The Image Biomarker Standardization Initiative: Standardized Convolutional Filters for Reproducible Radiomics and Enhanced Clinical Insights

Artificial Intelligence for Drug Discovery: An Update and Future Prospects

Artificial intelligence (AI) has become a pivotal tool for medical image analysis, significantly enhancing drug discovery through improved diagnostics, staging, prognostication, and response assessment. At a high level, AI-driven image analysis enables the quantification and synthesis of previously qualitative imaging characteristics, facilitating the identification of novel disease-specific biomarkers, patient risk stratification, prognostication, and adverse event prediction. In addition, AI can assist in response assessment by capturing changes in imaging "phenotype" over time, allowing for optimized treatment plans based on real-time analysis. Integrating this emerging technology into drug discovery pipelines has the potential to accelerate the identification and development of new pharmaceuticals by assisting in target identification and patient selection, as well as reducing the incidence, and therefore cost, of failed trials through high-throughput, reproducible, and data-driven insights. Continued progress in AI applications will shape the future of medical imaging, ultimately fostering more efficient, accurate, and tailored drug discovery processes. Herein, we offer a comprehensive overview of how AI enhances medical imaging to inform drug development and therapeutic strategies.

Machine learning based radiomic models outperform clinical biomarkers in predicting outcomes after immunotherapy for hepatocellular carcinoma

BACKGROUND

Atezolizumab plus bevacizumab (A/B) is a first-line therapy for unresectable hepatocellular carcinoma (HCC). Only a small proportion of patients respond to treatment. This study integrated radiomic and clinical data derived from routine pre-treatment imaging to predict outcomes after immunotherapy.

METHODS

152 patients from two international centres receiving A/B were retrospectively reviewed. Deep learning autosegmentation generated whole liver masks from pre-treatment CTs. Radiomic features combined with clinical variables were used to predict 12-month mortality post A/B. Radiomic and integrated radiomic-clinical models were developed using 7 machine learning models in combination with 13 feature selection techniques in the Imperial College London (ICL) cohort. K-means clustering identified high- and low-risk groups and predicted overall survival (OS), progression-free survival (PFS) and response. Model performance was assessed in the independent Assistance Publique-Hôpitaux de Paris (AP-HP) cohort.

RESULTS

The integrated radiomic-clinical model outperformed BCLC stage (AUC 0.61, p<0.001) and ALBI grade (AUC 0.48, p<0.001) in ICL (AUC 0.89, 95% CI 0.75-0.99) and AP-HP (AUC 0.75, 95% CI 0.64-0.85) cohorts. Integrated model-stratified high-risk patients had significantly shorter median OS (ICL: 5.6 months vs. 28.2 months; p<0.001; AP-HP: 5.8 months vs. 15.7 months; p<0.001) and PFS (ICL: 2.4 months vs. 14.6 months; p<0.001; AP-HP: 2.1 months vs. 6.1 months; p=0.046). Low-risk patients had significantly higher ICI response rates compared to high-risk patients (35.6% vs. 21.4%; p=0.038). In multivariable analysis, radiomic group was the strongest predictor of OS (HR 3.22, 95% CI 1.99-5.20; p<0.001) and PFS (HR 1.82, 95% CI 1.18-2.80; p=0.010).

CONCLUSION

Radiomic-based models predict survival outcomes and response to immunotherapy in patients with advanced HCC. Deep learning in combination with machine learning can stratify patients and allows for precision treatment strategies.

IMPACT AND IMPLICATIONS

There is a lack of prognostic markers predicting survival and response after immunotherapy in hepatocellular carcinoma. This study used deep learning and machine learning to develop and validate an integrated radiomic-clinical model which can predict survival and response to atezolizumab plus bevacizumab from pre-treatment imaging. Radiomic-based machine learning models can risk-stratify advanced HCC patients receiving atezolizumab plus bevacizumab.

Quality of radiomics research: comprehensive analysis of 1574 unique publications from 89 reviews

PURPOSE

This study aims to comprehensively evaluate the quality of radiomics research by examining unique papers from reviews using the radiomics quality score (RQS).

METHODS

A literature search was conducted in PubMed (last search date: April 14, 2024). Systematic or non-systematic reviews using the RQS to evaluate radiomic studies were potentially included. Exclusion was applied at two levels: first, at the review level, and second, at the study level (i.e., for the individual articles previously evaluated within the reviews). Score-wise and item-wise analyses were performed, along with trend, multivariable, and subgroup analyses based on baseline study characteristics and validation methods.

