Robust imaging habitat computation using voxel-wise radiomics features

MRI-based habitat imaging in cancer treatment: current technology, applications, and challenges

Extensive efforts have been dedicated to exploring the impact of tumor heterogeneity on cancer treatment at both histological and genetic levels. To accurately measure intra-tumoral heterogeneity, a non-invasive imaging technique, known as habitat imaging, was developed. The technique quantifies intra-tumoral heterogeneity by dividing complex tumors into distinct sub- regions, called habitats. This article reviews the following aspects of habitat imaging in cancer treatment, with a focus on radiotherapy: (1) Habitat imaging biomarkers for assessing tumor physiology; (2) Methods for habitat generation; (3) Efforts to combine radiomics, another imaging quantification method, with habitat imaging; (4) Technical challenges and potential solutions related to habitat imaging; (5) Pathological validation of habitat imaging and how it can be utilized to evaluate cancer treatment by predicting treatment response including survival rate, recurrence, and pathological response as well as ongoing open clinical trials.

Habitat radiomics and deep learning fusion nomogram to predict EGFR mutation status in stage I non-small cell lung cancer: a multicenter study

Develop a radiomics nomogram that integrates deep learning, radiomics, and clinical variables to predict epidermal growth factor receptor (EGFR) mutation status in patients with stage I non-small cell lung cancer (NSCLC). We retrospectively included 438 patients who underwent curative surgery and completed driver-gene mutation tests for stage I NSCLC from four academic medical centers. Predictive models were established by extracting and analyzing radiomic features in intratumoral, peritumoral, and habitat regions of CT images to identify EGFR mutation status in stage I NSCLC. Additionally, three deep learning models based on the intratumoral region were constructed. A nomogram was developed by integrating representative radiomic signatures, deep learning, and clinical features. Model performance was assessed by calculating the area under the receiver operating characteristic (ROC) curve. The established habitat radiomics features demonstrated encouraging performance in discriminating between EGFR mutant and wild-type, with predictive ability superior to other single models (AUC 0.886, 0.812, and 0.790 for the training, validation, and external test sets, respectively). The radiomics-based nomogram exhibited excellent performance, achieving the highest AUC values of 0.917, 0.837, and 0.809 in the training, validation, and external test sets, respectively. Decision curve analysis (DCA) indicated that the nomogram provided a higher net benefit than other radiomics models, offering valuable information for treatment.

Identification of Precise 3D CT Radiomics for Habitat Computation by Machine Learning in Cancer

Purpose To identify precise three-dimensional radiomics features in CT images that enable computation of stable and biologically meaningful habitats with machine learning for cancer heterogeneity assessment. Materials and Methods This retrospective study included 2436 liver or lung lesions from 605 CT scans (November 2010-December 2021) in 331 patients with cancer (mean age, 64.5 years ± 10.1 [SD]; 185 male patients). Three-dimensional radiomics were computed from original and perturbed (simulated retest) images with different combinations of feature computation kernel radius and bin size. The lower 95% confidence limit (LCL) of the intraclass correlation coefficient (ICC) was used to measure repeatability and reproducibility. Precise features were identified by combining repeatability and reproducibility results (LCL of ICC ≥ 0.50). Habitats were obtained with Gaussian mixture models in original and perturbed data using precise radiomics features and compared with habitats obtained using all features. The Dice similarity coefficient (DSC) was used to assess habitat stability. Biologic correlates of CT habitats were explored in a case study, with a cohort of 13 patients with CT, multiparametric MRI, and tumor biopsies. Results Three-dimensional radiomics showed poor repeatability (LCL of ICC: median [IQR], 0.442 [0.312-0.516]) and poor reproducibility against kernel radius (LCL of ICC: median [IQR], 0.440 [0.33-0.526]) but excellent reproducibility against bin size (LCL of ICC: median [IQR], 0.929 [0.853-0.988]). Twenty-six radiomics features were precise, differing in lung and liver lesions. Habitats obtained with precise features (DSC: median [IQR], 0.601 [0.494-0.712] and 0.651 [0.52-0.784] for lung and liver lesions, respectively) were more stable than those obtained with all features (DSC: median [IQR], 0.532 [0.424-0.637] and 0.587 [0.465-0.703] for lung and liver lesions, respectively; P < .001). In the case study, CT habitats correlated quantitatively and qualitatively with heterogeneity observed in multiparametric MRI habitats and histology. Conclusion Precise three-dimensional radiomics features were identified on CT images that enabled tumor heterogeneity assessment through stable tumor habitat computation. Keywords: CT, Diffusion-weighted Imaging, Dynamic Contrast-enhanced MRI, MRI, Radiomics, Unsupervised Learning, Oncology, Liver, Lung Supplemental material is available for this article. © RSNA, 2024 See also the commentary by Sagreiya in this issue.

