Radiomics-guided deep neural networks stratify lung adenocarcinoma prognosis from CT scans

Dose-Incorporated Deep Ensemble Learning for Improving Brain Metastasis Stereotactic Radiosurgery Outcome Prediction

PURPOSE

To develop a novel deep ensemble learning model for accurate prediction of brain metastasis (BM) local control outcomes after stereotactic radiosurgery (SRS).

METHODS AND MATERIALS

A total of 114 brain metastases (BMs) from 82 patients were evaluated, including 26 BMs that developed biopsy-confirmed local failure post-SRS. The SRS spatial dose distribution (Dmap) of each BM was registered to the planning contrast-enhanced T1 (T1-CE) magnetic resonance imaging (MRI). Axial slices of the Dmap, T1-CE, and planning target volume (PTV) segmentation (PTVseg) intersecting the BM center were extracted within a fixed field of view determined by the 60% isodose volume in Dmap. A spherical projection was implemented to transform planar image content onto a spherical surface using multiple projection centers, and the resultant T1-CE/Dmap/PTVseg projections were stacked as a 3-channel variable. Four Visual Geometry Group (VGG-19) deep encoders were used in an ensemble design, with each submodel using a different spherical projection formula as input for BM outcome prediction. In each submodel, clinical features after positional encoding were fused with VGG-19 deep features to generate logit results. The ensemble's outcome was synthesized from the 4 submodel results via logistic regression. In total, 10 model versions with random validation sample assignments were trained to study model robustness. Performance was compared with (1) a single VGG-19 encoder, (2) an ensemble with a T1-CE MRI as the sole image input after projections, and (3) an ensemble with the same image input design without clinical feature inclusion.

RESULTS

The ensemble model achieved an excellent area under the receiver operating characteristic curve (AUCROC: 0.89 ± 0.02) with high sensitivity (0.82 ± 0.05), specificity (0.84 ± 0.11), and accuracy (0.84 ± 0.08) results. This outperformed the MRI-only VGG-19 encoder (sensitivity: 0.35 ± 0.01, AUCROC: 0.64 ± 0.08), the MRI-only deep ensemble (sensitivity: 0.60 ± 0.09, AUCROC: 0.68 ± 0.06), and the 3-channel ensemble without clinical feature fusion (sensitivity: 0.78 ± 0.08, AUCROC: 0.84 ± 0.03).

CONCLUSIONS

Facilitated by the spherical image projection method, a deep ensemble model incorporating Dmap and clinical variables demonstrated excellent performance in predicting BM post-SRS local failure. Our novel approach could improve other radiation therapy outcome models and warrants further evaluation.

Cardiac ultrasomics for acute myocardial infarction risk stratification and prediction of all-cause mortality: a feasibility study

BACKGROUND

Current risk stratification tools for acute myocardial infarction (AMI) have limitations, particularly in predicting mortality. This study utilizes cardiac ultrasound radiomics (i.e., ultrasomics) to risk stratify AMI patients when predicting all-cause mortality.

RESULTS

The study included 197 patients: (a) retrospective internal cohort (n = 155) of non-ST-elevation myocardial infarction (n = 63) and ST-elevation myocardial infarction (n = 92) patients, and (b) external cohort from the multicenter Door-To-Unload in ST-segment-elevation myocardial infarction [DTU-STEMI] Pilot Trial (n = 42). Echocardiography images of apical 2, 3, and 4-chamber were processed through an automated deep-learning pipeline to extract ultrasomic features. Unsupervised machine learning (topological data analysis) generated AMI clusters followed by a supervised classifier to generate individual predicted probabilities. Validation included assessing the incremental value of predicted probabilities over the Global Registry of Acute Coronary Events (GRACE) risk score 2.0 to predict 1-year all-cause mortality in the internal cohort and infarct size in the external cohort. Three phenogroups were identified: Cluster A (high-risk), Cluster B (intermediate-risk), and Cluster C (low-risk). Cluster A patients had decreased LV ejection fraction (P < 0.01) and global longitudinal strain (P = 0.03) and increased mortality at 1-year (log rank P = 0.05). Ultrasomics features alone (C-Index: 0.74 vs. 0.70, P = 0.04) and combined with global longitudinal strain (C-Index: 0.81 vs. 0.70, P < 0.01) increased prediction of mortality beyond the GRACE 2.0 score. In the DTU-STEMI clinical trial, Cluster A was associated with larger infarct size (> 10% LV mass, P < 0.01), compared to remaining clusters.

