Dataset Information

CIRDataset: A large-scale Dataset for Clinically-Interpretable lung nodule Radiomics and malignancy prediction.

ABSTRACT: Spiculations/lobulations, sharp/curved spikes on the surface of lung nodules, are good predictors of lung cancer malignancy and hence, are routinely assessed and reported by radiologists as part of the standardized Lung-RADS clinical scoring criteria. Given the 3D geometry of the nodule and 2D slice-by-slice assessment by radiologists, manual spiculation/lobulation annotation is a tedious task and thus no public datasets exist to date for probing the importance of these clinically-reported features in the SOTA malignancy prediction algorithms. As part of this paper, we release a large-scale Clinically-Interpretable Radiomics Dataset, CIRDataset, containing 956 radiologist QA/QC'ed spiculation/lobulation annotations on segmented lung nodules from two public datasets, LIDC-IDRI (N=883) and LUNGx (N=73). We also present an end-to-end deep learning model based on multi-class Voxel2Mesh extension to segment nodules (while preserving spikes), classify spikes (sharp/spiculation and curved/lobulation), and perform malignancy prediction. Previous methods have performed malignancy prediction for LIDC and LUNGx datasets but without robust attribution to any clinically reported/actionable features (due to known hyperparameter sensitivity issues with general attribution schemes). With the release of this comprehensively-annotated CIRDataset and end-to-end deep learning baseline, we hope that malignancy prediction methods can validate their explanations, benchmark against our baseline, and provide clinically-actionable insights. Dataset, code, pretrained models, and docker containers are available at https://github.com/nadeemlab/CIR.

SUBMITTER: Choi W

PROVIDER: S-EPMC9527770 | biostudies-literature | 2022 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

CIRDataset: A large-scale Dataset for Clinically-Interpretable lung nodule Radiomics and malignancy prediction.

Choi Wookjin W Dahiya Navdeep N Nadeem Saad S

Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention 20220916

Spiculations/lobulations, sharp/curved spikes on the surface of lung nodules, are good predictors of lung cancer malignancy and hence, are routinely assessed and reported by radiologists as part of the standardized Lung-RADS clinical scoring criteria. Given the 3D geometry of the nodule and 2D slice-by-slice assessment by radiologists, manual spiculation/lobulation annotation is a tedious task and thus no public datasets exist to date for probing the importance of these clinically-reported featu ...[more]

PMID: 36198166

Similar Datasets

Project description:ObjectivesBoth radiomics and deep learning methods have shown great promise in predicting lesion malignancy in various image-based oncology studies. However, it is still unclear which method to choose for a specific clinical problem given the access to the same amount of training data. In this study, we try to compare the performance of a series of carefully selected conventional radiomics methods, end-to-end deep learning models, and deep-feature based radiomics pipelines for pulmonary nodule malignancy prediction on an open database that consists of 1297 manually delineated lung nodules.MethodsConventional radiomics analysis was conducted by extracting standard handcrafted features from target nodule images. Several end-to-end deep classifier networks, including VGG, ResNet, DenseNet, and EfficientNet were employed to identify lung nodule malignancy as well. In addition to the baseline implementations, we also investigated the importance of feature selection and class balancing, as well as separating the features learned in the nodule target region and the background/context region. By pooling the radiomics and deep features together in a hybrid feature set, we investigated the compatibility of these two sets with respect to malignancy prediction.ResultsThe best baseline conventional radiomics model, deep learning model, and deep-feature based radiomics model achieved AUROC values (mean ± standard deviations) of 0.792 ± 0.025, 0.801 ± 0.018, and 0.817 ± 0.032, respectively through 5-fold cross-validation analyses. However, after trying out several optimization techniques, such as feature selection and data balancing, as well as adding context features, the corresponding best radiomics, end-to-end deep learning, and deep-feature based models achieved AUROC values of 0.921 ± 0.010, 0.824 ± 0.021, and 0.936 ± 0.011, respectively. We achieved the best prediction accuracy from the hybrid feature set (AUROC: 0.938 ± 0.010).ConclusionThe end-to-end deep-learning model outperforms conventional radiomics out of the box without much fine-tuning. On the other hand, fine-tuning the models lead to significant improvements in the prediction performance where the conventional and deep-feature based radiomics models achieved comparable results. The hybrid radiomics method seems to be the most promising model for lung nodule malignancy prediction in this comparative study.

