Search

Local progression of primary skull base chordoma (PSBC) is a sign of treatment failure. Predicting the postoperative progression of PSBC can aid in the development of individualized treatment plans to improve patients' progression-free survival (PFS) after surgery. This study aimed to develop a multiparametric MRI-based fusion radiomic model (FRM) and clinicoradiomic model (CRM) via radiomic and clinical analysis and to explore their validity in predicting postoperative progression in PSBC patients before surgery. In this retrospective study, a total of 129 patients with PSBC from our institution, including 57 patients with progression, were enrolled and randomized to the training set (TS) or the validation set (VS) at a 2:1 ratio. Radiomic features were extracted and dimensionally reduced from 3.0 T/axial T2-weighted imaging (T2WI), T1-weighted imaging (T1WI) and contrast-enhanced T1-weighted imaging (CE-T1WI) sequences for each patient, and the features were used for radiomic analysis. Univariate and multivariate Cox regression analyses were used to screen for key clinical factors. We constructed models on the basis of multivariate logistic regression analysis. Receiver operating characteristic (ROC) curve, calibration curve, and decision curve analyses were performed to evaluate the performance of the clinical model (CM), FRM and CRM. Through analysis, we found that blood supply was the only significantly different clinical factor in the CM. For the FRM, the area under the receiver operating characteristic curve (AUC) of the TS was 0.925, and the calibration curves were consistent across the TS. In the CRM, the AUC of the TS was 0.929, the calibration curve analysis was consistent for both the TS and the VS, and the DCA showed that the net benefit was greater at a threshold probability of > 0% for both the TS and the VS. Our proposed FRM can help clinicians better predict PSBC progression preoperatively, and the use of the CRM can lead to the development of more appropriate protocols to improve patients' PFS after surgery., Competing Interests: Declarations. Competing interests: The authors declare no competing interests., (© 2024. The Author(s).)

This study aims to develop and evaluate radiomics-based machine learning (ML) models for predicting meningioma grades using multiparametric magnetic resonance imaging (MRI). The study utilized the BraTS-MEN dataset's training split, including 698 patients (524 with grade 1 and 174 with grade 2-3 meningiomas). We extracted 4872 radiomic features from T1, T1 with contrast, T2, and FLAIR MRI sequences using PyRadiomics. LASSO regression reduced features to 176. The data was split into training (60%), validation (20%), and test (20%) sets. Five ML algorithms (TabPFN, XGBoost, LightGBM, CatBoost, and Random Forest) were employed to build models differentiating low-grade (grade 1) from high-grade (grade 2-3) meningiomas. Hyperparameter tuning was performed using Optuna, optimizing model-specific parameters and feature selection. The CatBoost model demonstrated the best performance, achieving an area under the receiver operating characteristic curve (AUROC) of 0.838 [95% confidence interval (CI): 0.689-0.935], precision of 0.492 (95% CI: 0.371-0.623), recall of 0.838 (95% CI: 0.689-0.935), F1 score of 0.620 (95% CI: 0.495-0.722), accuracy of 0.729 (95% CI: 0.650-0.800), an area under the precision-recall curve (AUPRC) of 0.620 (95% CI: 0.433-0.753), and Brier score of 0.156 (95% CI: 0.122-0.200). Other models showed comparable performance, with mean AUROCs ranging from 0.752 to 0.784. The radiomics-based ML approach presented in this study showcases the potential for non-invasive and pre-operative grading of meningiomas using multiparametric MRI. Further validation on larger and independent datasets is necessary to establish the robustness and generalizability of these findings., (© 2024. The Author(s).)