CONCLUSIONS

Ultrasomics-based phenogroup clustering, augmented by TDA and supervised machine learning, provides a novel approach for AMI risk stratification.

Application of artificial intelligence in cancer diagnosis and tumor nanomedicine

Cancer is a major health concern due to its high incidence and mortality rates. Advances in cancer research, particularly in artificial intelligence (AI) and deep learning, have shown significant progress. The swift evolution of AI in healthcare, especially in tools like computer-aided diagnosis, has the potential to revolutionize early cancer detection. This technology offers improved speed, accuracy, and sensitivity, bringing a transformative impact on cancer diagnosis, treatment, and management. This paper provides a concise overview of the application of artificial intelligence in the realms of medicine and nanomedicine, with a specific emphasis on the significance and challenges associated with cancer diagnosis. It explores the pivotal role of AI in cancer diagnosis, leveraging structured, unstructured, and multimodal fusion data. Additionally, the article delves into the applications of AI in nanomedicine sensors and nano-oncology drugs. The fundamentals of deep learning and convolutional neural networks are clarified, underscoring their relevance to AI-driven cancer diagnosis. A comparative analysis is presented, highlighting the accuracy and efficiency of traditional methods juxtaposed with AI-based approaches. The discussion not only assesses the current state of AI in cancer diagnosis but also delves into the challenges faced by AI in this context. Furthermore, the article envisions the future development direction and potential application of artificial intelligence in cancer diagnosis, offering a hopeful prospect for enhanced cancer detection and improved patient prognosis.

Radiomics-Guided Deep Learning Networks Classify Differential Diagnosis of Parkinsonism

The differential diagnosis between atypical Parkinsonian syndromes may be challenging and critical. We aimed to proposed a radiomics-guided deep learning (DL) model to discover interpretable DL features and further verify the proposed model through the differential diagnosis of Parkinsonian syndromes. We recruited 1495 subjects for 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) scanning, including 220 healthy controls and 1275 patients diagnosed with idiopathic Parkinson's disease (IPD), multiple system atrophy (MSA), or progressive supranuclear palsy (PSP). Baseline radiomics and two DL models were developed and tested for the Parkinsonian diagnosis. The DL latent features were extracted from the last layer and subsequently guided by radiomics. The radiomics-guided DL model outperformed the baseline radiomics approach, suggesting the effectiveness of the DL approach. DenseNet showed the best diagnosis ability (sensitivity: 95.7%, 90.1%, and 91.2% for IPD, MSA, and PSP, respectively) using retained DL features in the test dataset. The retained DL latent features were significantly associated with radiomics features and could be interpreted through biological explanations of handcrafted radiomics features. The radiomics-guided DL model offers interpretable high-level abstract information for differential diagnosis of Parkinsonian disorders and holds considerable promise for personalized disease monitoring.

