Validated machine learning tools to distinguish immune checkpoint inhibitor, radiotherapy, COVID-19 and other infective pneumonitis.

BACKGROUND: Pneumonitis is a well-described, potentially disabling, or fatal adverse effect associated with both immune checkpoint inhibitors (ICI) and thoracic radiotherapy. Accurate differentiation between checkpoint inhibitor pneumonitis (CIP) radiation pneumonitis (RP), and infective pneumonitis (IP) is crucial for swift, appropriate, and tailored management to achieve optimal patient outcomes. However, correct diagnosis is often challenging, owing to overlapping clinical presentations and radiological patterns. METHODS: In this multi-centre study of 455 patients, we used machine learning with radiomic features extracted from chest CT imaging to develop and validate five models to distinguish CIP and RP from COVID-19, non-COVID-19 infective pneumonitis, and each other. Model performance was compared to that of two radiologists. RESULTS: Models to distinguish RP from COVID-19, CIP from COVID-19 and CIP from non-COVID-19 IP out-performed radiologists (test set AUCs of 0.92 vs 0.8 and 0.8; 0.68 vs 0.43 and 0.4; 0.71 vs 0.55 and 0.63 respectively). Models to distinguish RP from non-COVID-19 IP and CIP from RP were not superior to radiologists but demonstrated modest performance, with test set AUCs of 0.81 and 0.8 respectively. The CIP vs RP model performed less well on patients with prior exposure to both ICI and radiotherapy (AUC 0.54), though the radiologists also had difficulty distinguishing this test cohort (AUC values 0.6 and 0.6). CONCLUSION: Our results demonstrate the potential utility of such tools as a second or concurrent reader to support oncologists, radiologists, and chest physicians in cases of diagnostic uncertainty. Further research is required for patients with exposure to both ICI and thoracic radiotherapy.

Clinical application of machine learning models in patients with prostate cancer before prostatectomy.

BACKGROUND: To build machine learning predictive models for surgical risk assessment of extracapsular extension (ECE) in patients with prostate cancer (PCa) before radical prostatectomy; and to compare the use of decision curve analysis (DCA) and receiver operating characteristic (ROC) metrics for selecting input feature combinations in models. METHODS: This retrospective observational study included two independent data sets: 139 participants from a single institution (training), and 55 from 15 other institutions (external validation), both treated with Robotic Assisted Radical Prostatectomy (RARP). Five ML models, based on different combinations of clinical, semantic (interpreted by a radiologist) and radiomics features computed from T2W-MRI images, were built to predict extracapsular extension in the prostatectomy specimen (pECE+). DCA plots were used to rank the models' net benefit when assigning patients to prostatectomy with non-nerve-sparing surgery (NNSS) or nerve-sparing surgery (NSS), depending on the predicted ECE status. DCA model rankings were compared with those drived from ROC area under the curve (AUC). RESULTS: In the training data, the model using clinical, semantic, and radiomics features gave the highest net benefit values across relevant threshold probabilities, and similar decision curve was observed in the external validation data. The model ranking using the AUC was different in the discovery group and favoured the model using clinical + semantic features only. CONCLUSIONS: The combined model based on clinical, semantic and radiomic features may be used to predict pECE + in patients with PCa and results in a positive net benefit when used to choose between prostatectomy with NNS or NNSS.

PROMISE Registry: A prostate cancer registry of outcomes and germline mutations for improved survival and treatment effectiveness.

BACKGROUND: Recently approved treatments and updates to genetic testing recommendations for prostate cancer have created a need for correlated analyses of patient outcomes data via germline genetic mutation status. Genetic registries address these gaps by identifying candidates for recently approved targeted treatments, expanding clinical trial data examining specific gene mutations, and understanding effects of targeted treatments in the real-world setting. METHODS: The PROMISE Registry is a 20-year (5-year recruitment, 15-year follow-up), US-wide, prospective genetic registry for prostate cancer patients. Five thousand patients will be screened through an online at-home germline testing to identify and enroll 500 patients with germline mutations, including: pathogenic or likely pathogenic variants and variants of uncertain significance in genes of interest. Patients will be followed for 15 years and clinical data with real time patient reported outcomes will be collected. Eligible patients will enter long-term follow-up (6-month PRO surveys and medical record retrieval). As a virtual study with patient self-enrollment, the PROMISE Registry may fill gaps in genetics services in underserved areas and for patients within sufficient insurance coverage. RESULTS: The PROMISE Registry opened in May 2021. 2114 patients have enrolled to date across 48 US states and 23 recruiting sites. 202 patients have met criteria for long-term follow-up. PROMISE is on target with the study's goal of 5000 patients screened and 500 patients eligible for long-term follow-up by 2026. CONCLUSIONS: The PROMISE Registry is a novel, prospective, germline registry that will collect long-term patient outcomes data to address current gaps in understanding resulting from recently FDA-approved treatments and updates to genetic testing recommendations for prostate cancer. Through inclusion of a broad nationwide sample, including underserved patients and those unaffiliated with major academic centers, the PROMISE Registry aims to provide access to germline genetic testing and to collect data to understand disease characteristics and treatment responses across the disease spectrum for prostate cancer with rare germline genetic variants.