RESULTS

A total of 1574 unique papers (published online between 1999 and 2023) from 89 reviews were included in the final analysis. The median RQS percentage was 31% with an IQR of 25% (25th-75th percentiles, 14-39%). A positive correlation between median RQS percentage and publication year (2014-2023) was found, with Kendall's tau coefficient of 0.908 (p < 0.001), suggesting an improvement in quality over time. The quality of radiomics publications significantly varied according to different subfields of radiology (p < 0.001). Around one-third of the publications (32%) lacked a separate validation set. Papers with internal validation (54%) dominated those with external validation (14%). Higher-quality validation practices were significantly associated with better RQS percentage scores, independent of the validation's effect on the final score. Item-wise analysis revealed significant shortcomings in several areas.

CONCLUSION

Radiomics research quality is low but improving according to RQS. Significant variation exists across radiology subfields. Critical areas were identified for targeted improvement.

CLINICAL RELEVANCE STATEMENT

Our study shows that the quality of radiomics research is generally low but improving over time, with item-wise analysis highlighting critical areas needing improvement. It also reveals that the quality of radiomics research differs across subfields and validation methods.

KEY POINTS

Overall quality of radiomics research remains low and highly variable, although a significant positive trend suggests an improvement in quality over time. Considerable variations exist in the quality of radiomics publications across different subfields of radiology and validation types. The item-wise analysis highlights several critical areas requiring attention, emphasizing the need for targeted improvements.

Diagnostic Performance of Combined Conventional CT Imaging Features and Radiomics Signature in Differentiating Grade 1 Tumors from Higher-Grade Pancreatic Neuroendocrine Neoplasms

BACKGROUND/OBJECTIVES

Accurate identification of grade 1 (G1) pancreatic neuroendocrine tumors (PanNETs) is crucial due to their rising incidence and emerging nonsurgical management strategies. This study evaluated whether combining conventional CT imaging features, CT radiomics features, and clinical data improves differentiation of G1 PanNETs from higher-grade tumors (G2/G3 PanNETs and pancreatic neuroendocrine carcinomas [PanNECs]) compared to using these features individually.

METHODS

A retrospective analysis included 133 patients with pathologically confirmed PanNETs or PanNECs (70 males, 63 females; mean age, 58.5 years) who underwent pancreas protocol CT. A total of 28 conventional imaging features, 4892 radiomics features, and clinical data (age, gender, and tumor location) were analyzed using a support vector machine (SVM) model. Data were divided into 70% training and 30% testing sets.

RESULTS

The SVM model using the top 10 conventional imaging features (e.g., suspicious lymph nodes and hypoattenuating tumors) achieved 75% sensitivity, 81% specificity, and 79% accuracy for identifying higher-grade tumors (G2/G3 PanNETs and PanNECs). The top 10 radiomics features yielded 94% sensitivity, 46% specificity, and 69% accuracy. Combining all features (imaging, radiomics, and clinical data) improved performance, with 94% sensitivity, 69% specificity, 79% accuracy, and an F1-score of 0.77. The radiomics score demonstrated an AUC of 0.85 in the training and 0.83 in the testing set.

CONCLUSIONS

Conventional imaging features provided higher specificity, while radiomics offered greater sensitivity for identifying higher-grade tumors. Integrating all three features improved diagnostic accuracy, highlighting their complementary roles. This combined model may serve as a valuable tool for distinguishing higher-grade tumors from G1 PanNETs and potentially guiding patient management.

Integration of intratumoral and peritumoral CT radiomic features with machine learning algorithms for predicting induction therapy response in locally advanced non-small cell lung cancer

OBJECTIVES

To extract intratumoral, peritumoral, and integrated intratumoral-peritumoral CT radiomic features, develop multi-source radiomic models using various machine learning algorithms to identify the optimal model, and integrate clinical factors to establish a nomogram for predicting the therapeutic response to induction therapy(IT) in locally advanced non-small cell lung cancer.

METHODS

This study included 209 patients with locally advanced non-small cell lung cancer (LA-NSCLC) who received IT as the training cohort, and an external validation cohort comprising 50 patients from another center. Radiomic features were extracted from intratumoral, peritumoral, and integrated intratumoral-peritumoral regions by manually delineating the gross tumor volume (GTV) and an additional 3 mm surrounding area. Three machine learning algorithms-Support Vector Machine (SVM), XGBoost, and Gradient Boosting-were employed to construct radiomic models for each region. Model performance was evaluated in the external validation cohort using metrics such as Area Under the Curve (AUC), confusion matrix, accuracy, precision, recall, and F1 score. Finally, a comprehensive nomogram integrating the optimal radiomic model with independent clinical predictors was developed.