A Multichannel CT and Radiomics-Guided CNN-ViT (RadCT-CNNViT) Ensemble Network for Diagnosis of Pulmonary Sarcoidosis

Pulmonary sarcoidosis is a multisystem granulomatous interstitial lung disease (ILD) with a variable presentation and prognosis. The early accurate detection of pulmonary sarcoidosis may prevent progression to pulmonary fibrosis, a serious and potentially life-threatening form of the disease. However, the lack of a gold-standard diagnostic test and specific radiographic findings poses challenges in diagnosing pulmonary sarcoidosis. Chest computed tomography (CT) imaging is commonly used but requires expert, chest-trained radiologists to differentiate pulmonary sarcoidosis from lung malignancies, infections, and other ILDs. In this work, we develop a multichannel, CT and radiomics-guided ensemble network (RadCT-CNNViT) with visual explainability for pulmonary sarcoidosis vs. lung cancer (LCa) classification using chest CT images. We leverage CT and hand-crafted radiomics features as input channels, and a 3D convolutional neural network (CNN) and vision transformer (ViT) ensemble network for feature extraction and fusion before a classification head. The 3D CNN sub-network captures the localized spatial information of lesions, while the ViT sub-network captures long-range, global dependencies between features. Through multichannel input and feature fusion, our model achieves the highest performance with accuracy, sensitivity, specificity, precision, F1-score, and combined AUC of 0.93 ± 0.04, 0.94 ± 0.04, 0.93 ± 0.08, 0.95 ± 0.05, 0.94 ± 0.04, and 0.97, respectively, in a five-fold cross-validation study with pulmonary sarcoidosis (n = 126) and LCa (n = 93) cases. A detailed ablation study showing the impact of CNN + ViT compared to CNN or ViT alone, and CT + radiomics input, compared to CT or radiomics alone, is also presented in this work. Overall, the AI model developed in this work offers promising potential for triaging the pulmonary sarcoidosis patients for timely diagnosis and treatment from chest CT.

MRI-based radiomics for preoperative prediction of recurrence and metastasis in rectal cancer

OBJECTIVES

To explore the value of multi-parametric MRI (mp-MRI) radiomic model for preoperative prediction of recurrence and/or metastasis (RM) as well as survival benefits in patients with rectal cancer.

METHODS

A retrospective analysis of 234 patients from two centers with histologically confirmed rectal adenocarcinoma was conducted. All patients were divided into three groups: training, internal validation (in-vad) and external validation (ex-vad) sets. In the training set, radiomic features were extracted from T2WI, DWI, and contrast enhancement T1WI (CE-T1) sequence. Radiomic signature (RS) score was then calculated for feature screening to construct a rad-score model. Subsequently, preoperative clinical features with statistical significance were selected to construct a clinical model. Independent predictors from clinical and RS related to RM were selected to build the combined model and nomogram.

RESULTS

After feature extraction, 26 features were selected to construct the rad-score model. RS (OR = 0.007, p < 0.01), MR-detected T stage (mrT) (OR = 2.92, p = 0.03) and MR-detected circumferential resection margin (mrCRM) (OR = 4.70, p = 0.01) were identified as independent predictors of RM. Then, clinical model and combined model were constructed. ROC curve showed that the AUC, accuracy, sensitivity and specificity of the combined model were higher than that of the other two models in three sets. Kaplan-Meier curves showed that poorer disease-free survival (DFS) time was observed for patients in pT3-4 stages with low RS score (p < 0.001), similar results were also found in pCRM-positive patients (p < 0.05).

CONCLUSION

The mp-MRI radiomics model can be served as a noninvasive and accurate predictors of RM in rectal cancer that may support clinical decision-making.

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.

Foundation Models for Quantitative Biomarker Discovery in Cancer Imaging

Foundation models represent a recent paradigm shift in deep learning, where a single large-scale model trained on vast amounts of data can serve as the foundation for various downstream tasks. Foundation models are generally trained using self-supervised learning and excel in reducing the demand for training samples in downstream applications. This is especially important in medicine, where large labeled datasets are often scarce. Here, we developed a foundation model for imaging biomarker discovery by training a convolutional encoder through self-supervised learning using a comprehensive dataset of 11,467 radiographic lesions. The foundation model was evaluated in distinct and clinically relevant applications of imaging-based biomarkers. We found that they facilitated better and more efficient learning of imaging biomarkers and yielded task-specific models that significantly outperformed their conventional supervised counterparts on downstream tasks. The performance gain was most prominent when training dataset sizes were very limited. Furthermore, foundation models were more stable to input and inter-reader variations and showed stronger associations with underlying biology. Our results demonstrate the tremendous potential of foundation models in discovering novel imaging biomarkers that may extend to other clinical use cases and can accelerate the widespread translation of imaging biomarkers into clinical settings.