Project description:ObjectivesPost-imaging mathematical prediction models (MPMs) provide guidance for the management of solid pulmonary nodules by providing a lung cancer risk score from demographic and radiologists-indicated imaging characteristics. We hypothesized calibrating the MPM risk score threshold to a local study cohort would result in improved performance over the original recommended MPM thresholds. We compared the pre- and post-calibration performance of four MPM models and determined if improvement in MPM prediction occurs as nodules are imaged longitudinally.Materials and methodsA common cohort of 317 individuals with computed tomography-detected, solid nodules (80 malignant, 237 benign) were used to evaluate the MPM performance. We created a web-based application for this study that allows others to easily calibrate thresholds and analyze the performance of MPMs on their local cohort. Thirty patients with repeated imaging were tested for improved performance longitudinally.ResultsUsing calibrated thresholds, Mayo Clinic and Brock University (BU) MPMs performed the best (AUC = 0.63, 0.61) compared to the Veteran's Affairs (0.51) and Peking University (0.55). Only BU had consensus with the original MPM threshold; the other calibrated thresholds improved MPM accuracy. No significant improvements in accuracy were found longitudinally between time points.ConclusionsCalibration to a common cohort can select the best-performing MPM for your institution. Without calibration, BU has the most stable performance in solid nodules ≥ 8 mm but has only moderate potential to refine subjects into appropriate workup. Application of MPM is recommended only at initial evaluation as no increase in accuracy was achieved over time.Key points• Post-imaging lung cancer risk mathematical predication models (MPMs) perform poorly on local populations without calibration. • An application is provided to facilitate calibration to new study cohorts: the Mayo Clinic model, the U.S. Department of Veteran's Affairs model, the Brock University model, and the Peking University model. • No significant improvement in risk prediction occurred in nodules with repeated imaging sessions, indicating the potential value of risk prediction application is limited to the initial evaluation.

Project description:Crystal structure prediction (CSP) seeks to identify all thermodynamically accessible solid forms of a given compound and, crucially, to establish the relative thermodynamic stability between different polymorphs. The conventional hierarchical CSP workflow suggests that no single energy model can fulfill the needs of all stages in the workflow, and energy models across a spectrum of fidelities and computational costs are required. Hybrid ab initio/empirical force-field (HAIEFF) models have demonstrated a good balance of these two factors, but the force-field component presents a major bottleneck for model accuracy. Existing parameter estimation tools for fitting this empirical component are inefficient and have severe limitations on the manageable problem size. This, combined with a lack of reliable reference data for parameter fitting, has resulted in development in the force-field component of HAIEFF models having mostly stagnated. In this work, we address these barriers to progress. First, we introduce a curated database of 755 organic crystal structures, obtained using high quality, solid-state DFT-D calculations, which provide a complete set of geometry and energy data. Comparisons to various theoretical and experimental data sources indicate that this database provides suitable diversity for parameter fitting. In tandem, we also put forward a new parameter estimation algorithm implemented as the CrystalEstimator program. Our tests demonstrate that CrystalEstimator is capable of efficiently handling large-scale parameter estimation problems, simultaneously fitting as many as 62 model parameters based on data from 445 structures. This problem size far exceeds any previously reported works related to CSP force-field parametrization. These developments form a strong foundation for all future work involving parameter estimation of transferable or tailor-made force-fields for HAIEFF models. This ultimately opens the way for significant improvements in the accuracy achieved by the HAIEFF models.

