Imaging predictors of treatment outcomes in rectal cancer: An overview

Integrating Imaging and Circulating Tumor DNA Features for Predicting Patient Outcomes

Biomarkers for evaluating tumor response to therapy and estimating the risk of disease relapse represent tremendous areas of clinical need. To evaluate treatment efficacy, tumor response is routinely assessed using different imaging modalities like positron emission tomography/computed tomography or magnetic resonance imaging. More recently, the development of circulating tumor DNA detection assays has provided a minimally invasive approach to evaluate tumor response and prognosis through a blood test (liquid biopsy). Integrating imaging- and circulating tumor DNA-based biomarkers may lead to improvements in the prediction of patient outcomes. For this mini-review, we searched the scientific literature to find original articles that combined quantitative imaging and circulating tumor DNA biomarkers to build prediction models. Seven studies reported building prognostic models to predict distant recurrence-free, progression-free, or overall survival. Three discussed building models to predict treatment response using tumor volume, pathologic complete response, or objective response as endpoints. The limited number of articles and the modest cohort sizes reported in these studies attest to the infancy of this field of study. Nonetheless, these studies demonstrate the feasibility of developing multivariable response-predictive and prognostic models using regression and machine learning approaches. Larger studies are warranted to facilitate the building of highly accurate response-predictive and prognostic models that are generalizable to other datasets and clinical settings.

An automated deep learning pipeline for EMVI classification and response prediction of rectal cancer using baseline MRI: a multi-centre study

The classification of extramural vascular invasion status using baseline magnetic resonance imaging in rectal cancer has gained significant attention as it is an important prognostic marker. Also, the accurate prediction of patients achieving complete response with primary staging MRI assists clinicians in determining subsequent treatment plans. Most studies utilised radiomics-based methods, requiring manually annotated segmentation and handcrafted features, which tend to generalise poorly. We retrospectively collected 509 patients from 9 centres, and proposed a fully automated pipeline for EMVI status classification and CR prediction with diffusion weighted imaging and T2-weighted imaging. We applied nnUNet, a self-configuring deep learning model, for tumour segmentation and employed learned multiple-level image features to train classification models, named MLNet. This ensures a more comprehensive representation of the tumour features, in terms of both fine-grained detail and global context. On external validation, MLNet, yielding similar AUCs as internal validation, outperformed 3D ResNet10, a deep neural network with ten layers designed for analysing spatiotemporal data, in both CR and EMVI tasks. For CR prediction, MLNet showed better results than the current state-of-the-art model using imaging and clinical features in the same external cohort. Our study demonstrated that incorporating multi-level image representations learned by a deep learning based tumour segmentation model on primary MRI improves the results of EMVI classification and CR prediction with good generalisation to external data. We observed variations in the contributions of individual feature maps to different classification tasks. This pipeline has the potential to be applied in clinical settings, particularly for EMVI classification.

Photoacoustic clinical applications: Musculoskeletal and abdominal imaging

Photoacoustic (PA) imaging has been extensively investigated in application in biomedicine over the last decade. This article reviews the motivation, significance, and system configuration of a few ongoing studies of implementing photoacoustic technology in musculoskeletal imaging, abdominal imaging, and interstitial sensing. The review then summarizes the methodologies and latest progress of relevant projects. Finally, we discuss our expectations for the future of translation research in PA imaging.

Baseline and early digital [ 18 F]FDG PET/CT and multiparametric MRI contain promising features to predict response to neoadjuvant therapy in locally advanced rectal cancer patients: a pilot study

OBJECTIVE

In this pilot study, we investigated the feasibility of response prediction using digital [ 18 F]FDG PET/computed tomography (CT) and multiparametric MRI before, during, and after neoadjuvant chemoradiation therapy in locally advanced rectal cancer (LARC) patients and aimed to select the most promising imaging modalities and timepoints for further investigation in a larger trial.

METHODS

Rectal cancer patients scheduled to undergo neoadjuvant chemoradiation therapy were prospectively included in this trial, and underwent multiparametric MRI and [ 18 F]FDG PET/CT before, 2 weeks into, and 6-8 weeks after chemoradiation therapy. Two groups were created based on pathological tumor regression grade, that is, good responders (TRG1-2) and poor responders (TRG3-5). Using binary logistic regression analysis with a cutoff value of P  ≤ 0.2, promising predictive features for response were selected.

RESULTS

Nineteen patients were included. Of these, 5 were good responders, and 14 were poor responders. Patient characteristics of these groups were similar at baseline. Fifty-seven features were extracted, of which 13 were found to be promising predictors of response. Baseline [T2: volume, diffusion-weighted imaging (DWI): apparent diffusion coefficient (ADC) mean, DWI: difference entropy], early response (T2: volume change, DWI: ADC mean change) and end-of-treatment presurgical evaluation MRI (T2: gray level nonuniformity, DWI: inverse difference normalized, DWI: gray level nonuniformity normalized), as well as baseline (metabolic tumor volume, total lesion glycolysis) and early response PET/CT (Δ maximum standardized uptake value, Δ peak standardized uptake value corrected for lean body mass), were promising features.

CONCLUSION

Both multiparametric MRI and [ 18 F]FDG PET/CT contain promising imaging features to predict response to neoadjuvant chemoradiotherapy in LARC patients. A future larger trial should investigate baseline, early response, and end-of-treatment presurgical evaluation MRI and baseline and early response PET/CT.

Associations between Response to Commonly Used Neo-Adjuvant Schedules in Rectal Cancer and Routinely Collected Clinical and Imaging Parameters

Complete pathological response (pCR) is achieved in 10−20% of rectal cancers when treated with short-course radiotherapy (scRT) or long-course chemoradiotherapy (CRT) and in 28% with total neoadjuvant therapy (scRT/CRT + CTX). pCR is associated with better outcomes and a “watch-and-wait” strategy (W&W). The aim of this study was to identify baseline clinical or imaging factors predicting pCR. All patients with preoperative treatment and delays to surgery in Uppsala-Dalarna (n = 359) and Stockholm (n = 635) were included. Comparison of pCR versus non-pCR was performed with binary logistic regression models. Receiver operating characteristics (ROC) models for predicting pCR were built using factors with p < 0.10 in multivariate analyses. A pCR was achieved in 12% of the 994 patients (scRT 8% [33/435], CRT 13% [48/358], scRT/CRT + CTX 21% [43/201]). In univariate and multivariate analyses, choice of CRT (OR 2.62; 95%CI 1.34−5.14, scRT reference) or scRT/CRT + CTX (4.70; 2.23−9.93), cT1−2 (3.37; 1.30−8.78; cT4 reference), tumour length ≤ 3.5 cm (2.27; 1.24−4.18), and CEA ≤ 5 µg/L (1.73; 1.04−2.90) demonstrated significant associations with achievement of pCR. Age < 70 years, time from radiotherapy to surgery > 11 weeks, leucocytes ≤ 109/L, and thrombocytes ≤ 4009/L were significant only in univariate analyses. The associations were not fundamentally different between treatments. A model including T-stage, tumour length, CEA, and leucocytes (with scores of 0, 0.5, or 1 for each factor, maximum 4 points) showed an area under the curve (AUC) of 0.66 (95%CI 0.60−0.71) for all patients, and 0.65−0.73 for the three treatments separately. The choice of neoadjuvant treatment in combination with low CEA, short tumour length, low cT-stage, and normal leucocytes provide support in predicting pCR and, thus, could offer guidance for selecting patients for organ preservation.

Pattern of Rectal Cancer Recurrence Following Potentially Curative Surgical Treatment

Survival in rectal cancer has been related mainly to clinical and pathological staging. Recurrence is the most challenging issue when surgical treatment of rectal cancer is concerned. This study aims to establish a recurrence pattern for rectal adenocarcinoma submitted to surgical treatment between June 2003 and July 2021. After applying the exclusion criteria to 305 patients, 166 patients were analyzed. Global recurrence was found in 18.7% of them, while 7.8% have had local recurrence. Recurrences were diagnosed from 5 to 92 months after the surgical procedure, with a median of 32.5 months. Follow-up varied from 6 to 115 months. Recurrence, in literature, is usually between 3 and 35% in 5 years and shows a 5-year survival rate of only 5%. In around 50% of cases, recurrence is local, confined to the pelvis. This study was consonant with the literature in most aspects evaluated, although a high rate of local recurrence remains a challenge in seeking better surgical outcomes.