RESULTS

Through a comparison of optimal machine learning algorithms, INTRAPERI, INTRA, and PERI achieved the best performance with Gradient Boosting, SVM, and XGBoost, respectively. Compared to the INTRA_SVM and PERI_XGBoost INTRA models, the fusion model that integrates INTRA and peritumoral regions within a 3 mm margin around the tumor (INTRAPERI_GradientBoosting) showed better predictive performance in the training set, with AUCs of 93.7%, 82.5%, and 89.4%, respectively. In the clinical model, the PS score was identified as an independent predictive factor. The nomogram combining clinical factors with the INTRAPERI_GradientBoosting score demonstrated clinical predictive value.

CONCLUSION

The INTRAPERI_GradientBoosting model, which integrates intra-tumoral and peritumoral features, performs better than the INTRA intra-tumoral and PERI peritumoral radiomics models in predicting the efficacy of IT therapy in LA-NSCLC. Additionally, the nomogram based on INTRAPERI intra-tumoral and peritumoral features combined with independent clinical predictors has clinical predictive value.

Robust vs. Non-robust radiomic features: the quest for optimal machine learning models using phantom and clinical studies

PURPOSE

This study aimed to select robust features against lung motion in a phantom study and use them as input to feature selection algorithms and machine learning classifiers in a clinical study to predict the lymphovascular invasion (LVI) of non-small cell lung cancer (NSCLC). The results of robust features were also compared with conventional techniques without considering the robustness of radiomic features.

METHODS

An in-house developed lung phantom was developed with two 22mm lesion sizes based on a clinical study. A specific motor was built to simulate motion in two orthogonal directions. Lesions of both clinical and phantom studies were segmented using a Fuzzy C-means-based segmentation algorithm. After inducing motion and extracting 105 radiomic features in 4 feature sets, including shape, first-, second-, and higher-order statistics features from each region of interest (ROI) of the phantom image, statistical analyses were performed to select robust features against motion. Subsequently, these robust features and a total of 105 radiomic features were extracted from 126 clinical data. Various feature selection (FS) and multiple machine learning (ML) classifiers were implemented to predict the LVI of NSCLC, followed by comparing the results of predicting LVI using robust features with common conventional techniques not considering the robustness of radiomic features.

RESULTS

Our results demonstrated that selecting robust features as input to FS algorithms and ML classifiers surges the sensitivity, which has a gentle negative effect on the accuracy and the area under the curve (AUC) of predictions compared with commonly used methods in 12 of 15 outcomes. The top performance of the LVI prediction was achieved by the NB classifier and RFE FS without considering the robustness of radiomic features with 95% area under the curve of AUC, 67% accuracy, and 100% sensitivity. Moreover, the top performance of the LVI prediction using robust features belonged to the NB classifier and Boruta feature selection with 92% AUC, 86% accuracy, and 100% sensitivity.

CONCLUSION

Robustness over various influential factors is critical and should be considered in a radiomic study. Selecting robust features is a solution to overcome the low reproducibility of radiomic features. Although setting robust features against motion in a phantom study has a minor negative impact on the accuracy and AUC of LVI prediction, it boosts the sensitivity of prediction to a large extent.

Transfer learning radiomic model predicts intratumoral tertiary lymphoid structures in hepatocellular carcinoma: a multicenter study

BACKGROUND

Intratumoral tertiary lymphoid structures (iTLS) in hepatocellular carcinoma (HCC) are associated with improved survival and may influence treatment decisions. However, their non-invasive detection remains challenging in HCC. We aim to develop a non-invasive model using baseline contrast-enhanced MRI to predict the iTLS status.

METHODS

A total of 660 patients with HCC who underwent surgery were retrospectively recruited from four centers between October 2015 and January 2023 and divided into training, internal test, and external validation sets. After features dimensionality and selection, corresponding features were used to construct transfer learning radiomic (TLR) models for diagnosing iTLS, and model interpretability was explored with pathway analysis in The Cancer Genome Atlas-Liver HCC. The performances of models were assessed using the area under the receiver operating characteristic curve (AUC). The log-rank test was used to evaluate the prognostic value of the TLR model. The combination therapy set of 101 patients with advanced HCC treated with first-line anti-programmed death 1 or ligand 1 plus antiangiogenic treatment between January 2021 and January 2024 was used to investigate the value of the TLR model for evaluating the treatment response.