Project description:Objective: To explore the diagnostic value of CT radiographic images and radiomics features for invasive classification of lung adenocarcinoma manifesting as ground-glass nodules (GGNs) in computer tomography (CT). Methods: A total of 312 GGNs were enrolled in this retrospective study. All GGNs were randomly divided into training set (n = 219) and test set (n = 93). Univariate and multivariate logistic regressions were used to establish a clinical model, while the minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) algorithm were used to select the radiomics features and construct the radiomics model. A combined model was finally built by combining these two models. The performance of these models was assessed in both training and test set. A combined nomogram was developed based on the combined model and evaluated with its calibration curves and C-index. Results: Diameter [odds ratio (OR), 1.159; p ＜ 0.001], lobulation (OR, 2.953; p = 0.002), and vascular changes (OR, 3.431; p ＜ 0.001) were retained as independent predictors of the invasive adenocarcinoma (IAC) group. Eleven radiomics features were selected by mRMR and LASSO method to established radiomics model. The clinical model and radiomics mode showed good predictive ability in both training set and test set. When two models were combined, the diagnostic area under the curve (AUC) value was higher than the single clinical or radiomics model (training set: 0.86 vs. 0.83 vs. 0.82; test set: 0.80 vs. 0.78 vs. 0.79). The constructed combined nomogram could effectively quantify the risk degree of 3 image features and Rad score with a C-index of 0.855 (95%: 0.805∼0.905). Conclusion: Radiographic and radiomics features show high accuracy in the invasive diagnosis of GGNs, and their combined analysis can improve the diagnostic efficacy of IAC manifesting as GGNs. The nomogram, serving as a noninvasive and accurate predictive tool, can help judge the invasiveness of GGNs prior to surgery and assist clinicians in creating personalized treatment strategies.

Project description:BackgroundAdvancements in transcriptomic profiling have led to the emergence of new challenges regarding data integration and interpretability. Variability between measurement platforms makes it difficult to compare between cohorts, and large numbers of gene features have encouraged the use black box methods that are not easily translated into biologically and clinically meaningful findings. We propose that gene rankings and algorithms that rely on relative expression within gene pairs can address such obstacles.MethodsWe implemented an innovative process to evaluate the performance of five feature selection methods on simulated gene-pair data. Along with TSP, we consider other methods that retain more information in their score calculations, including the magnitude of gene expression change as well as within-class variation. Tree-based rule extraction was also applied to serum microRNA (miRNA) pairs in order to devise a noninvasive screening tool for pancreatic and ovarian cancer.ResultsGene pair data were simulated using different types of signal and noise. Pairs were filtered using feature selection approaches, including top-scoring pairs (TSP), absolute differences between gene ranks, and Fisher scores. Methods that retain more information, such as the magnitude of expression change and within-class variance, yielded higher classification accuracy using a random forest model. We then demonstrate two powerful applications of gene pairs by first performing large-scale integration of 52 breast cancer datasets consisting of 10,350 patients. Not only did we confirm known oncogenes, but we also propose novel tumorigenic genes, such as BSDC1 and U2AF1, that could distinguish between tumor subtypes. Finally, circulating miRNA pairs were filtered and salient rules were extracted to build simplified tree ensemble learners (STELs) for four types of cancer. These accessible clinical frameworks detected pancreatic and ovarian cancer with 84.8 and 93.6% accuracy, respectively.ConclusionRank-based gene pair classification benefits from careful feature selection methods that preserve maximal information. Gene pairs enable dataset integration for greater statistical power and discovery of robust biomarkers as well as facilitate construction of user-friendly clinical screening tools.

Dataset Information

CIRDataset: A large-scale Dataset for Clinically-Interpretable lung nodule Radiomics and malignancy prediction.

Publications

CIRDataset: A large-scale Dataset for Clinically-Interpretable lung nodule Radiomics and malignancy prediction.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets