Challenges in ensuring the generalizability of image quantitation methods for MRI

The power of many brains: Catalyzing neuropsychiatric discovery through open neuroimaging data and large-scale collaboration

Recent advances in open neuroimaging data are enhancing our comprehension of neuropsychiatric disorders. By pooling images from various cohorts, statistical power has increased, enabling the detection of subtle abnormalities and robust associations, and fostering new research methods. Global collaborations in imaging have furthered our knowledge of the neurobiological foundations of brain disorders and aided in imaging-based prediction for more targeted treatment. Large-scale magnetic resonance imaging initiatives are driving innovation in analytics and supporting generalizable psychiatric studies. We also emphasize the significant role of big data in understanding neural mechanisms and in the early identification and precise treatment of neuropsychiatric disorders. However, challenges such as data harmonization across different sites, privacy protection, and effective data sharing must be addressed. With proper governance and open science practices, we conclude with a projection of how large-scale imaging resources and collaborations could revolutionize diagnosis, treatment selection, and outcome prediction, contributing to optimal brain health.

Radiomic profiles improve prognostication and reveal targets for therapy in cervical cancer

Cervical cancer (CC) is a major global health problem with 570,000 new cases and 266,000 deaths annually. Prognosis is poor for advanced stage disease, and few effective treatments exist. Preoperative diagnostic imaging is common in high-income countries and MRI measured tumor size routinely guides treatment allocation of cervical cancer patients. Recently, the role of MRI radiomics has been recognized. However, its potential to independently predict survival and treatment response requires further clarification. This retrospective cohort study demonstrates how non-invasive, preoperative, MRI radiomic profiling may improve prognostication and tailoring of treatments and follow-ups for cervical cancer patients. By unsupervised clustering based on 293 radiomic features from 132 patients, we identify three distinct clusters comprising patients with significantly different risk profiles, also when adjusting for FIGO stage and age. By linking their radiomic profiles to genomic alterations, we identify putative treatment targets for the different patient clusters (e.g., immunotherapy, CDK4/6 and YAP-TEAD inhibitors and p53 pathway targeting treatments).

Discrimination of benign, atypical, and malignant peripheral nerve sheath tumors in neurofibromatosis type 1 using diffusion-weighted MRI

Background

Neurofibromatosis type 1 (NF1) is associated with the development of benign (BPNST) and malignant (MPNST) peripheral nerve sheath tumors. Recently described atypical neurofibromas (ANF) are considered pre-malignant precursor lesions to MPNSTs. Previous studies indicate that diffusion-weighted magnetic resonance imaging (DW-MRI) can reliably discriminate MPNSTs from BPNSTs. We therefore investigated the diagnostic accuracy of DW-MRI for the discrimination of benign, atypical, and malignant peripheral nerve sheath tumors.

Methods

In this prospective explorative single-center phase II diagnostic study, 44 NF1 patients (23 male; 30.1 ± 11.8 years) underwent DW-MRI (b-values 0-800 s/mm²) at 3T. Two radiologists independently assessed mean and minimum apparent diffusion coefficients (ADCmean/min) in areas of largest tumor diameters and ADCdark in areas of lowest signal intensity by manual contouring of the tumor margins of 60 BPNSTs, 13 ANFs, and 21 MPNSTs. Follow-up of ≥ 24 months (BPNSTs) or histopathological evaluation (ANFs + MPNSTs) served as diagnostic reference standard. Diagnostic ADC-based cut-off values for discrimination of the three tumor groups were chosen to yield the highest possible specificity while maintaining a clinically acceptable sensitivity.

Results

ADC values of pre-malignant ANFs clustered between BPNSTs and MPNSTs. Best BPNST vs. ANF + MPNST discrimination was obtained using ADCdark at a cut-off value of 1.6 × 10-3 mm2/s (85.3% sensitivity, 93.3% specificity), corresponding to an AUC of 94.3% (95% confidence interval: 85.2-98.0). Regarding BPNST + ANF vs. MPNST, best discrimination was obtained using an ADCdark cut-off value of 1.4 × 10-3 mm2/s (83.3% sensitivity, 94.5% specificity).

Conclusions

DW-MRI using ADCdark allows specific and noninvasive discrimination of benign, atypical, and malignant nerve sheath tumors in NF1.

Assessing the Feasibility of Using Artificial Intelligence-Segmented Dominant Intraprostatic Lesion for Focal Intraprostatic Boost With External Beam Radiation Therapy

PURPOSE

The delineation of dominant intraprostatic gross tumor volumes (GTVs) on multiparametric magnetic resonance imaging (mpMRI) can be subject to interobserver variability. We evaluated whether deep learning artificial intelligence (AI)-segmented GTVs can provide a similar degree of intraprostatic boosting with external beam radiation therapy (EBRT) as radiation oncologist (RO)-delineated GTVs.