Evaluation of simultaneous multi-slice acquisition with advanced processing for free-breathing diffusion-weighted imaging in patients with liver metastasis.

OBJECTIVES: Diffusion-weighted imaging (DWI) with simultaneous multi-slice (SMS) acquisition and advanced processing can accelerate acquisition time and improve MR image quality. This study evaluated the image quality and apparent diffusion coefficient (ADC) measurements of free-breathing DWI acquired from patients with liver metastases using a prototype SMS-DWI acquisition (with/without an advanced processing option) and conventional DWI. METHODS: Four DWI schemes were compared in a pilot 5-patient cohort; three DWI schemes were further assessed in a 24-patient cohort. Two readers scored image quality of all b-value images and ADC maps across the three methods. ADC measurements were performed, for all three methods, in left and right liver parenchyma, spleen, and liver metastases. The Friedman non-parametric test (post-hoc Wilcoxon test with Bonferroni correction) was used to compare image quality scoring; t-test was used for ADC comparisons. RESULTS: SMS-DWI was faster (by 24%) than conventional DWI. Both readers scored the SMS-DWI with advanced processing as having the best image quality for highest b-value images (b750) and ADC maps; Cohen's kappa inter-reader agreement was 0.6 for b750 image and 0.56 for ADC maps. The prototype SMS-DWI sequence with advanced processing allowed a better visualization of the left lobe of the liver. ADC measured in liver parenchyma, spleen, and liver metastases using the SMS-DWI with advanced processing option showed lower values than those derived from the SMS-DWI method alone (t-test, p < 0.0001; p < 0.0001; p = 0.002). CONCLUSIONS: Free-breathing SMS-DWI with advanced processing was faster and demonstrated better image quality versus a conventional DWI protocol in liver patients. CLINICAL RELEVANCE STATEMENT: Free-breathing simultaneous multi-slice- diffusion-weighted imaging (DWI) with advanced processing was faster and demonstrated better image quality versus a conventional DWI protocol in liver patients. KEY POINTS: • Diffusion-weighted imaging (DWI) with simultaneous multi-slice (SMS) can accelerate acquisition time and improve image quality. • Apparent diffusion coefficients (ADC) measured in liver parenchyma, spleen, and liver metastases using the simultaneous multi-slice DWI with advanced processing were significantly lower than those derived from the simultaneous multi-slice DWI method alone. • Simultaneous multi-slice DWI sequence with inline advanced processing was faster and demonstrated better image quality in liver patients.

A CT-based radiomics classification model for the prediction of histological type and tumour grade in retroperitoneal sarcoma (RADSARC-R): a retrospective multicohort analysis.

BACKGROUND: Retroperitoneal sarcomas are tumours with a poor prognosis. Upfront characterisation of the tumour is difficult, and under-grading is common. Radiomics has the potential to non-invasively characterise the so-called radiological phenotype of tumours. We aimed to develop and independently validate a CT-based radiomics classification model for the prediction of histological type and grade in retroperitoneal leiomyosarcoma and liposarcoma. METHODS: A retrospective discovery cohort was collated at our centre (Royal Marsden Hospital, London, UK) and an independent validation cohort comprising patients recruited in the phase 3 STRASS study of neoadjuvant radiotherapy in retroperitoneal sarcoma. Patients aged older than 18 years with confirmed primary leiomyosarcoma or liposarcoma proceeding to surgical resection with available contrast-enhanced CT scans were included. Using the discovery dataset, a CT-based radiomics workflow was developed, including manual delineation, sub-segmentation, feature extraction, and predictive model building. Separate probabilistic classifiers for the prediction of histological type and low versus intermediate or high grade tumour types were built and tested. Independent validation was then performed. The primary objective of the study was to develop radiomic classification models for the prediction of retroperitoneal leiomyosarcoma and liposarcoma type and histological grade. FINDINGS: 170 patients recruited between Oct 30, 2016, and Dec 23, 2020, were eligible in the discovery cohort and 89 patients recruited between Jan 18, 2012, and April 10, 2017, were eligible in the validation cohort. In the discovery cohort, the median age was 63 years (range 27-89), with 83 (49%) female and 87 (51%) male patients. In the validation cohort, median age was 59 years (range 33-77), with 46 (52%) female and 43 (48%) male patients. The highest performing model for the prediction of histological type had an area under the receiver operator curve (AUROC) of 0·928 on validation, based on a feature set of radiomics and approximate radiomic volume fraction. The highest performing model for the prediction of histological grade had an AUROC of 0·882 on validation, based on a radiomics feature set. INTERPRETATION: Our validated radiomics model can predict the histological type and grade of retroperitoneal sarcomas with excellent performance. This could have important implications for improving diagnosis and risk stratification in retroperitoneal sarcomas. FUNDING: Wellcome Trust, European Organisation for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group, the National Institutes for Health, and the National Institute for Health and Care Research Biomedical Research Centre at The Royal Marsden NHS Foundation Trust and The Institute of Cancer Research.