A brain tumor reporting and data system to optimize imaging surveillance and prognostication in high-grade gliomas

BACKGROUND AND PURPOSE

High-grade glioma (HGG), including glioblastoma, is the most common primary brain neoplasm and has a dismal prognosis. After initial treatment, follow-up decisions are guided by longitudinal MRI performed at routine intervals. The Brain Tumor Reporting and Data System (BT-RADS) is a proposed structured reporting system for posttreatment brain MRIs. The purpose of this study is to determine the relationship between BT-RADS scores and overall survival in HGG patients.

METHODS

Chart review of grade 4 glioma patients who had an MRI at a single institution from November 2018 to November 2019 was performed. BT-RADS scores, tumor characteristics, and overall survival were recorded. Likelihood of improvement, stability, or worsening on the subsequent study was calculated for each score. Survival analysis was performed using Kaplan-Meier method, log-rank test, and a time-dependent cox model. Significance level of .05 was used.

RESULTS

The study identified 91 HGG patients who underwent a total of 538 MRIs. Mean age of patients was 57 years old. Score with the highest likelihood for worsening on the next follow-up was 3b. The risk of death was 53% higher with each incremental increase in BT-RADS scores (hazard ratio, 1.53; 95% confidence interval [CI], 1.07-2.19; p = .019). The risk of death was 167% higher in O-6-methylguanine-DNA-methyltransferase unmethylated tumors (hazard ratio, 2.67; 95% CI, 1.34-5.33; p = .005).

CONCLUSIONS

BT-RADS scores can be used as a reference guide to anticipate whether patients' subsequent MRI will be improved, stable, or worsened. The scoring system can also be used to predict clinical outcomes and prognosis.

Preoperative Prediction of Perineural Invasion Status of Rectal Cancer Based on Radiomics Nomogram of Multiparametric Magnetic Resonance Imaging

Objectives

To compare the predictive performance of different radiomics signatures from multiparametric magnetic resonance imaging (mpMRI), including four sequences when used individually or combined, and to establish and validate an optimal nomogram for predicting perineural invasion (PNI) in rectal cancer (RC) patients.

Methods

Our retrospective study included 279 RC patients without preoperative antitumor therapy (194 in the training dataset and 85 in the test dataset) who underwent preoperative mpMRI scan between January 2017 and January 2021. Among them, 72 cases were PNI-positive. Then, clinical and radiological variables were collected, including carcinoembryonic antigen (CEA), radiological tumour stage (T_1-4), lymph node stage (N_0-2) and so on. Quantitative radiomics features were extracted and selected from oblique axial T₂-weighted imaging (T₂WI), T₁-weighted imaging (T₁WI), apparent diffusion coefficient (ADC), and enhanced T₁WI (T₁CE) sequences. The clinical model was constructed by integrating the final selected clinical and radiological variables. The radiomics signatures included four single-sequence signatures and one fusion signature were built using the respective remaining optimized features. And the nomogram was constructed based on the independent predictors by using multivariable logistic regression. The area under curve (AUC), DeLong test, calibration curve, and decision curve analysis (DCA) were used to evaluate the performance.

Results

Ultimately, 20 radiomics features were retained from the four sequences—T₁WI (n = 4), T₂WI (n = 5), ADC (n = 5), and T₁CE (n = 6)—to construct four single-sequence radiomics signatures and one fusion radiomics signature. The fusion radiomics signature performed better than four single-sequence radiomics signatures and clinical model (AUCs of 0.835 and 0.773 vs. 0.680-0.737 and 0.666-0.709 in the training and test datasets, respectively). The nomogram constructed by incorporating CEA, tumour stage and rad-score performed best, with AUCs of 0.869 and 0.864 in the training and test datasets, respectively. Delong test showed that the nomogram was significantly different from the clinical model and four single-sequence radiomics signatures (P < 0.05). Moreover, calibration curves demonstrated good agreement, and DCA highlighted benefits of the nomogram.

Conclusions

The comprehensive nomogram can preoperatively and noninvasively predict PNI status, provide a convenient and practical tool for treatment strategy, and help optimize individualized clinical decision-making in RC patients.

MRI radiomics independent of clinical baseline characteristics and neoadjuvant treatment modalities predicts response to neoadjuvant therapy in rectal cancer

Background

To analyse the performance of multicentre pre-treatment MRI-based radiomics (MBR) signatures combined with clinical baseline characteristics and neoadjuvant treatment modalities to predict complete response to neoadjuvant (chemo)radiotherapy in locally advanced rectal cancer (LARC).