Clinico-biological-radiomics (CBR) based machine learning for improving the diagnostic accuracy of FDG-PET false-positive lymph nodes in lung cancer

BACKGROUND

The main problem of positron emission tomography/computed tomography (PET/CT) for lymph node (LN) staging is the high false positive rate (FPR). Thus, we aimed to explore a clinico-biological-radiomics (CBR) model via machine learning (ML) to reduce FPR and improve the accuracy for predicting the hypermetabolic mediastinal-hilar LNs status in lung cancer than conventional PET/CT.

METHODS

A total of 260 lung cancer patients with hypermetabolic mediastinal-hilar LNs (SUVmax ≥ 2.5) were retrospectively reviewed. Patients were treated with surgery with systematic LN resection and pathologically divided into the LN negative (LN-) and positive (LN +) groups, and randomly assigned into the training (n = 182) and test (n = 78) sets. Preoperative CBR dataset containing 1738 multi-scale features was constructed for all patients. Prediction models for hypermetabolic LNs status were developed using the features selected by the supervised ML algorithms, and evaluated using the classical diagnostic indicators. Then, a nomogram was developed based on the model with the highest area under the curve (AUC) and the lowest FPR, and validated by the calibration plots.

RESULTS

In total, 109 LN- and 151 LN + patients were enrolled in this study. 6 independent prediction models were developed to differentiate LN- from LN + patients using the selected features from clinico-biological-image dataset, radiomics dataset, and their combined CBR dataset, respectively. The DeLong test showed that the CBR Model containing all-scale features held the highest predictive efficiency and the lowest FPR among all of established models (p < 0.05) in both the training and test sets (AUCs of 0.90 and 0.89, FPRs of 12.82% and 6.45%, respectively) (p < 0.05). The quantitative nomogram based on CBR Model was validated to have a good consistency with actual observations.

CONCLUSION

This study presents an integrated CBR nomogram that can further reduce the FPR and improve the accuracy of hypermetabolic mediastinal-hilar LNs evaluation than conventional PET/CT in lung cancer, thereby greatly reducing the risk of overestimation and assisting for precision treatment.

Technical note: Intensity-based quality assurance criteria for deformable image registration in image-guided radiotherapy

BACKGROUND

Deformable image registration is increasingly used in radiotherapy to adapt the treatment plan and accumulate the delivered dose. Consequently, clinical workflows using deformable image registration require quick and reliable quality assurance to accept registrations. Additionally, for online adaptive radiotherapy, quality assurance without the need for an operator to delineate contours while the patient is on the treatment table is needed. Established quality assurance criteria such as the Dice similarity coefficient or Hausdorff distance lack these qualities and also display a limited sensitivity to registration errors beyond soft tissue boundaries.

PURPOSE

The purpose of this study is to investigate the existing intensity-based quality assurance criteria structural similarity and normalized mutual information for their ability to quickly and reliably identify registration errors for (online) adaptive radiotherapy and compare them to contour-based quality assurance criteria.

METHODS

All criteria were tested using synthetic and simulated biomechanical deformations of 3D MR images as well as manually annotated 4D CT data. The quality assurance criteria were scored for classification performance, for their ability to predict the registration error, and for their spatial information.

RESULTS

We found that besides being fast and operator-independent, the intensity-based criteria have the highest area under the receiver operating characteristic curve and provide the best input for models to predict the registration error on all data sets. Structural similarity furthermore provides spatial information with a higher gamma pass rate of the predicted registration error than commonly used spatial quality assurance criteria.

CONCLUSIONS

Intensity-based quality assurance criteria can provide the required confidence in decisions about using mono-modal registrations in clinical workflows. They thereby enable automated quality assurance for deformable image registration in adaptive radiotherapy treatments.

Assessment of brain cancer atlas maps with multimodal imaging features

BACKGROUND

Glioblastoma Multiforme (GBM) is a fast-growing and highly aggressive brain tumor that invades the nearby brain tissue and presents secondary nodular lesions across the whole brain but generally does not spread to distant organs. Without treatment, GBM can result in death in about 6 months. The challenges are known to depend on multiple factors: brain localization, resistance to conventional therapy, disrupted tumor blood supply inhibiting effective drug delivery, complications from peritumoral edema, intracranial hypertension, seizures, and neurotoxicity.