Interpretability of radiomics models is improved when using feature group selection strategies for predicting molecular and clinical targets in clear-cell renal cell carcinoma: insights from the TRACERx Renal study.

BACKGROUND: The aim of this work is to evaluate the performance of radiomics predictions for a range of molecular, genomic and clinical targets in patients with clear cell renal cell carcinoma (ccRCC) and demonstrate the impact of novel feature selection strategies and sub-segmentations on model interpretability. METHODS: Contrast-enhanced CT scans from the first 101 patients recruited to the TRACERx Renal Cancer study (NCT03226886) were used to derive radiomics classification models to predict 20 molecular, histopathology and clinical target variables. Manual 3D segmentation was used in conjunction with automatic sub-segmentation to generate radiomics features from the core, rim, high and low enhancing sub-regions, and the whole tumour. Comparisons were made between two classification model pipelines: a Conventional pipeline reflecting common radiomics practice, and a Proposed pipeline including two novel feature selection steps designed to improve model interpretability. For both pipelines nested cross-validation was used to estimate prediction performance and tune model hyper-parameters, and permutation testing was used to evaluate the statistical significance of the estimated performance measures. Further model robustness assessments were conducted by evaluating model variability across the cross-validation folds. RESULTS: Classification performance was significant (p < 0.05, H0:AUROC = 0.5) for 11 of 20 targets using either pipeline and for these targets the AUROCs were within ± 0.05 for the two pipelines, except for one target where the Proposed pipeline performance increased by > 0.1. Five of these targets (necrosis on histology, presence of renal vein invasion, overall histological stage, linear evolutionary subtype and loss of 9p21.3 somatic alteration marker) had AUROC > 0.8. Models derived using the Proposed pipeline contained fewer feature groups than the Conventional pipeline, leading to more straightforward model interpretations without loss of performance. Sub-segmentations lead to improved performance and/or improved interpretability when predicting the presence of sarcomatoid differentiation and tumour stage. CONCLUSIONS: Use of the Proposed pipeline, which includes the novel feature selection methods, leads to more interpretable models without compromising prediction performance. TRIAL REGISTRATION: NCT03226886 (TRACERx Renal).

Radiomic applications in upper gastrointestinal cancer surgery.

INTRODUCTION: Cross-sectional imaging plays an integral role in the management of upper gastrointestinal (UGI) cancer, from initial diagnosis and staging to determining appropriate treatment strategies. Subjective imaging interpretation has known limitations. The field of radiomics has evolved to extract quantitative data from medical imaging and relate these to biological processes. The key concept behind radiomics is that the high-throughput analysis of quantitative imaging features can provide predictive or prognostic information, with the goal of providing individualised care. OBJECTIVE: Radiomic studies have shown promising utility in upper gastrointestinal oncology, highlighting a potential role in determining stage of disease and degree of tumour differentiation and predicting recurrence-free survival. This narrative review aims to provide an insight into the concepts underpinning radiomics, as well as its potential applications for guiding treatment and surgical decision-making in upper gastrointestinal malignancy. CONCLUSION: Outcomes from studies to date have been promising; however, further standardisation and collaboration are required. Large prospective studies with external validation and evaluation of radiomic integration into clinical pathways are needed. Future research should now focus on translating the promising utility of radiomics into meaningful patient outcomes.

Phase II Randomized Study of Salvage Radiation Therapy Plus Enzalutamide or Placebo for High-Risk Prostate-Specific Antigen Recurrent Prostate Cancer After Radical Prostatectomy: The SALV-ENZA Trial.