Head and neck cancer treatment outcome prediction: a comparison between machine learning with conventional radiomics features and deep learning radiomics

Background

Radiomics can provide in-depth characterization of cancers for treatment outcome prediction. Conventional radiomics rely on extraction of image features within a pre-defined image region of interest (ROI) which are typically fed to a classification algorithm for prediction of a clinical endpoint. Deep learning radiomics allows for a simpler workflow where images can be used directly as input to a convolutional neural network (CNN) with or without a pre-defined ROI.

Purpose

The purpose of this study was to evaluate (i) conventional radiomics and (ii) deep learning radiomics for predicting overall survival (OS) and disease-free survival (DFS) for patients with head and neck squamous cell carcinoma (HNSCC) using pre-treatment 18F-fluorodeoxuglucose positron emission tomography (FDG PET) and computed tomography (CT) images.

Materials and methods

FDG PET/CT images and clinical data of patients with HNSCC treated with radio(chemo)therapy at Oslo University Hospital (OUS; n = 139) and Maastricht University Medical Center (MAASTRO; n = 99) were collected retrospectively. OUS data was used for model training and initial evaluation. MAASTRO data was used for external testing to assess cross-institutional generalizability. Models trained on clinical and/or conventional radiomics features, with or without feature selection, were compared to CNNs trained on PET/CT images without or with the gross tumor volume (GTV) included. Model performance was measured using accuracy, area under the receiver operating characteristic curve (AUC), Matthew's correlation coefficient (MCC), and the F1 score calculated for both classes separately.

Results

CNNs trained directly on images achieved the highest performance on external data for both endpoints. Adding both clinical and radiomics features to these image-based models increased performance further. Conventional radiomics including clinical data could achieve competitive performance. However, feature selection on clinical and radiomics data lead to overfitting and poor cross-institutional generalizability. CNNs without tumor and node contours achieved close to on-par performance with CNNs including contours.

Conclusion

High performance and cross-institutional generalizability can be achieved by combining clinical data, radiomics features and medical images together with deep learning models. However, deep learning models trained on images without contours can achieve competitive performance and could see potential use as an initial screening tool for high-risk patients.

Value of radiomics in differentiating synchronous double primary lung adenocarcinomas from intrapulmonary metastasis

BACKGROUND

Distinguishing synchronous double primary lung adenocarcinoma (SDPLA) from intrapulmonary metastasis (IPM) of lung cancer has significant therapeutic and prognostic values. This study aimed to develop and validate a CT-based radiomics model to differentiate SDPLA from IPM.

METHODS

A total of 153 patients (93 SDPLA and 60 IPM) with 306 pathologically confirmed lesions were retrospectively studied. CT morphological features were also recorded. Region of interest (ROI) segmentation was performed semiautomatically, and 1,037 radiomics features were extracted from every segmented lesion The differences of radiomics features were defined as the relative net difference in radiomics features between the two lesions on CT. Those low reliable (ICC <0.75) and redundant (r>0.9) features were excluded by intraclass correlation coefficients (ICC) and Pearson's correlation. Multivariate logistic regression (LR) algorithm was used to establish the classification model according to the selected features. The radiomics model was based on the four most contributing differences of radiomics features. Clinical-CT model and MixModel were based on selected clinical and CT features only and the combination of clinical-CT and Rad-score, respectively.