Methods

Baseline MRI and clinical characteristics with neoadjuvant treatment modalities at four centres were collected. Decision tree, support vector machine and five-fold cross-validation were applied for two non-imaging and three radiomics-based models’ development and validation.

Results

We finally included 674 patients. Pre-treatment CEA, T stage, and histologic grade were selected to generate two non-imaging models: C model (clinical baseline characteristics alone) and CT model (clinical baseline characteristics combining neoadjuvant treatment modalities). The prediction performance of both non-imaging models were poor. The MBR signatures comprising 30 selected radiomics features, the MBR signatures combining clinical baseline characteristics (CMBR), and the CMBR incorporating neoadjuvant treatment modalities (CTMBR) all showed good discrimination with mean AUCs of 0.7835, 0.7871 and 0.7916 in validation sets, respectively. The three radiomics-based models had insignificant discrimination in performance.

Conclusions

The performance of the radiomics-based models were superior to the non-imaging models. MBR signatures seemed to reflect LARC’s true nature more accurately than clinical parameters and helped identify patients who can undergo organ preservation strategies.

Computed tomography-derived biomarker for predicting the treatment response to neoadjuvant chemoradiotherapy of rectal cancer

BACKGROUND

Malignant tumor essentially implies structural heterogeneity. Analysis of medical imaging can quantify this structural heterogeneity, which can be a new biomarker. This study aimed to evaluate the usefulness of texture analysis of computed tomography (CT) imaging as a biomarker for predicting the therapeutic response of neoadjuvant chemoradiotherapy (nCRT) for locally advanced rectal cancer.

METHODS

We enrolled 76 patients with rectal cancer who underwent curative surgery after nCRT. Texture analyses (Fractal analysis and Histogram analysis) were applied to contrast-enhanced CT images, and fractal dimension (FD), skewness, and kurtosis of the tumor were calculated. These CT-derived parameters were compared with the therapeutic response and prognosis.

RESULTS

Forty-six of 76 patients were diagnosed as clinical responders after nCRT. Kurtosis was significantly higher in the responders group than in the non-responders group (4.17 ± 4.16 vs. 2.62 ± 3.19, p = 0.04). Nine of 76 patients were diagnosed with pathological complete response (pCR) after surgery. FD of the pCR group was significantly lower than that of the non-pCR group (0.90 ± 0.12 vs. 1.01 ± 0.12, p = 0.009). The area under the receiver-operating characteristics curve of tumor FD for predicting pCR was 0.77, and the optimal cut-off value was 0.84 (accuracy; 93.4%). Furthermore, patients with lower FD tumors tended to show better relapse-free survival and disease-specific survival than those with higher FD tumors (5-year, 80.8 vs. 66.6%, 94.4 vs. 80.2%, respectively), although it was not statistically significant (p = 0.14, 0.11).

CONCLUSIONS

CT-derived texture parameters could be potential biomarkers for predicting the therapeutic response of rectal cancer.

Radiomics as a New Frontier of Imaging for Cancer Prognosis: A Narrative Review

The evaluation of the efficacy of different therapies is of paramount importance for the patients and the clinicians in oncology, and it is usually possible by performing imaging investigations that are interpreted, taking in consideration different response evaluation criteria. In the last decade, texture analysis (TA) has been developed in order to help the radiologist to quantify and identify parameters related to tumor heterogeneity, which cannot be appreciated by the naked eye, that can be correlated with different endpoints, including cancer prognosis. The aim of this work is to analyze the impact of texture in the prediction of response and in prognosis stratification in oncology, taking into consideration different pathologies (lung cancer, breast cancer, gastric cancer, hepatic cancer, rectal cancer). Key references were derived from a PubMed query. Hand searching and clinicaltrials.gov were also used. This paper contains a narrative report and a critical discussion of radiomics approaches related to cancer prognosis in different fields of diseases.

Tumor Volume as Predictor of Pathologic Complete Response Following Neoadjuvant Chemoradiation in Locally Advanced Rectal Cancer

PURPOSE

Neoadjuvant chemoradiation followed by surgery is the current standard of care in the treatment of locally advanced rectal cancer. Those who achieved pathologic complete response, following this standard of care, complete pathologic response (pCR) had better outcome. Until now there are no reliable clinical parameters to predict this response. The purpose of the study was to evaluate whether tumor volume may serve as a predictive factor in patients treated with neoadjuvant chemoradiotherapy.