Kidney diseases pose a global healthcare burden, with millions requiring renal replacement therapy. Ischemia/reperfusion injury (IRI) is a common pathology of acute kidney injury, causing hypoxia and subsequent inflammation-induced kidney damage. Accurate detection of acute kidney injury due to IRI is crucial for timely intervention. We used longitudinal, multi-parametric magnetic resonance imaging (MRI) employing arterial spin labelling (ASL), diffusion weighted imaging (DWI), and dynamic contrast enhanced (DCE)-MRI to assess IRI induced changes in both the injured and healthy contralateral kidney, in a unilateral IRI mouse model (n = 9). Multi-parametric MRI demonstrated significant differences in kidney volume (p = 0.001), blood flow (p = 0.002), filtration coefficient (p = 0.038), glomerular filtration rate (p = 0.005) and apparent diffusion coefficient (p = 0.048) between the injured kidney and contralateral kidney on day 1 post-IRI surgery. Identification of the injured kidney using principal component analysis including most of the imaging parameters demonstrated an area under the curve (AUC) of 0.97. These results point to the utility of multi-parametric MRI in early detection of IRI-induced kidney damage suggesting that the combination of various MRI parameters may be suitable for monitoring the extent of injury in this model., (© 2024. The Author(s).)

In addition to focal lesions, diffusely abnormal white matter (DAWM) is seen on brain MRI of multiple sclerosis (MS) patients and may represent early or distinct disease processes. The role of MRI-observed DAWM is understudied due to a lack of automated assessment methods. Supervised deep learning (DL) methods are highly capable in this domain, but require large sets of labeled data. To overcome this challenge, a DL-based network (DAWM-Net) was trained using semi-supervised learning on a limited set of labeled data for segmentation of DAWM, focal lesions, and normal-appearing brain tissues on multiparametric MRI. DAWM-Net segmentation performance was compared to a previous intensity thresholding-based method on an independent test set from expert consensus (N = 25). Segmentation overlap by Dice Similarity Coefficient (DSC) and Spearman correlation of DAWM volumes were assessed. DAWM-Net showed DSC > 0.93 for normal-appearing brain tissues and DSC > 0.81 for focal lesions. For DAWM-Net, the DAWM DSC was 0.49 ± 0.12 with a moderate volume correlation (ρ = 0.52, p < 0.01). The previous method showed lower DAWM DSC of 0.26 ± 0.08 and lacked a significant volume correlation (ρ = 0.23, p = 0.27). These results demonstrate the feasibility of DL-based DAWM auto-segmentation with semi-supervised learning. This tool may facilitate future investigation of the role of DAWM in MS., (© 2024. The Author(s).)

Triple-negative breast cancer (TNBC) is often treated with neoadjuvant systemic therapy (NAST). We investigated if radiomic models based on multiparametric Magnetic Resonance Imaging (MRI) obtained early during NAST predict pathologic complete response (pCR). We included 163 patients with stage I-III TNBC with multiparametric MRI at baseline and after 2 (C2) and 4 cycles of NAST. Seventy-eight patients (48%) had pCR, and 85 (52%) had non-pCR. Thirty-six multivariate models combining radiomic features from dynamic contrast-enhanced MRI and diffusion-weighted imaging had an area under the receiver operating characteristics curve (AUC) > 0.7. The top-performing model combined 35 radiomic features of relative difference between C2 and baseline; had an AUC = 0.905 in the training and AUC = 0.802 in the testing set. There was high inter-reader agreement and very similar AUC values of the pCR prediction models for the 2 readers. Our data supports multiparametric MRI-based radiomic models for early prediction of NAST response in TNBC., (© 2024. The Author(s).)

Prostate cancer is one of the most common and fatal diseases among men, and its early diagnosis can have a significant impact on the treatment process and prevent mortality. Since it does not have apparent clinical symptoms in the early stages, it is difficult to diagnose. In addition, the disagreement of experts in the analysis of magnetic resonance images is also a significant challenge. In recent years, various research has shown that deep learning, especially convolutional neural networks, has appeared successfully in machine vision (especially in medical image analysis). In this research, a deep learning approach was used on multi-parameter magnetic resonance images, and the synergistic effect of clinical and pathological data on the accuracy of the model was investigated. The data were collected from Trita Hospital in Tehran, which included 343 patients (data augmentation and learning transfer methods were used during the process). In the designed model, four different types of images are analyzed with four separate ResNet50 deep convolutional networks, and their extracted features are transferred to a fully connected neural network and combined with clinical and pathological features. In the model without clinical and pathological data, the maximum accuracy reached 88%, but by adding these data, the accuracy increased to 96%, which shows the significant impact of clinical and pathological data on the accuracy of diagnosis., (© 2024. The Author(s).)

Prediction of glioma is crucial to provide a precise treatment plan to optimize the prognosis of children with glioma. However, studies on the grading of pediatric gliomas using radiomics are limited. Meanwhile, existing methods are mainly based on only radiomics features, ignoring intuitive information about tumor morphology on traditional imaging features. This study aims to utilize multiparametric magnetic resonance imaging (MRI) to identify high-grade and low-grade gliomas in children and establish a classification model based on radiomics features and clinical features. A total of 85 children with gliomas underwent tumor resection, and part of the tumor tissue was examined pathologically. Patients were categorized into high-grade and low-grade groups according to World Health Organization guidelines. Preoperative multiparametric MRI data, including contrast-enhanced T1-weighted imaging, T2-weighted imaging, T2-weighted fluid-attenuated inversion recovery, diffusion-weighted images, and apparent diffusion coefficient sequences, were obtained and labeled by two radiologists. The images were preprocessed, and radiomics features were extracted for each MRI sequence. Feature selection methods were used to select radiomics features, and statistically significant clinical features were identified using t-tests. The selected radiomics features and conventional MRI features were used to train the AutoGluon models. The improved model, based on radiomics features and conventional MRI features, achieved a balanced classification accuracy of 66.59%. The cross-validated areas under the receiver operating characteristic curve for the classifier of AutoGluon frame were 0.8071 on the test dataset. The results indicate that the performance of AutoGluon models can be improved by incorporating conventional MRI features, highlighting the importance of the experience of radiologists in accurately grading pediatric gliomas. This method can help predict the grade of pediatric glioma before pathological examination and assist in determining the appropriate treatment plan, including radiotherapy, chemotherapy, drugs, and gene surgery., (© 2024. The Author(s).)

This study aimed to differentiate isocitrate dehydrogenase (IDH) mutation status with the voxel-wise clustering method of multiparametric magnetic resonance imaging (MRI) and to discover biological underpinnings of the clusters. A total of 69 patients with treatment-naïve diffuse glioma were scanned with pH-sensitive amine chemical exchange saturation transfer MRI, diffusion-weighted imaging, fluid-attenuated inversion recovery, and contrast-enhanced T1-weighted imaging at 3 T. An unsupervised two-level clustering approach was used for feature extraction from acquired images. The logarithmic ratio of the labels in each class within tumor regions was applied to a support vector machine to differentiate IDH status. The highest performance to predict IDH mutation status was found for 10-class clustering, with a mean area under the curve, accuracy, sensitivity, and specificity of 0.94, 0.91, 0.90, and 0.91, respectively. Targeted biopsies revealed that the tissues with labels 7-10 showed high expression levels of hypoxia-inducible factor 1-alpha, glucose transporter 3, and hexokinase 2, which are typical of IDH wild-type glioma, whereas those with labels 1 showed low expression of these proteins. In conclusion, A machine learning model successfully predicted the IDH mutation status of gliomas, and the resulting clusters properly reflected the metabolic status of the tumors., (© 2022. The Author(s).)

Stereotactic radiosurgery planning for cerebral arteriovenous malformations (AVM) is complicated by the variability in appearance of an AVM nidus across different imaging modalities. We developed a deep learning approach to automatically segment cerebrovascular-anatomical maps from multiple high-resolution magnetic resonance imaging/angiography (MRI/MRA) sequences in AVM patients, with the goal of facilitating target delineation. Twenty-three AVM patients who were evaluated for radiosurgery and underwent multi-parametric MRI/MRA were included. A hybrid semi-automated and manual approach was used to label MRI/MRAs with arteries, veins, brain parenchyma, cerebral spinal fluid (CSF), and embolized vessels. Next, these labels were used to train a convolutional neural network to perform this task. Imaging from 17 patients (6362 image slices) was used for training, and 6 patients (1224 slices) for validation. Performance was evaluated by Dice Similarity Coefficient (DSC). Classification performance was good for arteries, veins, brain parenchyma, and CSF, with DSCs of 0.86, 0.91, 0.98, and 0.91, respectively in the validation image set. Performance was lower for embolized vessels, with a DSC of 0.75. This demonstrates the proof of principle that accurate, high-resolution cerebrovascular-anatomical maps can be generated from multiparametric MRI/MRA. Clinical validation of their utility in radiosurgery planning is warranted., (© 2022. The Author(s).)

Magnetic resonance imaging (MRI) is playing an increasingly important role in evaluating chronic kidney disease (CKD). It has the potential to be used not only for evaluation of physiological and pathological states, but also for prediction of disease course. Although different MRI sequences have been employed in renal disease, there are few studies that have compared the different sequences. We compared several multiparametric MRI sequences, and compared their results with the estimated glomerular filtration rate. Principal component analysis showed a similarity between T1 values and tissue perfusion (arterial spin labelling), and between fractional anisotropy (diffusion tensor imaging) and apparent diffusion coefficient values (diffusion-weighted imaging). In multiple regression analysis, only T2* values, derived from the blood oxygenation level-dependent (BOLD) MRI sequence, were associated with estimated glomerular filtration rate slope after adjusting for degree of proteinuria, a classic prognostic factor for CKD. In receiver operating characteristic curve analysis, T2* values were a good predictor of rapid deterioration, regardless of the degree of proteinuria. This suggests further study of the use of BOLD-derived T2* values in the workup of CKD, especially to predict the disease course., (© 2021. The Author(s).)

Amyloid plaques are a hallmark of Alzheimer's disease (AD) that develop in its earliest stages. Thus, non-invasive detection of these plaques would be invaluable for diagnosis and the development and monitoring of treatments, but this remains a challenge due to their small size. Here, we investigated the utility of manganese-enhanced MRI (MEMRI) for visualizing plaques in transgenic rodent models of AD across two species: 5xFAD mice and TgF344-AD rats. Animals were given subcutaneous injections of MnCl 2 and imaged in vivo using a 9.4 T Bruker scanner. MnCl 2 improved signal-to-noise ratio but was not necessary to detect plaques in high-resolution images. Plaques were visible in all transgenic animals and no wild-types, and quantitative susceptibility mapping showed that they were more paramagnetic than the surrounding tissue. This, combined with beta-amyloid and iron staining, indicate that plaque MR visibility in both animal models was driven by plaque size and iron load. Longitudinal relaxation rate mapping revealed increased manganese uptake in brain regions of high plaque burden in transgenic animals compared to their wild-type littermates. This was limited to the rhinencephalon in the TgF344-AD rats, while it was most significantly increased in the cortex of the 5xFAD mice. Alizarin Red staining suggests that manganese bound to plaques in 5xFAD mice but not in TgF344-AD rats. Multi-parametric MEMRI is a simple, viable method for detecting amyloid plaques in rodent models of AD. Manganese-induced signal enhancement can enable higher-resolution imaging, which is key to visualizing these small amyloid deposits. We also present the first in vivo evidence of manganese as a potential targeted contrast agent for imaging plaques in the 5xFAD model of AD.

High-grade gliomas are an aggressive and invasive malignancy which are susceptible to treatment resistance due to heterogeneity in intratumoral properties such as cell proliferation and density and perfusion. Non-invasive imaging approaches can measure these properties, which can then be used to calibrate patient-specific mathematical models of tumor growth and response. We employed multiparametric magnetic resonance imaging (MRI) to identify tumor extent (via contrast-enhanced T 1 -weighted, and T 2 -FLAIR) and capture intratumoral heterogeneity in cell density (via diffusion-weighted imaging) to calibrate a family of mathematical models of chemoradiation response in nine patients with unresected or partially resected disease. The calibrated model parameters were used to forecast spatially-mapped individual tumor response at future imaging visits. We then employed the Akaike information criteria to select the most parsimonious member from the family, a novel two-species model describing the enhancing and non-enhancing components of the tumor. Using this model, we achieved low error in predictions of the enhancing volume (median: - 2.5%, interquartile range: 10.0%) and a strong correlation in total cell count (Kendall correlation coefficient 0.79) at 3-months post-treatment. These preliminary results demonstrate the plausibility of using multiparametric MRI data to inform spatially-informative, biologically-based predictive models of tumor response in the setting of clinical high-grade gliomas.