RESULTS

In both the training and testing cohorts, the area under the curves (AUCs) of the radiomics model were larger than those of the clinical-CT model (0.944 vs. 0.793 and 0.886 vs. 0.735 on training and testing cohorts, respectively), and statistically significant differences between the two models in the testing set were found (P<0.001). Meanwhile, three radiologists had sensitivities of 84.2%, 63.9%, and 68.4%, and specificities of 76.9%, 69.2%, and 76.9% in differentiating 19 SDPLA cases from 13 cases of IPM in the testing set. Compared with the performance of the three radiologists, the radiomics model showed better accuracy to the patients in both the training and testing cohorts. Among the three models, the radiomics model showed the best net benefits.

CONCLUSIONS

The differences of radiomics features showed excellent diagnostic performance for preoperative differentiation between synchronous double primary lung adenocarcinoma from interpulmonary metastasis, superior to the clinical model and decisions made by radiologists.

An ensemble deep learning model for risk stratification of invasive lung adenocarcinoma using thin-slice CT

Lung cancer screening using computed tomography (CT) has increased the detection rate of small pulmonary nodules and early-stage lung adenocarcinoma. It would be clinically meaningful to accurate assessment of the nodule histology by CT scans with advanced deep learning algorithms. However, recent studies mainly focus on predicting benign and malignant nodules, lacking of model for the risk stratification of invasive adenocarcinoma. We propose an ensemble multi-view 3D convolutional neural network (EMV-3D-CNN) model to study the risk stratification of lung adenocarcinoma. We include 1075 lung nodules (≤30 mm and ≥4 mm) with preoperative thin-section CT scans and definite pathology confirmed by surgery. Our model achieves a state-of-art performance of 91.3% and 92.9% AUC for diagnosis of benign/malignant and pre-invasive/invasive nodules, respectively. Importantly, our model outperforms senior doctors in risk stratification of invasive adenocarcinoma with 77.6% accuracy [i.e., Grades 1, 2, 3]). It provides detailed predictive histological information for the surgical management of pulmonary nodules. Finally, for user-friendly access, the proposed model is implemented as a web-based system ( https://seeyourlung.com.cn ).

Novel Mechanisms and Future Opportunities for the Management of Radiation Necrosis in Patients Treated for Brain Metastases in the Era of Immunotherapy

Radiation necrosis, also known as treatment-induced necrosis, has emerged as an important adverse effect following stereotactic radiotherapy (SRS) for brain metastases. The improved survival of patients with brain metastases and increased use of combined systemic therapy and SRS have contributed to a growing incidence of necrosis. The cyclic GMP-AMP (cGAMP) synthase (cGAS) and stimulator of interferon genes (STING) pathway (cGAS-STING) represents a key biological mechanism linking radiation-induced DNA damage to pro-inflammatory effects and innate immunity. By recognizing cytosolic double-stranded DNA, cGAS induces a signaling cascade that results in the upregulation of type 1 interferons and dendritic cell activation. This pathway could play a key role in the pathogenesis of necrosis and provides attractive targets for therapeutic development. Immunotherapy and other novel systemic agents may potentiate activation of cGAS-STING signaling following radiotherapy and increase necrosis risk. Advancements in dosimetric strategies, novel imaging modalities, artificial intelligence, and circulating biomarkers could improve the management of necrosis. This review provides new insights into the pathophysiology of necrosis and synthesizes our current understanding regarding the diagnosis, risk factors, and management options of necrosis while highlighting novel avenues for discovery.

Characterization of Mediastinal Bulky Lymphomas with FDG-PET-Based Radiomics and Machine Learning Techniques

BACKGROUND

This study tested the diagnostic value of 18F-FDG PET/CT (FDG-PET) volumetric and texture parameters in the histological differentiation of mediastinal bulky disease due to classical Hodgkin lymphoma (cHL), primary mediastinal B-cell lymphoma (PMBCL) and grey zone lymphoma (GZL), using machine learning techniques.