MATERIALS AND METHODS

Between September 2015 and September 2019, patients diagnosed with stage IIA to IIIC rectal adenocarcinoma, who were treated with neoadjuvant chemoradiation, were enrolled to this study. All patients underwent rectal ultrasound, pelvic magnetic resonance imaging, fluorodeoxyglucose-positron emission tomography-computed tomography and the diagnosis was confirmed by pathology report. Radiation therapy was consisted of 50 Gy delivered to the tumor site, 2 Gy a day, 5 times a week and to the pelvic lymph nodes for a total of 45 Gy in 1.8 Gy a day, 5 times a week. The gross tumor volume (GTV) was contoured by radiation oncology expert, reviewed by radiology and nuclear medicine expert and approved by radiation therapy tumor board. Chemotherapy was consisted of either capecitabine 875 mg/m2 twice a day or continuous. IV infusion of 5 fluorouracil 375 mg/m2 for 4 consecutive days in a 3 weeks apart. Operation, either low anterior or abdominoperineal resection was carried out 6 to 8 weeks following completion of treatment. Patients were assigned to either complete pathologic response (pCR) or non-pCR groups. GTV, among other clinical and treatment parameters, were evaluated for prediction of pCR. Statistical methods included independent t test, logistic regression, area under the curve-receiver operating characteristic, Bayesian independent statistics and multilayer perceptron model.

RESULTS

One hundred ninety-three patients were enrolled to this study, 6 were excluded due to metastatic disease detected at the time of operation. Seventy had stage II and 117 had stage III. Forty-four of 187 (23.5%) patients achieved pCR and 143 patients had either partial or no response/progressive disease. Among the 44 pCR group, 21 had stage II and 23 had stage III disease. Treatment interruption, defined as either a delay of up to 1 week in radiation, and a dose reduction to 75%, was occurred in 42 patients. Sex, ethnicity, distance from anal verge to tumor, height, weight, age, delivered radiation dose, radiotherapy techniques, clinical T and N stage and GTV were evaluated for prediction of pCR. GTV at the volume of <39.5 cm3 was the only significant predictive factor to detect pCR by logistic regression model (P<0.01) and by Bayesian independent test (P=0.026). Area under the receiver operating characteristic curve of GTV <39.5 cm3 showed area under the curve of 0.715 (P=0.009) for stage II and area under the curve of 0.62 (P>0.05) for stage III.

CONCLUSION

GTV may serve as a predictive factor for achieving pCR in locally advanced rectal cancer after neoadjuvant chemoradiotherapy.

Assessing Rectal Cancer Treatment Response Using Coregistered Endorectal Photoacoustic and US Imaging Paired with Deep Learning

Background Conventional radiologic modalities perform poorly in the radiated rectum and are often unable to differentiate residual cancer from treatment scarring. Purpose To report the development and initial patient study of an imaging system comprising an endorectal coregistered photoacoustic (PA) microscopy (PAM) and US system paired with a convolution neural network (CNN) to assess the rectal cancer treatment response. Materials and Methods In this prospective study (ClinicalTrials.gov identifier NCT04339374), participants completed radiation and chemotherapy from September 2019 to September 2020 and images were obtained with the PAM/US system prior to surgery. Another group's colorectal specimens were studied ex vivo. The PAM/US system consisted of an endorectal imaging probe, a 1064-nm laser, and one US ring transducer. The PAM CNN and US CNN models were trained and validated to distinguish normal from malignant colorectal tissue using ex vivo and in vivo patient data. The PAM CNN and US CNN were then tested using additional in vivo patient data that had not been seen by the CNNs during training and validation. Results Twenty-two patients' ex vivo specimens and five patients' in vivo images (a total of 2693 US regions of interest [ROIs] and 2208 PA ROIs) were used for CNN training and validation. Data from five additional patients were used for testing. A total of 32 participants (mean age, 60 years; range, 35-89 years) were evaluated. Unique PAM imaging markers of the complete tumor response were found, specifically including recovery of normal submucosal vascular architecture within the treated tumor bed. The PAM CNN model captured this recovery process and correctly differentiated these changes from the residual tumor. The imaging system remained highly capable of differentiating tumor from normal tissue, achieving an area under the receiver operating characteristic curve of 0.98 (95% CI: 0.98, 0.99) for data from five participants. By comparison, the US CNN had an area under the receiver operating characteristic curve of 0.71 (95% CI: 0.70, 0.73). Conclusion An endorectal coregistered photoacoustic microscopy/US system paired with a convolutional neural network model showed high diagnostic performance in assessing the rectal cancer treatment response and demonstrated potential for optimizing posttreatment management. © RSNA, 2021 Supplemental material is available for this article. See also the editorial by Klibanov in this issue.