PURPOSE: We sought to investigate whether enzalutamide (ENZA), without concurrent androgen deprivation therapy, increases freedom from prostate-specific antigen (PSA) progression (FFPP) when combined with salvage radiation therapy (SRT) in men with recurrent prostate cancer after radical prostatectomy (RP). PATIENTS AND METHODS: Men with biochemically recurrent prostate cancer after RP were enrolled into a randomized, double-blind, phase II, placebo-controlled, multicenter study of SRT plus ENZA or placebo (ClinicalTrials.gov identifier: NCT02203695). Random assignment (1:1) was stratified by center, surgical margin status (R0 v R1), PSA before salvage treatment (PSA ≥ 0.5 v < 0.5 ng/mL), and pathologic Gleason sum (7 v 8-10). Patients were assigned to receive either ENZA 160 mg once daily or matching placebo for 6 months. After 2 months of study drug therapy, external-beam radiation (66.6-70.2 Gy) was administered to the prostate bed (no pelvic nodes). The primary end point was FFPP in the intention-to-treat population. Secondary end points were time to local recurrence within the radiation field, metastasis-free survival, and safety as determined by frequency and severity of adverse events. RESULTS: Eighty-six (86) patients were randomly assigned, with a median follow-up of 34 (range, 0-52) months. Trial arms were well balanced. The median pre-SRT PSA was 0.3 (range, 0.06-4.6) ng/mL, 56 of 86 patients (65%) had extraprostatic disease (pT3), 39 of 86 (45%) had a Gleason sum of 8-10, and 43 of 86 (50%) had positive surgical margins (R1). FFPP was significantly improved with ENZA versus placebo (hazard ratio [HR], 0.42; 95% CI, 0.19 to 0.92; P = .031), and 2-year FFPP was 84% versus 66%, respectively. Subgroup analyses demonstrated differential benefit of ENZA in men with pT3 (HR, 0.22; 95% CI, 0.07 to 0.69) versus pT2 disease (HR, 1.54; 95% CI, 0.43 to 5.47; Pinteraction = .019) and R1 (HR, 0.14; 95% CI, 0.03 to 0.64) versus R0 disease (HR, 1.00; 95% CI, 0.36 to 2.76; Pinteraction = .023). There were insufficient secondary end point events for analysis. The most common adverse events were grade 1-2 fatigue (65% ENZA v 53% placebo) and urinary frequency (40% ENZA v 49% placebo). CONCLUSION: SRT plus ENZA monotherapy for 6 months in men with PSA-recurrent high-risk prostate cancer after RP is safe and delays PSA progression relative to SRT alone. The impact of ENZA on distant metastasis or survival is unknown at this time.

Gross tumour volume radiomics for prognostication of recurrence & death following radical radiotherapy for NSCLC.

Recurrence occurs in up to 36% of patients treated with curative-intent radiotherapy for NSCLC. Identifying patients at higher risk of recurrence for more intensive surveillance may facilitate the earlier introduction of the next line of treatment. We aimed to use radiotherapy planning CT scans to develop radiomic classification models that predict overall survival (OS), recurrence-free survival (RFS) and recurrence two years post-treatment for risk-stratification. A retrospective multi-centre study of >900 patients receiving curative-intent radiotherapy for stage I-III NSCLC was undertaken. Models using radiomic and/or clinical features were developed, compared with 10-fold cross-validation and an external test set, and benchmarked against TNM-stage. Respective validation and test set AUCs (with 95% confidence intervals) for the radiomic-only models were: (1) OS: 0.712 (0.592-0.832) and 0.685 (0.585-0.784), (2) RFS: 0.825 (0.733-0.916) and 0.750 (0.665-0.835), (3) Recurrence: 0.678 (0.554-0.801) and 0.673 (0.577-0.77). For the combined models: (1) OS: 0.702 (0.583-0.822) and 0.683 (0.586-0.78), (2) RFS: 0.805 (0.707-0.903) and 0·755 (0.672-0.838), (3) Recurrence: 0·637 (0.51-0.·765) and 0·738 (0.649-0.826). Kaplan-Meier analyses demonstrate OS and RFS difference of >300 and >400 days respectively between low and high-risk groups. We have developed validated and externally tested radiomic-based prediction models. Such models could be integrated into the routine radiotherapy workflow, thus informing a personalised surveillance strategy at the point of treatment. Our work lays the foundations for future prospective clinical trials for quantitative personalised risk-stratification for surveillance following curative-intent radiotherapy for NSCLC.