METHODS

We reviewed 80 cHL, 29 PMBCL and 8 GZL adult patients with mediastinal bulky disease and histopathological diagnoses who underwent FDG-PET pre-treatment. Volumetric and radiomic parameters were measured using FDG-PET both for bulky lesions (BL) and for all lesions (AL) using LIFEx software (threshold SUV ≥ 2.5). Binary and multiclass classifications were performed with various machine learning techniques fed by a relevant subset of radiomic features.

RESULTS

The analysis showed significant differences between the lymphoma groups in terms of SUVmax, SUVmean, MTV, TLG and several textural features of both first- and second-order grey level. Among machine learning classifiers, the tree-based ensembles achieved the best performance both for binary and multiclass classifications in histological differentiation.

CONCLUSIONS

Our results support the value of metabolic heterogeneity as an imaging biomarker, and the use of radiomic features for early characterization of mediastinal bulky lymphoma.

Incremental benefits of size-zone matrix-based radiomics features for the prognosis of lung adenocarcinoma: advantage of spatial partitioning on tumor evaluation

OBJECTIVES

Prognostic models of lung adenocarcinoma (ADC) can be built using radiomics features from various categories. The size-zone matrix (SZM) features have a strong biological basis related to tumor partitioning, but their incremental benefits have not been fully explored. In our study, we aimed to evaluate the incremental benefits of SZM features for the prognosis of lung ADC.

METHODS

A total of 298 patients were included and their pretreatment computed tomography images were analyzed in fivefold cross-validation. We built a risk model of overall survival using SZM features and compared it with a conventional radiomics risk model and a clinical variable-based risk model. We also compared it with other models incorporating various combinations of SZM features, other radiomics features, and clinical variables. A total of seven risk models were compared and evaluated using the hazard ratio (HR) on the left-out test fold.

RESULTS

As a baseline, the clinical variable risk model showed an HR of 2.739. Combining the radiomics signature with SZM feature was better (HR 4.034) than using radiomics signature alone (HR 3.439). Combining radiomics signature, SZM feature, and clinical variable (HR 6.524) fared better than just combining radiomics signature and clinical variables (HR 4.202). These results confirmed the added benefits of SZM features for prognosis in lung ADC.

CONCLUSION

Combining SZM feature with the radiomics signature was better than using the radiomics signature alone and the benefits of SZM features were maintained when clinical variables were added confirming the incremental benefits of SZM features for lung ADC prognosis.

KEY POINTS

• Size-zone matrix (SZM) features provide incremental benefits for the prognosis of lung adenocarcinoma. • Combining the radiomics signature with SZM features performed better than using a radiomics signature alone.

Radiomics in Urolithiasis: Systematic Review of Current Applications, Limitations, and Future Directions

Radiomics is increasingly applied to the diagnosis, management, and outcome prediction of various urological conditions. Urolithiasis is a common benign condition with a high incidence and recurrence rate. The purpose of this scoping review is to evaluate the current evidence of the application of radiomics in urolithiasis, especially its utility in diagnostics and therapeutics. An electronic literature search on radiomics in the setting of urolithiasis was conducted on PubMed, EMBASE, and Scopus from inception to 21 March 2022. A total of 7 studies were included. Radiomics has been successfully applied in the field of urolithiasis to differentiate phleboliths from calculi and classify stone types and composition pre-operatively. More importantly, it has also been utilized to predict outcomes and complications after endourological procedures. Although radiomics in urolithiasis is still in its infancy, it has the potential for large-scale implementation. Its greatest potential lies in the correlation with conventional established diagnostic and therapeutic factors.