Magnetic resonance imaging as a predictor of surgical outcome in patients with local pelvic recurrence of colorectal cancer

INTRODUCTION

Pelvic local recurrence of colorectal cancer (PRCRC) may be cured if radical surgery is performed. Preoperative assessment normally includes magnetic resonance imaging (MRI). The aim of this study was to evaluate the influence of specific MRI-related findings on outcome of surgery of PRCRC.

MATERIALS AND METHODS

Clinical data from 95 consecutive patients, operated with a curative intent for PRCRC at Karolinska University Hospital during 2003-2013, were collected from medical records. Preoperative MRI examinations of the PRCRC were re-evaluated. The potential influence of clinical factors and specific MRI-findings (location, solid/mucinous, size, volume and border) on surgical resection margins (R0-R1) and survival were calculated with logistic and cox regression.

RESULTS

Eighty-seven patients had available MRI scans and were included in the study. Sixty-five patients (75%) had a R0 resection and 22 patients (25%) had a R1 resection of their PRCRC. In all, 47 patients (54%) had an involved lateral compartment. Lateral location was the only MRI finding associated with both an increased risk of R1 resection (OR 3.97, 95%CI: 1.31-12.04) and death (HR 1.94, 95%CI: 1.07-3.51). Lateral location entailed an increased risk of death also after R0 resection (HR2.09, 95%CI: 1.07-4.10). Five-year survival was 35% for all patients, 44% after R0 resection and 7% after R1 resection.

CONCLUSION

Tumour involvement of the lateral and posterior compartments on MRI was a predictor for R1 resection, but only lateral involvement was associated with an increased risk of death. An increased risk of death associated with lateral involvement was still present after R0 resection.

Radiomics performs comparable to morphologic assessment by expert radiologists for prediction of response to neoadjuvant chemoradiotherapy on baseline staging MRI in rectal cancer

PURPOSE

To compare the performance of advanced radiomics analysis to morphological assessment by expert radiologists to predict a good or complete response to chemoradiotherapy in rectal cancer using baseline staging MRI.

MATERIALS AND METHODS

We retrospectively assessed the primary staging MRIs [prior to chemoradiotherapy (CRT)] of 133 rectal cancer patients from 2 centers. First, two expert radiologists subjectively estimated the likelihood of achieving a "complete response" (ypT0) and "good response" (TRG 1-2), using a 5-point score (based on TN-stage, MRF/EMVI-status, size/signal/shape). Next, tumor volumes were segmented on high b value DWI (semi-automated, corrected by 2 non-expert and 2-expert readers, resulting in 5 segmentations), copied to the remaining sequences after which a total of 2505 radiomic features were extracted from T2W, low and high b value DWI and ADC. Stability of features for noise due to inter-reader and inter-scanner and protocol variations was assessed using intraclass correlation (ICC) and the Kruskal-Wallis test. Using data from center 1 (n = 86; training set), top 9 features were selected using minimum Redundancy Maximum Relevance and combined in a logistic regression model. Finally, diagnostic performance of the fitted models was assessed on data from center 2 (n = 47; validation set) and compared to the performance of the radiologists.

RESULTS

The Radiomic models resulted in AUCs of 0.69-0.79 (with similar results for the segmentations performed by expert/non-expert readers) to predict response, results similar to the morphologic prediction by the expert radiologists (AUC 0.67-0.83). Radiomics using semi-automatically generated segmentations (without manual input) did not result in significant predictive performance.

CONCLUSIONS

Radiomics could predict response to therapy with comparable diagnostic performance as expert radiologists, regardless of whether image segmentation was performed by non-expert or expert readers, indicating that expert input is not required in order for the radiomics workflow to produce significant predictive performance.