METHODS AND MATERIALS

We identified 124 patients who underwent mpMRI followed by EBRT between 2010 and 2013. A reference GTV was delineated by an RO and approved by a board-certified radiologist. We trained an AI algorithm for GTV delineation on 89 patients, and tested the algorithm on 35 patients, each with at least 1 PI-RADS (Prostate Imaging Reporting and Data System) 4 or 5 lesion (46 total lesions). We then asked 5 additional ROs to independently delineate GTVs on the test set. We compared lesion detectability and geometric accuracy of the GTVs from AI and 5 ROs against the reference GTV. Then, we generated EBRT plans (77 Gy prostate) that boosted each observer-specific GTV to 95 Gy. We compared reference GTV dose (D98%) across observers using a mixed-effects model.

RESULTS

On a lesion level, AI GTV exhibited a sensitivity of 82.6% and positive predictive value of 86.4%. Respective ranges among the 5 RO GTVs were 84.8% to 95.7% and 95.1% to 100.0%. Among 30 GTVs mutually identified by all observers, no significant differences in Dice coefficient were detected between AI and any of the 5 ROs. Across all patients, only 2 of 5 ROs had a reference GTV D98% that significantly differed from that of AI by 2.56 Gy (P = .02) and 3.20 Gy (P = .003). The presence of false-negative (-5.97 Gy; P < .001) but not false-positive (P = .24) lesions was associated with reference GTV D98%.

CONCLUSIONS

AI-segmented GTVs demonstrate potential for intraprostatic boosting, although the degree of boosting may be adversely affected by false-negative lesions. Prospective review of AI-segmented GTVs remains essential.

Development and multicenter validation of a multiparametric imaging model to predict treatment response in rectal cancer

OBJECTIVES

To develop and validate a multiparametric model to predict neoadjuvant treatment response in rectal cancer at baseline using a heterogeneous multicenter MRI dataset.

METHODS

Baseline staging MRIs (T2W (T2-weighted)-MRI, diffusion-weighted imaging (DWI) / apparent diffusion coefficient (ADC)) of 509 patients (9 centres) treated with neoadjuvant chemoradiotherapy (CRT) were collected. Response was defined as (1) complete versus incomplete response, or (2) good (Mandard tumor regression grade (TRG) 1-2) versus poor response (TRG3-5). Prediction models were developed using combinations of the following variable groups: (1) Non-imaging: age/sex/tumor-location/tumor-morphology/CRT-surgery interval (2) Basic staging: cT-stage/cN-stage/mesorectal fascia involvement, derived from (2a) original staging reports, or (2b) expert re-evaluation (3) Advanced staging: variables from 2b combined with cTN-substaging/invasion depth/extramural vascular invasion/tumor length (4) Quantitative imaging: tumour volume + first-order histogram features (from T2W-MRI and DWI/ADC) Models were developed with data from 6 centers (n = 412) using logistic regression with the Least Absolute Shrinkage and Selector Operator (LASSO) feature selection, internally validated using repeated (n = 100) random hold-out validation, and externally validated using data from 3 centers (n = 97).

RESULTS

After external validation, the best model (including non-imaging and advanced staging variables) achieved an area under the curve of 0.60 (95%CI=0.48-0.72) to predict complete response and 0.65 (95%CI=0.53-0.76) to predict a good response. Quantitative variables did not improve model performance. Basic staging variables consistently achieved lower performance compared to advanced staging variables.

CONCLUSIONS

Overall model performance was moderate. Best results were obtained using advanced staging variables, highlighting the importance of good-quality staging according to current guidelines. Quantitative imaging features had no added value (in this heterogeneous dataset).

CLINICAL RELEVANCE STATEMENT

Predicting tumour response at baseline could aid in tailoring neoadjuvant therapies for rectal cancer. This study shows that image-based prediction models are promising, though are negatively affected by variations in staging quality and MRI acquisition, urging the need for harmonization.

KEY POINTS

This multicenter study combining clinical information and features derived from MRI rendered disappointing performance to predict response to neoadjuvant treatment in rectal cancer. Best results were obtained with the combination of clinical baseline information and state-of-the-art image-based staging variables, highlighting the importance of good quality staging according to current guidelines and staging templates. No added value was found for quantitative imaging features in this multicenter retrospective study. This is likely related to acquisition variations, which is a major problem for feature reproducibility and thus model generalizability.