Radiomic Features From Diffusion-Weighted MRI of Retroperitoneal Soft-Tissue Sarcomas Are Repeatable and Exhibit Change After Radiotherapy.

Background: Size-based assessments are inaccurate indicators of tumor response in soft-tissue sarcoma (STS), motivating the requirement for new response imaging biomarkers for this rare and heterogeneous disease. In this study, we assess the test-retest repeatability of radiomic features from MR diffusion-weighted imaging (DWI) and derived maps of apparent diffusion coefficient (ADC) in retroperitoneal STS and compare baseline repeatability with changes in radiomic features following radiotherapy (RT). Materials and Methods: Thirty patients with retroperitoneal STS received an MR examination prior to treatment, of whom 23/30 were investigated in our repeatability analysis having received repeat baseline examinations and 14/30 patients were investigated in our post-treatment analysis having received an MR examination after completing pre-operative RT. One hundred and seven radiomic features were extracted from the full manually delineated tumor region using PyRadiomics. Test-retest repeatability was assessed using an intraclass correlation coefficient (baseline ICC), and post-radiotherapy variance analysis (post-RT-IMS) was used to compare the change in radiomic feature value to baseline repeatability. Results: For the ADC maps and DWI images, 101 and 102 features demonstrated good baseline repeatability (baseline ICC > 0.85), respectively. Forty-three and 2 features demonstrated both good baseline repeatability and a high post-RT-IMS (>0.85), respectively. Pearson correlation between the baseline ICC and post-RT-IMS was weak (0.432 and 0.133, respectively). Conclusions: The ADC-based radiomic analysis shows better test-retest repeatability compared with features derived from DWI images in STS, and some of these features are sensitive to post-treatment change. However, good repeatability at baseline does not imply sensitivity to post-treatment change.

Deep learning DCE-MRI parameter estimation: Application in pancreatic cancer.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is an MRI technique for quantifying perfusion that can be used in clinical applications for classification of tumours and other types of diseases. Conventionally, the non-linear least squares (NLLS) methods is used for tracer-kinetic modelling of DCE data. However, despite promising results, NLLS suffers from long processing times (minutes-hours) and noisy parameter maps due to the non-convexity of the cost function. In this work, we investigated physics-informed deep neural networks for estimating physiological parameters from DCE-MRI signal-curves. Three voxel-wise temporal frameworks (FCN, LSTM, GRU) and two spatio-temporal frameworks (CNN, U-Net) were investigated. The accuracy and precision of parameter estimation by the temporal frameworks were evaluated in simulations. All networks showed higher precision than the NLLS. Specifically, the GRU showed to decrease the random error on ve by a factor of 4.8 with respect to the NLLS for noise (SD) of 1/20. The accuracy was better for the prediction of the ve parameter in all networks compared to the NLLS. The GRU and LSTM worked with arbitrary acquisition lengths. The GRU was selected for in vivo evaluation and compared to the spatio-temporal frameworks in 28 patients with pancreatic cancer. All neural network approaches showed less noisy parameter maps than the NLLS. The GRU had better test-retest repeatability than the NLLS for all three parameters and was able to detect one additional patient with significant changes in DCE parameters post chemo-radiotherapy. Although the U-Net and CNN had even better test-retest characteristics than the GRU, and were able to detect even more responders, they also showed potential systematic errors in the parameter maps. Therefore, we advise using our GRU framework for analysing DCE data.

A comparison of machine learning methods for predicting recurrence and death after curative-intent radiotherapy for non-small cell lung cancer: Development and validation of multivariable clinical prediction models.

BACKGROUND: Surveillance is universally recommended for non-small cell lung cancer (NSCLC) patients treated with curative-intent radiotherapy. High-quality evidence to inform optimal surveillance strategies is lacking. Machine learning demonstrates promise in accurate outcome prediction for a variety of health conditions. The purpose of this study was to utilise readily available patient, tumour, and treatment data to develop, validate and externally test machine learning models for predicting recurrence, recurrence-free survival (RFS) and overall survival (OS) at 2 years from treatment. METHODS: A retrospective, multicentre study of patients receiving curative-intent radiotherapy for NSCLC was undertaken. A total of 657 patients from 5 hospitals were eligible for inclusion. Data pre-processing derived 34 features for predictive modelling. Combinations of 8 feature reduction methods and 10 machine learning classification algorithms were compared, producing risk-stratification models for predicting recurrence, RFS and OS. Models were compared with 10-fold cross validation and an external test set and benchmarked against TNM-stage and performance status. Youden Index was derived from validation set ROC curves to distinguish high and low risk groups and Kaplan-Meier analyses performed. FINDINGS: Median follow-up time was 852 days. Parameters were well matched across training-validation and external test sets: Mean age was 73 and 71 respectively, and recurrence, RFS and OS rates at 2 years were 43% vs 34%, 54% vs 47% and 54% vs 47% respectively. The respective validation and test set AUCs were as follows: 1) RFS: 0·682 (0·575-0·788) and 0·681 (0·597-0·766), 2) Recurrence: 0·687 (0·582-0·793) and 0·722 (0·635-0·81), and 3) OS: 0·759 (0·663-0·855) and 0·717 (0·634-0·8). Our models were superior to TNM stage and performance status in predicting recurrence and OS. INTERPRETATION: This robust and ready to use machine learning method, validated and externally tested, sets the stage for future clinical trials entailing quantitative personalised risk-stratification and surveillance following curative-intent radiotherapy for NSCLC. FUNDING: A full list of funding bodies that contributed to this study can be found in the Acknowledgements section.

Spatial patterns of tumour growth impact clonal diversification in a computational model and the TRACERx Renal study.

Genetic intra-tumour heterogeneity fuels clonal evolution, but our understanding of clinically relevant clonal dynamics remain limited. We investigated spatial and temporal features of clonal diversification in clear cell renal cell carcinoma through a combination of modelling and real tumour analysis. We observe that the mode of tumour growth, surface or volume, impacts the extent of subclonal diversification, enabling interpretation of clonal diversity in patient tumours. Specific patterns of proliferation and necrosis explain clonal expansion and emergence of parallel evolution and microdiversity in tumours. In silico time-course studies reveal the appearance of budding structures before detectable subclonal diversification. Intriguingly, we observe radiological evidence of budding structures in early-stage clear cell renal cell carcinoma, indicating that future clonal evolution may be predictable from imaging. Our findings offer a window into the temporal and spatial features of clinically relevant clonal evolution.

DCE-MRI is more sensitive than IVIM-DWI for assessing anti-angiogenic treatment-induced changes in colorectal liver metastases.

BACKGROUND: Diffusion weighted imaging (DWI) with intravoxel incoherent motion (IVIM) modelling can inform on tissue perfusion without exogenous contrast administration. Dynamic-contrast-enhanced (DCE) MRI can also characterise tissue perfusion, but requires a bolus injection of a Gadolinium-based contrast agent. This study compares the use of DCE-MRI and IVIM-DWI methods in assessing response to anti-angiogenic treatment in patients with colorectal liver metastases in a cohort with confirmed treatment response. METHODS: This prospective imaging study enrolled 25 participants with colorectal liver metastases to receive Regorafenib treatment. A target metastasis > 2 cm in each patient was imaged before and at 15 days after treatment on a 1.5T MR scanner using slice-matched IVIM-DWI and DCE-MRI protocols. MRI data were motion-corrected and tumour volumes of interest drawn on b=900 s/mm2 diffusion-weighted images were transferred to DCE-MRI data for further analysis. The median value of four IVIM-DWI parameters [diffusion coefficient D (10-3 mm2/s), perfusion fraction f (ml/ml), pseudodiffusion coefficient D* (10-3 mm2/s), and their product fD* (mm2/s)] and three DCE-MRI parameters [volume transfer constant Ktrans (min-1), enhancement fraction EF (%), and their product KEF (min-1)] were recorded at each visit, before and after treatment. Changes in pre- and post-treatment measurements of all MR parameters were assessed using Wilcoxon signed-rank tests (P<0.05 was considered significant). DCE-MRI and IVIM-DWI parameter correlations were evaluated with Spearman rank tests. Functional MR parameters were also compared against Response Evaluation Criteria In Solid Tumours v.1.1 (RECIST) evaluations. RESULTS: Significant treatment-induced reductions of DCE-MRI parameters across the cohort were observed for EF (91.2 to 50.8%, P<0.001), KEF (0.095 to 0.045 min-1, P<0.001) and Ktrans (0.109 to 0.078 min-1, P=0.002). For IVIM-DWI, only D (a non-perfusion parameter) increased significantly post treatment (0.83 to 0.97 × 10-3 mm2/s, P<0.001), while perfusion-related parameters showed no change. No strong correlations were found between DCE-MRI and IVIM-DWI parameters. A moderate correlation was found, after treatment, between Ktrans and D* (r=0.60; P=0.002) and fD* (r=0.67; P<0.001). When compared to RECIST v.1.1 evaluations, KEF and D correctly identified most clinical responders, whilst non-responders were incorrectly identified. CONCLUSION: IVIM-DWI perfusion-related parameters showed limited sensitivity to the anti-angiogenic effects of Regorafenib treatment in colorectal liver metastases and showed low correlation with DCE-MRI parameters, despite profound and significant post-treatment reductions in DCE-MRI measurements. TRIAL REGISTRATION: NCT03010722 clinicaltrials.gov; registration date 6th January 2015.

Radiomics in Oncology: A Practical Guide.

Radiomics refers to the extraction of mineable data from medical imaging and has been applied within oncology to improve diagnosis, prognostication, and clinical decision support, with the goal of delivering precision medicine. The authors provide a practical approach for successfully implementing a radiomic workflow from planning and conceptualization through manuscript writing. Applications in oncology typically are either classification tasks that involve computing the probability of a sample belonging to a category, such as benign versus malignant, or prediction of clinical events with a time-to-event analysis, such as overall survival. The radiomic workflow is multidisciplinary, involving radiologists and data and imaging scientists, and follows a stepwise process involving tumor segmentation, image preprocessing, feature extraction, model development, and validation. Images are curated and processed before segmentation, which can be performed on tumors, tumor subregions, or peritumoral zones. Extracted features typically describe the distribution of signal intensities and spatial relationship of pixels within a region of interest. To improve model performance and reduce overfitting, redundant and nonreproducible features are removed. Validation is essential to estimate model performance in new data and can be performed iteratively on samples of the dataset (cross-validation) or on a separate hold-out dataset by using internal or external data. A variety of noncommercial and commercial radiomic software applications can be used. Guidelines and artificial intelligence checklists are useful when planning and writing up radiomic studies. Although interest in the field continues to grow, radiologists should be familiar with potential pitfalls to ensure that meaningful conclusions can be drawn. Online supplemental material is available for this article. Published under a CC BY 4.0 license.

Descriptive analysis of MRI functional changes occurring during reduced dose radiotherapy for myxoid liposarcomas.

OBJECTIVES: Myxoid liposarcomas (MLS) show enhanced response to radiotherapy due to their distinctive vascular pattern and therefore could be effectively treated with lower radiation doses. This is a descriptive study to explore the use of functional MRI to identify response in a uniform cohort of MLS patients treated with reduced dose radiotherapy. METHODS: 10 patients with MLS were imaged pre-, during, and post-radiotherapy receiving reduced dose radiotherapy and the response to treatment was histopathologically assessed post-radiotherapy. Apparent diffusion coefficient (ADC), T2* relaxation time, volume transfer constant (Ktrans), initial area under the gadolinium curve over 60 s (IAUGC60) and (Gd) were estimated for a central tumour volume. RESULTS: All parameters showed large inter- and intrasubject variabilities. Pre-treatment (Gd), IAUGC60 and Ktrans were significantly different between responders and non-responders. Post-radiotherapy reductions from baseline were demonstrated for T2*, (Gd), IAUGC60 and Ktrans for responders. No statistically significant ADC differences were demonstrated between the two response groups. Significantly greater early tumour volume reductions were observed for responders. CONCLUSIONS: MLS are heterogenous lesions, characterised by a slow gradual contrast-agent uptake. Pre-treatment vascular parameters, early changes to tumour volume, vascular parameters and T2* have potential in identifying response to treatment. The delayed (Gd) is a suitable descriptive parameter, relying simply on T1 measurements. Volume changes precede changes in MLS functionality and could be used to identify early response. ADVANCES IN KNOWLEDGE: MLS are are characterised by slow gradual contrast-agent uptake. Measurement of the delayed contrast-agent uptake (Gd) is simple to implement and able to discriminate response.

"Real-world" radiomics from multi-vendor MRI: an original retrospective study on the prediction of nodal status and disease survival in breast cancer, as an exemplar to promote discussion of the wider issues.

BACKGROUND: Most MRI radiomics studies to date, even multi-centre ones, have used "pure" datasets deliberately accrued from single-vendor, single-field-strength scanners. This does not reflect aspirations for the ultimate generalisability of AI models. We therefore investigated the development of a radiomics signature from heterogeneous data originating on six different imaging platforms, for a breast cancer exemplar, in order to provide input into future discussions of the viability of radiomics in "real-world" scenarios where image data are not controlled by specific trial protocols but reflective of routine clinical practice. METHODS: One hundred fifty-six patients with pathologically proven breast cancer underwent multi-contrast MRI prior to neoadjuvant chemotherapy and/or surgery. From these, 92 patients were identified for whom T2-weighted, diffusion-weighted and contrast-enhanced T1-weighted sequences were available, as well as key clinicopathological variables. Regions-of-interest were drawn on the above image types and, from these, semantic and calculated radiomics features were derived. Classification models using a variety of methods, both with and without recursive feature elimination, were developed to predict pathological nodal status. Separately, we applied the same methods to analyse the information carried by the radiomic features regarding the originating scanner type and field strength. Repeated, ten-fold cross-validation was employed to verify the results. In parallel work, survival modelling was performed using random survival forests. RESULTS: Prediction of nodal status yielded mean cross-validated AUC values of 0.735 ± 0.15 (SD) for clinical variables alone, 0.673 ± 0.16 (SD) for radiomic features only, and 0.764 ± 0.16 (SD) for radiomics and clinical features together. Prediction of scanner platform from the radiomics features yielded extremely high values of AUC between 0.91 and 1 for the different classes examined indicating the presence of confounding features for the nodal status classification task. Survival analysis, gave out-of-bag prediction errors of 19.3% (clinical features only), 36.9-51.8% (radiomic features from different combinations of image contrasts), and 26.7-35.6% (clinical plus radiomics features). CONCLUSIONS: Radiomic classification models whose predictive ability was consistent with previous single-vendor, single-field strength studies have been obtained from multi-vendor, multi-field-strength data, despite clear confounding information being present. However, our sample size was too small to obtain useful survival modelling results.

Repeatability and reproducibility of apparent diffusion coefficient and fat fraction measurement of focal myeloma lesions on whole body magnetic resonance imaging.

OBJECTIVE: To assess intra- and inter-reader variability of apparent diffusion coefficient (ADC) and fat fraction (FF) measurement in focal myeloma bone lesions and the influence of lesion size. METHODS: 22 myeloma patients with focal active disease on whole body MRI were included. Two readers outlined a small (5-10 mm) and large lesion (>10 mm) in each subject on derived ADC and FF maps; one reader performed this twice. Intra- and inter-reader agreement for small and large lesion groups were calculated for derived statistics from each map using within-subject standard deviation, coefficient of variation, interclass correlation coefficient measures, and visualized with Bland-Altman plots. RESULTS: For mean ADC, intra- and inter-reader repeatability demonstrated equivalently low coefficient of variation (3.0-3.6%) and excellent interclass correlation coefficient (0.975-0.982) for both small and large lesions. For mean FF, intra- and inter-reader repeatability was significantly poorer for small lesions compared to large lesions (intra-reader within-subject standard variation estimate is 2.7 times higher for small lesions than large lesions (p = 0.0071), and for inter-reader variations is 3.8 times higher (p = 0.0070)). CONCLUSION: There is excellent intra- and inter-reader agreement for mean ADC estimates, even for lesions as small as 5 mm. For FF measurements, there is a significant increase in coefficient of variation for smaller lesions, suggesting lesions >10 mm should be selected for lesion FF measurement. ADVANCES IN KNOWLEDGE: ADC measurements of focal myeloma have excellent intra- and inter-reader agreement. FF measurements are more susceptible to lesion size as intra- and inter-reader agreement is significantly impaired in lesions less than 10 mm.

Quantifying MRI T1 relaxation in flowing blood: implications for arterial input function measurement in DCE-MRI.

OBJECTIVES: To investigate the feasibility of accurately quantifying the concentration of MRI contrast agent in flowing blood by measuring its T1 in a large vessel. Such measures are often used to obtain patient-specific arterial input functions for the accurate fitting of pharmacokinetic models to dynamic contrast enhanced MRI data. Flow is known to produce errors with this technique, but these have so far been poorly quantified and characterised in the context of pulsatile flow with a rapidly changing T1 as would be expected in vivo. METHODS: A phantom was developed which used a mechanical pump to pass fluid at physiologically relevant rates. Measurements of T1 were made using high temporal resolution gradient recalled sequences suitable for DCE-MRI of both constant and pulsatile flow. These measures were used to validate a virtual phantom that was then used to simulate the expected errors in the measurement of an AIF in vivo. RESULTS: The relationship between measured T1 values and flow velocity was found to be non-linear. The subsequent error in quantification of contrast agent concentration in a measured AIF was shown. CONCLUSIONS: The T1 measurement of flowing blood using standard DCE- MRI sequences are subject to large measurement errors which are non-linear in relation to flow velocity. ADVANCES IN KNOWLEDGE: This work qualitatively and quantitatively demonstrates the difficulties of accurately measuring the T1 of flowing blood using DCE-MRI over a wide range of physiologically realistic flow velocities and pulsatilities. Sources of error are identified and proposals made to reduce these.