MAIN TEXT

Imaging techniques are routinely used to obtain accurate detections of lesions that localize brain tumors. Especially magnetic resonance imaging (MRI) delivers multimodal images both before and after the administration of contrast, which results in displaying enhancement and describing physiological features as hemodynamic processes. This review considers one possible extension of the use of radiomics in GBM studies, one that recalibrates the analysis of targeted segmentations to the whole organ scale. After identifying critical areas of research, the focus is on illustrating the potential utility of an integrated approach with multimodal imaging, radiomic data processing and brain atlases as the main components. The templates associated with the outcome of straightforward analyses represent promising inference tools able to spatio-temporally inform on the GBM evolution while being generalizable also to other cancers.

CONCLUSIONS

The focus on novel inference strategies applicable to complex cancer systems and based on building radiomic models from multimodal imaging data can be well supported by machine learning and other computational tools potentially able to translate suitably processed information into more accurate patient stratifications and evaluations of treatment efficacy.

Clinical features combined with ultrasound-based radiomics nomogram for discrimination between benign and malignant lesions in ultrasound suspected supraclavicular lymphadenectasis

Background

Conventional ultrasound (CUS) is the first choice for discrimination benign and malignant lymphadenectasis in supraclavicular lymph nodes (SCLNs), which is important for the further treatment. Radiomics provide more comprehensive and richer information than radiographic images, which are imperceptible to human eyes.

Objective

This study aimed to explore the clinical value of CUS-based radiomics analysis in preoperative differentiation of malignant from benign lymphadenectasis in CUS suspected SCLNs.

Methods

The characteristics of CUS images of 189 SCLNs were retrospectively analyzed, including 139 pathologically confirmed benign SCLNs and 50 malignant SCLNs. The data were randomly divided (7:3) into a training set (n=131) and a validation set (n=58). A total of 744 radiomics features were extracted from CUS images, radiomics score (Rad-score) built were using least absolute shrinkage and selection operator (LASSO) logistic regression. Rad-score model, CUS model, radiomics-CUS (Rad-score + CUS) model, clinic-radiomics (Clin + Rad-score) model, and combined CUS-clinic-radiomics (Clin + CUS + Rad-score) model were built using logistic regression. Diagnostic accuracy was assessed by receiver operating characteristic (ROC) curve analysis.

Results

A total of 20 radiomics features were selected from 744 radiomics features and calculated to construct Rad-score. The AUCs of Rad-score model, CUS model, Clin + Rad-score model, Rad-score + CUS model, and Clin + CUS + Rad-score model were 0.80, 0.72, 0.85, 0.83, 0.86 in the training set and 0.77, 0.80, 0.82, 0.81, 0.85 in the validation set. There was no statistical significance among the AUC of all models in the training and validation set. The calibration curve also indicated the good predictive performance of the proposed nomogram.

Conclusions

The Rad-score model, derived from supraclavicular ultrasound images, showed good predictive effect in differentiating benign from malignant lesions in patients with suspected supraclavicular lymphadenectasis.

Tumor cellularity beyond the visible in soft tissue sarcomas: Results of an ADC-based, single center, and preliminary radiomics study

Purpose

Soft tissue sarcomas represent approximately 1% of all malignancies, and diagnostic radiology plays a significant role in the overall management of this rare group of tumors. Recently, quantitative imaging and, in particular, radiomics demonstrated to provide significant novel information, for instance, in terms of prognosis and grading. The aim of this study was to evaluate the prognostic role of radiomic variables extracted from apparent diffusion coefficient (ADC) maps collected at diagnosis in patients with soft tissue sarcomas in terms of overall survival and metastatic spread as well as to assess the relationship between radiomics and the tumor grade.

Methods

Patients with histologically proven soft tissue sarcomas treated in our tertiary center from 2016 to 2019 who underwent an Magnetic Resonance (MR) scan at diagnosis including diffusion-weighted imaging were included in this retrospective institution review board–approved study. Each primary lesion was segmented using the b50 images; the volumetric region of interest was then applied on the ADC map. A total of 33 radiomic features were extracted, and highly correlating features were selected by factor analysis. In the case of feature/s showing statistically significant results, the diagnostic accuracy was computed. The Spearman correlation coefficient was used to evaluate the relationship between the tumor grade and radiomic features selected by factor analysis. All analyses were performed applying p&lt;0.05 as a significant level.