A Review on the Use of Imaging Biomarkers in Oncology Clinical Trials: Quality Assurance Strategies for Technical Validation

Imaging biomarkers (IBs) have been proposed in medical literature that exploit images in a quantitative way, going beyond the visual assessment by an imaging physician. These IBs can be used in the diagnosis, prognosis, and response assessment of several pathologies and are very often used for patient management pathways. In this respect, IBs to be used in clinical practice and clinical trials have a requirement to be precise, accurate, and reproducible. Due to limitations in imaging technology, an error can be associated with their value when considering the entire imaging chain, from data acquisition to data reconstruction and subsequent analysis. From this point of view, the use of IBs in clinical trials requires a broadening of the concept of quality assurance and this can be a challenge for the responsible medical physics experts (MPEs). Within this manuscript, we describe the concept of an IB, examine some examples of IBs currently employed in clinical practice/clinical trials and analyze the procedure that should be carried out to achieve better accuracy and reproducibility in their use. We anticipate that this narrative review, written by the components of the EFOMP working group on "the role of the MPEs in clinical trials"-imaging sub-group, can represent a valid reference material for MPEs approaching the subject.

Delivering a Quantitative Imaging Agenda

In a digital image, each voxel contains quantitative information dependent on the technique used to generate the image [...].

Whole-Lesion Apparent Diffusion Coefficient Histogram Analysis for Assessing Normal-Sized Lymph Node Metastasis in Cervical Cancer: Comparison Between Readout-Segmented and Single-Shot Echo-Planar Diffusion-Weighted Imaging

OBJECTIVE

To compare the value of whole-lesion apparent diffusion coefficient (ADC) histogram analysis derived from readout-segmented echo-planar imaging (RS-EPI) and single-shot echo-planar imaging (SS-EPI) diffusion-weighted imaging (DWI) in evaluating normal-sized lymph node metastasis (LNM) in cervical cancer.

METHODS

Seventy-six pathologically confirmed cervical cancer patients (stages IB and IIA) were enrolled, including 61 patients with non-LNM (group A) and 15 patients with normal-sized LNM (group B). The recorded tumor volume on T2-weighted imaging was the reference against which both DWIs were evaluated. Each ADC histogram parameter (including ADCmax, ADC90, ADCmedian, ADCmean, ADC10, ADCmin, ADCskewness, ADCkurtosis, and ADCentropy) was compared between SS-EPI and RS-EPI and between the 2 groups.

RESULTS

There was no significant difference in tumor volume between the 2 DWIs and T2-weighted imaging (both P > 0.05). Higher ADCmax and ADCentropy but lower ADC10, ADCmin and ADCskewness were found in SS-EPI than those in RS-EPI (all P < 0.05). For SS-EPI, lower ADC90 and higher ADCkurtosis were found in group B than those in group A (both P < 0.05). For RS-EPI, lower ADC90 and higher ADCkurtosis and ADCentropy were found in group B than those in group A (all P < 0.05). Readout-segmented echo-planar imaging ADCkurtosis showed the highest area under the curve of 0.792 in the differentiation of the 2 groups (sensitivity, 80%; specificity, 73.77%).

CONCLUSIONS

Compared with SS-EPI, the ADC histogram parameters derived from RS-EPI were more accurate, and ADCkurtosis held great potential in differentiating normal-sized LNM in cervical cancer.

RESOLVE-based radiomics in cervical cancer: improved image quality means better feature reproducibility?

AIM

To compare the reproducibility of apparent diffusion coefficient (ADC)-based radiomic features between readout-segmented echo-planar diffusion-weighted imaging (RESOLVE) and single-shot echo-planar diffusion-weighted imaging (SS-EPI DWI) in cervical cancer.

MATERIALS AND METHODS

The RESOLVE and SS-EPI DWI images of 36 patients with histopathologically confirmed cervical cancer were collected retrospectively. Two observers independently delineated the whole tumour on RESOLVE and SS-EPI DWI, and then copied them to the corresponding ADC maps. Shape, first-order, and texture features were extracted from ADC maps in the original and filtered (Laplacian of Gaussian [LoG] and wavelet) images. Thereafter, 1,316 features were generated in each RESOLVE and SS-EPI DWI, respectively. The reproducibility of radiomic features was assessed using intraclass correlation coefficient (ICC).

RESULTS

In the original images, RESOLVE showed 92.86%, 66.67%, and 86.67% of features with excellent reproducibility in shape, first-order, and texture features, while SS-EPI DWI showed 85.71%, 72.22%, and 60% of features with excellent reproducibility, respectively. In the LoG and wavelet filtered images, RESOLVE had 56.77% and 65.32% of features with excellent reproducibility and SS-EPI DWI had 44.95% and 61.96% of features with excellent reproducibility, respectively.

CONCLUSION

Compared with SS-EPI DWI, the feature reproducibility of RESOLVE was better in cervical cancer, especially for texture features. The filtered images cannot improve the feature reproducibility compared with the original images for both SS-EPI DWI and RESOLVE.