With the rapid growth and increasing use of brain MRI, there is an interest in automated image classification to aid human interpretation and improve workflow. We aimed to train a deep convolutional neural network and assess its performance in identifying abnormal brain MRIs and critical intracranial findings including acute infarction, acute hemorrhage and mass effect. A total of 13,215 clinical brain MRI studies were categorized to training (74%), validation (9%), internal testing (8%) and external testing (8%) datasets. Up to eight contrasts were included from each brain MRI and each image volume was reformatted to common resolution to accommodate for differences between scanners. Following reviewing the radiology reports, three neuroradiologists assigned each study to abnormal vs normal, and identified three critical findings including acute infarction, acute hemorrhage, and mass effect. A deep convolutional neural network was constructed by a combination of localization feature extraction (LFE) modules and global classifiers to identify the presence of 4 variables in brain MRIs including abnormal, acute infarction, acute hemorrhage and mass effect. Training, validation and testing sets were randomly defined on a patient basis. Training was performed on 9845 studies using balanced sampling to address class imbalance. Receiver operating characteristic (ROC) analysis was performed. The ROC analysis of our models for 1050 studies within our internal test data showed AUC/sensitivity/specificity of 0.91/83%/86% for normal versus abnormal brain MRI, 0.95/92%/88% for acute infarction, 0.90/89%/81% for acute hemorrhage, and 0.93/93%/85% for mass effect. For 1072 studies within our external test data, it showed AUC/sensitivity/specificity of 0.88/80%/80% for normal versus abnormal brain MRI, 0.97/90%/97% for acute infarction, 0.83/72%/88% for acute hemorrhage, and 0.87/79%/81% for mass effect. Our proposed deep convolutional network can accurately identify abnormal and critical intracranial findings on individual brain MRIs, while addressing the fact that some MR contrasts might not be available in individual studies.

Despite the key-role of the Prostate Imaging and Reporting and Data System (PI-RADS) in the diagnosis and characterization of prostate cancer (PCa), this system remains to be affected by several limitations, primarily associated with the interpretation of equivocal PI-RADS 3 lesions and with the debated role of Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI), which is only used to upgrade peripheral PI-RADS category 3 lesions to PI-RADS category 4 if enhancement is focal. We aimed at investigating the usefulness of radiomics for detection of PCa lesions (Gleason Score ≥ 6) in PI-RADS 3 lesions and in peripheral PI-RADS 3 upgraded to PI-RADS 4 lesions (upPI-RADS 4). Multiparametric MRI (mpMRI) data of patients who underwent prostatic mpMRI between April 2013 and September 2018 were retrospectively evaluated. Biopsy results were used as gold standard. PI-RADS 3 and PI-RADS 4 lesions were re-scored according to the PI-RADS v2.1 before and after DCE-MRI evaluation. Radiomic features were extracted from T2-weighted MRI (T2), Apparent diffusion Coefficient (ADC) map and DCE-MRI subtracted images using PyRadiomics. Feature selection was performed using Wilcoxon-ranksum test and Minimum Redundancy Maximum Relevance (mRMR). Predictive models were constructed for PCa detection in PI-RADS 3 and upPI-RADS 4 lesions using at each step an imbalance-adjusted bootstrap resampling (IABR) on 1000 samples. 41 PI-RADS 3 and 32 upPI-RADS 4 lesions were analyzed. Among 293 radiomic features, the top selected features derived from T2 and ADC. For PI-RADS 3 stratification, second order model showed higher performances (Area Under the Receiver Operating Characteristic Curve-AUC- = 80%), while for upPI-RADS 4 stratification, first order model showed higher performances respect to superior order models (AUC = 89%). Our results support the significant role of T2 and ADC radiomic features for PCa detection in lesions scored as PI-RADS 3 and upPI-RADS 4. Radiomics models showed high diagnostic efficacy in classify PI-RADS 3 and upPI-RADS 4 lesions, outperforming PI-RADS v2.1 performance.

Differentiating pseudoprogression from true tumor progression has become a significant challenge in follow-up of diffuse infiltrating gliomas, particularly high grade, which leads to a potential treatment delay for patients with early glioma recurrence. In this study, we proposed to use a multiparametric MRI data as a sequence input for the convolutional neural network with the recurrent neural network based deep learning structure to discriminate between pseudoprogression and true tumor progression. In this study, 43 biopsy-proven patient data identified as diffuse infiltrating glioma patients whose disease progressed/recurred were used. The dataset consists of five original MRI sequences; pre-contrast T1-weighted, post-contrast T1-weighted, T2-weighted, FLAIR, and ADC images as well as two engineered sequences; T1post-T1pre and T2-FLAIR. Next, we used three CNN-LSTM models with a different set of sequences as input sequences to pass through CNN-LSTM layers. We performed threefold cross-validation in the training dataset and generated the boxplot, accuracy, and ROC curve, AUC from each trained model with the test dataset to evaluate models. The mean accuracy for VGG16 models ranged from 0.44 to 0.60 and the mean AUC ranged from 0.47 to 0.59. For CNN-LSTM model, the mean accuracy ranged from 0.62 to 0.75 and the mean AUC ranged from 0.64 to 0.81. The performance of the proposed CNN-LSTM with multiparametric sequence data was found to outperform the popular convolutional CNN with a single MRI sequence. In conclusion, incorporating all available MRI sequences into a sequence input for a CNN-LSTM model improved diagnostic performance for discriminating between pseudoprogression and true tumor progression.

Non-alcoholic steatohepatitis (NASH) is major health burden lacking effective pharmacological therapies. Clinical trials enrol patients with histologically-defined NAFLD (non-alcoholic fatty liver disease) activity score (NAS) ≥ 4 and Kleiner-Brunt fibrosis stage (F) ≥ 2; however, screen failure rates are often high following biopsy. This study evaluated a non-invasive MRI biomarker, iron-corrected T1 mapping (cT1), as a diagnostic pre-screening biomarker for NASH. In a retrospective analysis of 86 biopsy confirmed NAFLD patients we explored the potential of blood and imaging biomarkers, both in isolation and in combination, to discriminate those who have NAS ≥ 4 and F ≥ 2 from those without. Stepwise logistic regression was performed to select the optimal combination of biomarkers, diagnostic accuracy was determined using area under the receiver operator curve and model validated confirmed with and fivefold cross-validation. Results showed that levels of cT1, AST, GGT and fasting glucose were all good predictors of NAS ≥ 4 and F ≥ 2, and the model identified the combination of cT1-AST-fasting glucose (cTAG) as far superior to any individual biomarker (AUC 0.90 [0.84-0.97]). This highlights the potential utility of the composite cTAG score for screening patients prior to biopsy to identify those suitable for NASH clinical trial enrolment.