Results

A total of 36 patients matched the inclusion criteria (15 women; mean age 58.9 ± 15 years old). The most frequent histotype was myxofibrosarcoma (16.6%), and most of the patients were affected by high-grade lesions (77.7%). Seven patients had pulmonary metastases, and, altogether, eight were deceased. Only the feature Imc1 turned out to be a predictor of metastatic spread (p=0.045 after Bonferroni correction) with 76.7% accuracy. The value -0.16 showed 73.3% sensitivity and 71.4% specificity, and patients with metastases showed lower values (mean Imc1 of metastatic patients -0.31). None of the examined variables was a predictor of the overall outcome (p&gt;0.05, each). A moderate statistically significant correlation emerged only between Imc1 and the tumor grade (r=0.457, p=0.005).

Conclusions

In conclusion, the radiomic feature Imc1 acts as a predictor of metastatic spread in patients with soft tissue sarcomas and correlates with the tumor grade.

Radiomic Signatures Associated with CD8+ Tumour-Infiltrating Lymphocytes: A Systematic Review and Quality Assessment Study

The tumour immune microenvironment influences the efficacy of immune checkpoint inhibitors. Within this microenvironment are CD8-expressing tumour-infiltrating lymphocytes (CD8+ TILs), which are an important mediator and marker of anti-tumour response. In practice, the assessment of CD8+ TILs via tissue sampling involves logistical challenges. Radiomics, the high-throughput extraction of features from medical images, may offer a novel and non-invasive alternative. We performed a systematic review of the available literature reporting radiomic signatures associated with CD8+ TILs. We also aimed to evaluate the methodological quality of the identified studies using the Radiomics Quality Score (RQS) tool, and the risk of bias and applicability with the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. Articles were searched from inception until 31 December 2021, in three electronic databases, and screened against eligibility criteria. Twenty-seven articles were included. A wide variety of cancers have been studied. The reported radiomic signatures were heterogeneous, with very limited reproducibility between studies of the same cancer group. The overall quality of studies was found to be less than desirable (mean RQS = 33.3%), indicating a need for technical maturation. Some potential avenues for further investigation are also discussed.

Voxel-wise supervised analysis of tumors with multimodal engineered features to highlight interpretable biological patterns

BACKGROUND

Translation of predictive and prognostic image-based learning models to clinical applications is challenging due in part to their lack of interpretability. Some deep-learning-based methods provide information about the regions driving the model output. Yet, due to the high-level abstraction of deep features, these methods do not completely solve the interpretation challenge. In addition, low sample size cohorts can lead to instabilities and suboptimal convergence for models involving a large number of parameters such as convolutional neural networks.

PURPOSE

Here, we propose a method for designing radiomic models that combines the interpretability of handcrafted radiomics with a sub-regional analysis.

MATERIALS AND METHODS

Our approach relies on voxel-wise engineered radiomic features with average global aggregation and logistic regression. The method is illustrated using a small dataset of 51 soft tissue sarcoma (STS) patients where the task is to predict the risk of lung metastasis occurrence during the follow-up period.

RESULTS

Using positron emission tomography/computed tomography and two magnetic resonance imaging sequences separately to build two radiomic models, we show that our approach produces quantitative maps that highlight the signal that contributes to the decision within the tumor region of interest. In our STS example, the analysis of these maps identified two biological patterns that are consistent with STS grading systems and knowledge: necrosis development and glucose metabolism of the tumor.

CONCLUSIONS

We demonstrate how that method makes it possible to spatially and quantitatively interpret radiomic models amenable to sub-regions identification and biological interpretation for patient stratification.

An Intelligent Radiomic Approach for Lung Cancer Screening

The efficiency of lung cancer screening for reducing mortality is hindered by the high rate of false positives. Artificial intelligence applied to radiomics could help to early discard benign cases from the analysis of CT scans. The available amount of data and the fact that benign cases are a minority, constitutes a main challenge for the successful use of state of the art methods (like deep learning), which can be biased, over-fitted and lack of clinical reproducibility. We present an hybrid approach combining the potential of radiomic features to characterize nodules in CT scans and the generalization of the feed forward networks. In order to obtain maximal reproducibility with minimal training data, we propose an embedding of nodules based on the statistical significance of radiomic features for malignancy detection. This representation space of lesions is the input to a feed forward network, which architecture and hyperparameters are optimized using own-defined metrics of the diagnostic power of the whole system. Results of the best model on an independent set of patients achieve 100% of sensitivity and 83% of specificity (AUC = 0.94) for malignancy detection.