Multimodal MRI radiomic models to predict genomic mutations in diffuse intrinsic pontine glioma with missing imaging modalities

Purpose

Predicting H3.1, TP53, and ACVR1 mutations in DIPG could aid in the selection of therapeutic options. The contribution of clinical data and multi-modal MRI were studied for these three predictive tasks. To keep the maximum number of subjects, which is essential for a rare disease, missing data were considered. A multi-modal model was proposed, collecting all available data for each patient, without performing any imputation.

Methods

A retrospective cohort of 80 patients with confirmed DIPG and at least one of the four MR modalities (T1w, T1c, T2w, and FLAIR), acquired with two different MR scanners was built. A pipeline including standardization of MR data and extraction of radiomic features within the tumor was applied. The values of radiomic features between the two MR scanners were realigned using the ComBat method. For each prediction task, the most robust features were selected based on a recursive feature elimination with cross-validation. Five different models, one based on clinical data and one per MR modality, were developed using logistic regression classifiers. The prediction of the multi-modal model was defined as the average of all possible prediction results among five for each patient. The performances of the models were compared using a leave-one-out approach.

Results

The percentage of missing modalities ranged from 6 to 11% across modalities and tasks. The performance of each individual model was dependent on each specific task, with an AUC of the ROC curve ranging from 0.63 to 0.80. The multi-modal model outperformed the clinical model for each prediction tasks, thus demonstrating the added value of MRI. Furthermore, regardless of performance criteria, the multi-modal model came in the first place or second place (very close to first). In the leave-one-out approach, the prediction of H3.1 (resp. ACVR1 and TP53) mutations achieved a balanced accuracy of 87.8% (resp. 82.1 and 78.3%).

Conclusion

Compared with a single modality approach, the multi-modal model combining multiple MRI modalities and clinical features was the most powerful to predict H3.1, ACVR1, and TP53 mutations and provided prediction, even in the case of missing modality. It could be proposed in the absence of a conclusive biopsy.

The Road Toward Reproducibility of Parametric Mapping of the Heart: A Technical Review

Parametric mapping of the heart has become an essential part of many cardiovascular magnetic resonance imaging exams, and is used for tissue characterization and diagnosis in a broad range of cardiovascular diseases. These pulse sequences are used to quantify the myocardial T₁, T₂, T2*, and T_1ρ relaxation times, which are unique surrogate indices of fibrosis, edema and iron deposition that can be used to monitor a disease over time or to compare patients to one another. Parametric mapping is now well-accepted in the clinical setting, but its wider dissemination is hindered by limited inter-center reproducibility and relatively long acquisition times. Recently, several new parametric mapping techniques have appeared that address both of these problems, but substantial hurdles remain for widespread clinical adoption. This review serves both as a primer for newcomers to the field of parametric mapping and as a technical update for those already well at home in it. It aims to establish what is currently needed to improve the reproducibility of parametric mapping of the heart. To this end, we first give an overview of the metrics by which a mapping technique can be assessed, such as bias and variability, as well as the basic physics behind the relaxation times themselves and what their relevance is in the prospect of myocardial tissue characterization. This is followed by a summary of routine mapping techniques and their variations. The problems in reproducibility and the sources of bias and variability of these techniques are reviewed. Subsequently, novel fast, whole-heart, and multi-parametric techniques and their merits are treated in the light of their reproducibility. This includes state of the art segmentation techniques applied to parametric maps, and how artificial intelligence is being harnessed to solve this long-standing conundrum. We finish up by sketching an outlook on the road toward inter-center reproducibility, and what to expect in the future.

MRSaiFE: An AI-based Approach Towards the Real-Time Prediction of Specific Absorption Rate

The purpose of this study is to investigate feasibility of estimating the specific absorption rate (SAR) in MRI in real time. To this goal, SAR maps are predicted from 3T- and 7T-simulated magnetic resonance (MR) images in 10 realistic human body models via a convolutional neural network. Two-dimensional (2-D) U-Net architectures with varying contraction layers and different convolutional filters were designed to estimate the SAR distribution in realistic body models. Sim4Life (ZMT, Switzerland) was used to create simulated anatomical images and SAR maps at 3T and 7T imaging frequencies for Duke, Ella, Charlie, and Pregnant Women (at 3, 7, and 9 month gestational stages) body models. Mean squared error (MSE) was used as the cost function and the structural similarity index (SSIM) was reported. A 2-D U-Net with 4 contracting (and 4 expanding) layers and 64 convolutional filters at the initial stage showed the best compromise to estimate SAR distributions. Adam optimizer outperformed stochastic gradient descent (SGD) for all cases with an average SSIM of 90.5∓3.6 % and an average MSE of 0.7∓0.6% for head images at 7T, and an SSIM of >85.1∓6.2 % and an MSE of 0.4∓0.4% for 3T body imaging. Algorithms estimated the SAR maps for 224×224 slices under 30 ms. The proposed methodology shows promise to predict real-time SAR in clinical imaging settings without using extra mapping techniques or patient-specific calibrations.