were acquired. Dose difference (DD) and structural similarity (SSIM) maps were calculated to quantify changes in the fluence maps. Three types of machine learning (ML) models that utilize radiomics features of the DD maps (ML 1 models), features of the SSIM maps (ML 2 models), and features of both DD and SSIM maps (ML 3 models) as inputs were used to perform three types of position error classification, namely a binary classification of the isocenter error (type 1), three binary classifications of LR, SI, and AP direction errors (type 2), and an eight-element classification of the combined LR, SI, and AP direction errors (type 3). Convolutional neural network (CNN) was also used to classify position errors using the DD and SSIM maps as input.

Results

The best-performing ML 1 model was XGBoost, which achieved accuracies of 0.889, 0.755, 0.778, 0.833, and 0.532 in the type 1, type 2-LR, type 2-AP, type 2-SI, and type 3 classification, respectively. The best ML 2 model was XGBoost, which achieved accuracies of 0.856, 0.731, 0.736, 0.949, and 0.491, respectively. The best ML 3 model was linear discriminant classifier (LDC), which achieved accuracies of 0.903, 0.792, 0.870, 0.931, and 0.671, respectively. The CNN achieved classification accuracies of 0.925, 0.833, 0.875, 0.949, and 0.689, respectively.

Conclusion

ML models and CNN using combined DD and SSIM maps can analyze EPID transmission fluence maps to identify position errors in the treatment of GO patients. Further studies with large sample sizes are needed to improve the accuracy of CNN.

Impact of Lesion Delineation and Intensity Quantisation on the Stability of Texture Features from Lung Nodules on CT: A Reproducible Study

Computer-assisted analysis of three-dimensional imaging data (radiomics) has received a lot of research attention as a possible means to improve the management of patients with lung cancer. Building robust predictive models for clinical decision making requires the imaging features to be stable enough to changes in the acquisition and extraction settings. Experimenting on 517 lung lesions from a cohort of 207 patients, we assessed the stability of 88 texture features from the following classes: first-order (13 features), Grey-level Co-Occurrence Matrix (24), Grey-level Difference Matrix (14), Grey-level Run-length Matrix (16), Grey-level Size Zone Matrix (16) and Neighbouring Grey-tone Difference Matrix (five). The analysis was based on a public dataset of lung nodules and open-access routines for feature extraction, which makes the study fully reproducible. Our results identified 30 features that had good or excellent stability relative to lesion delineation, 28 to intensity quantisation and 18 to both. We conclude that selecting the right set of imaging features is critical for building clinical predictive models, particularly when changes in lesion delineation and/or intensity quantisation are involved.