The aim of this prospective single-institution clinical trial (NCT02002455) was to evaluate the potential of advanced post-processing methods for 18 F-Fluciclovine PET and multisequence multiparametric MRI in the prediction of prostate cancer (PCa) aggressiveness, defined by Gleason Grade Group (GGG). 21 patients with PCa underwent PET/CT, PET/MRI and MRI before prostatectomy. DWI was post-processed using kurtosis (ADC k , K), mono- (ADC m ), and biexponential functions (f, D p , D f ) while Logan plots were used to calculate volume of distribution (V T ). In total, 16 unique PET (V T , SUV) and MRI derived quantitative parameters were evaluated. Univariate and multivariate analysis were carried out to estimate the potential of the quantitative parameters and their combinations to predict GGG 1 vs >1, using logistic regression with a nested leave-pair out cross validation (LPOCV) scheme and recursive feature elimination technique applied for feature selection. The second order rotating frame imaging (RAFF), monoexponential and kurtosis derived parameters had LPOCV AUC in the range of 0.72 to 0.92 while the corresponding value for V T was 0.85. T he best performance for GGG prediction was achieved by K parameter of kurtosis function followed by quantitative parameters based on DWI, RAFF and 18 F-FACBC PET. No major improvement was achieved using parameter combinations with or without feature selection. Addition of 18 F-FACBC PET derived parameters (V T , SUV) to DWI and RAFF derived parameters did not improve LPOCV AUC.

Glioblastoma (GBM) is a heterogeneous and lethal brain cancer. These tumors are followed using magnetic resonance imaging (MRI), which is unable to precisely identify tumor cell invasion, impairing effective surgery and radiation planning. We present a novel hybrid model, based on multiparametric intensities, which combines machine learning (ML) with a mechanistic model of tumor growth to provide spatially resolved tumor cell density predictions. The ML component is an imaging data-driven graph-based semi-supervised learning model and we use the Proliferation-Invasion (PI) mechanistic tumor growth model. We thus refer to the hybrid model as the ML-PI model. The hybrid model was trained using 82 image-localized biopsies from 18 primary GBM patients with pre-operative MRI using a leave-one-patient-out cross validation framework. A Relief algorithm was developed to quantify relative contributions from the data sources. The ML-PI model statistically significantly outperformed (p < 0.001) both individual models, ML and PI, achieving a mean absolute predicted error (MAPE) of 0.106 ± 0.125 versus 0.199 ± 0.186 (ML) and 0.227 ± 0.215 (PI), respectively. Associated Pearson correlation coefficients for ML-PI, ML, and PI were 0.838, 0.518, and 0.437, respectively. The Relief algorithm showed the PI model had the greatest contribution to the result, emphasizing the importance of the hybrid model in achieving the high accuracy.

In this study we assessed the repeatability of radiomics features on small prostate tumors using test-retest Multiparametric Magnetic Resonance Imaging (mpMRI). The premise of radiomics is that quantitative image-based features can serve as biomarkers for detecting and characterizing disease. For such biomarkers to be useful, repeatability is a basic requirement, meaning its value must remain stable between two scans, if the conditions remain stable. We investigated repeatability of radiomics features under various preprocessing and extraction configurations including various image normalization schemes, different image pre-filtering, and different bin widths for image discretization. Although we found many radiomics features and preprocessing combinations with high repeatability (Intraclass Correlation Coefficient > 0.85), our results indicate that overall the repeatability is highly sensitive to the processing parameters. Neither image normalization, using a variety of approaches, nor the use of pre-filtering options resulted in consistent improvements in repeatability. We urge caution when interpreting radiomics features and advise paying close attention to the processing configuration details of reported results. Furthermore, we advocate reporting all processing details in radiomics studies and strongly recommend the use of open source implementations.

Multiparametric magnetic resonance imaging (mpMRI) has become increasingly important for the clinical assessment of prostate cancer (PCa), but its interpretation is generally variable due to its relatively subjective nature. Radiomics and classification methods have shown potential for improving the accuracy and objectivity of mpMRI-based PCa assessment. However, these studies are limited to a small number of classification methods, evaluation using the AUC score only, and a non-rigorous assessment of all possible combinations of radiomics and classification methods. This paper presents a systematic and rigorous framework comprised of classification, cross-validation and statistical analyses that was developed to identify the best performing classifier for PCa risk stratification based on mpMRI-derived radiomic features derived from a sizeable cohort. This classifier performed well in an independent validation set, including performing better than PI-RADS v2 in some aspects, indicating the value of objectively interpreting mpMRI images using radiomics and classification methods for PCa risk assessment.