How Many Private Data Are Needed for Deep Learning in Lung Nodule Detection on CT Scans? A Retrospective Multicenter Study

Simple Summary

The early detection of lung nodules is important for patient treatment and follow-up. Many researchers are investigating deep-learning-based lung nodule detection to ease the burden of lung nodule detection by radiologists. The purpose of this paper is to provide guidelines for collecting lung nodule data to facilitate research. We collected chest computed tomography scans reviewed by radiologists at three hospitals. In addition, several experiments were conducted using the large-scale open dataset, LUNA16. As a result of the experiment, it was possible to prove the value of using the collected data compared to using LUNA16. We also demonstrated the effectiveness of transfer learning from pre-trained learning weights in LUNA16. Finally, our study provides information on the amount of lung nodule data that must be collected to stabilize lung nodule detection performance.

Abstract

Early detection of lung nodules is essential for preventing lung cancer. However, the number of radiologists who can diagnose lung nodules is limited, and considerable effort and time are required. To address this problem, researchers are investigating the automation of deep-learning-based lung nodule detection. However, deep learning requires large amounts of data, which can be difficult to collect. Therefore, data collection should be optimized to facilitate experiments at the beginning of lung nodule detection studies. We collected chest computed tomography scans from 515 patients with lung nodules from three hospitals and high-quality lung nodule annotations reviewed by radiologists. We conducted several experiments using the collected datasets and publicly available data from LUNA16. The object detection model, YOLOX was used in the lung nodule detection experiment. Similar or better performance was obtained when training the model with the collected data rather than LUNA16 with large amounts of data. We also show that weight transfer learning from pre-trained open data is very useful when it is difficult to collect large amounts of data. Good performance can otherwise be expected when reaching more than 100 patients. This study offers valuable insights for guiding data collection in lung nodules studies in the future.

A polygenic stacking classifier revealed the complicated platelet transcriptomic landscape of adult immune thrombocytopenia

Immune thrombocytopenia (ITP) is an autoimmune disease with the typical symptom of a low platelet count in blood. ITP demonstrated age and sex biases in both occurrences and prognosis, and adult ITP was mainly induced by the living environments. The current diagnosis guideline lacks the integration of molecular heterogenicity. This study recruited the largest cohort of platelet transcriptome samples. A comprehensive procedure of feature selection, feature engineering, and stacking classification was carried out to detect the ITP biomarkers using RNA sequencing (RNA-seq) transcriptomes. The 40 detected biomarkers were loaded to train the final ITP detection model, with an overall accuracy 0.974. The biomarkers suggested that ITP onset may be associated with various transcribed components, including protein-coding genes, long intergenic non-coding RNA (lincRNA) genes, and pseudogenes with apparent transcriptions. The delivered ITP detection model may also be utilized as a complementary ITP diagnosis tool. The code and the example dataset is freely available on http://www.healthinformaticslab.org/supp/resources.php

Towards a safe and efficient clinical implementation of machine learning in radiation oncology by exploring model interpretability, explainability and data-model dependency

The interest in machine learning (ML) has grown tremendously in recent years, partly due to the performance leap that occurred with new techniques of deep learning, convolutional neural networks for images, increased computational power, and wider availability of large datasets. Most fields of medicine follow that popular trend and, notably, radiation oncology is one of those that are at the forefront, with already a long tradition in using digital images and fully computerized workflows. ML models are driven by data, and in contrast with many statistical or physical models, they can be very large and complex, with countless generic parameters. This inevitably raises two questions, namely, the tight dependence between the models and the datasets that feed them, and the interpretability of the models, which scales with its complexity. Any problems in the data used to train the model will be later reflected in their performance. This, together with the low interpretability of ML models, makes their implementation into the clinical workflow particularly difficult. Building tools for risk assessment and quality assurance of ML models must involve then two main points: interpretability and data-model dependency. After a joint introduction of both radiation oncology and ML, this paper reviews the main risks and current solutions when applying the latter to workflows in the former. Risks associated with data and models, as well as their interaction, are detailed. Next, the core concepts of interpretability, explainability, and data-model dependency are formally defined and illustrated with examples. Afterwards, a broad discussion goes through key applications of ML in workflows of radiation oncology as well as vendors' perspectives for the clinical implementation of ML.