RESULTS

The presence of iTLS was identified in 46.0% (n=308) patients. The TLR model demonstrated excellent performance in predicting the presence of iTLS in training (AUC=0.91, 95% CI: 0.87, 0.94), internal test (AUC=0.85, 95% CI: 0.77, 0.93) and external validation set (AUC=0.85, 95% CI: 0.81, 0.90). The TLR model-predicted iTLS group has favorable overall survival (HR=0.66; 95% CI: 0.48, 0.90; p=0.007) and relapse-free survival (HR=0.64; 95% CI: 0.48, 0.85; p=0.001) in the external validation set. The model-predicted iTLS status was associated with inflammatory response and specific tumor-associated signaling activation (all p<0.001). The proportion of treatment responders was significantly higher in the model-predicted group with iTLS than in the group without iTLS (36% vs 13.73%, p=0.009).

CONCLUSION

The TLR model has indicated accurate prediction of iTLS status, which may assist in the risk stratification for patients with HCC in clinical practice.

ESR Essentials: radiomics-practice recommendations by the European Society of Medical Imaging Informatics

Radiomics is a method to extract detailed information from diagnostic images that cannot be perceived by the naked eye. Although radiomics research carries great potential to improve clinical decision-making, its inherent methodological complexities make it difficult to comprehend every step of the analysis, often causing reproducibility and generalizability issues that hinder clinical adoption. Critical steps in the radiomics analysis and model development pipeline-such as image, application of image filters, and selection of feature extraction parameters-can greatly affect the values of radiomic features. Moreover, common errors in data partitioning, model comparison, fine-tuning, assessment, and calibration can reduce reproducibility and impede clinical translation. Clinical adoption of radiomics also requires a deep understanding of model explainability and the development of intuitive interpretations of radiomic features. To address these challenges, it is essential for radiomics model developers and clinicians to be well-versed in current best practices. Proper knowledge and application of these practices is crucial for accurate radiomics feature extraction, robust model development, and thorough assessment, ultimately increasing reproducibility, generalizability, and the likelihood of successful clinical translation. In this article, we have provided researchers with our recommendations along with practical examples to facilitate good research practices in radiomics. KEY POINTS: Radiomics' inherent methodological complexity should be understood to ensure rigorous radiomic model development to improve clinical decision-making. Adherence to radiomics-specific checklists and quality assessment tools ensures methodological rigor. Use of standardized radiomics tools and best practices enhances clinical translation of radiomics models.

Evaluating the impact of the Radiomics Quality Score: a systematic review and meta-analysis

OBJECTIVES

Conduct a systematic review and meta-analysis on the application of the Radiomics Quality Score (RQS).

MATERIALS AND METHODS

A search was conducted from January 1, 2022, to December 31, 2023, for systematic reviews which implemented the RQS. Identification of articles prior to 2022 was via a previously published review. Quality scores of individual radiomics papers, their associated criteria scores, and these scores from all readers were extracted. Errors in the application of RQS criteria were noted and corrected. The RQS of radiomics papers were matched with the publication date, imaging modality, and country, where available.

RESULTS

A total of 130 systematic reviews were included, and individual quality scores 117/130 (90.0%), criteria scores 98/130 (75.4%), and multiple reader data 24/130 (18.5%) were extracted. 3258 quality scores were correlated with the radiomics study date of publication. Criteria scoring errors were discovered in 39/98 (39.8%) of articles. Overall mean RQS was 9.4 ± 6.4 (95% CI, 9.1-9.6) (26.1% ± 17.8% (25.3%-26.7%)). Quality scores were positively correlated with publication year (Pearson R = 0.32, p < 0.01) and significantly higher after publication of the RQS (year < 2018, 5.6 ± 6.1 (5.1-6.1); year ≥ 2018, 10.1 ± 6.1 (9.9-10.4); p < 0.01). Only 233/3258 (7.2%) scores were ≥ 50% of the maximum RQS. Quality scores were significantly different across imaging modalities (p < 0.01). Ten criteria were positively correlated with publication year, and one was negatively correlated.

CONCLUSION

Radiomics study adherence to the RQS is increasing with time, although a vast majority of studies are developmental and rarely provide a high level of evidence to justify the clinical translation of proposed models.

KEY POINTS

Question What level of adherence to the Radiomics Quality Score have radiomics studies achieved to date, has it increased with time, and is it sufficient? Findings A meta-analysis of 3258 quality scores extracted from 130 review articles resulted in a mean score of 9.4 ± 6.4. Quality scores were positively correlated with time. Clinical relevance Although quality scores of radiomics studies have increased with time, many studies have not demonstrated sufficient evidence for clinical translation. As new appraisal tools emerge, the current role of the Radiomics Quality Score may change.

Adapting to evolving MRI data: A transfer learning approach for Alzheimer's disease prediction

Integrating 3D magnetic resonance imaging (MRI) with machine learning has shown promising results in healthcare, especially in detecting Alzheimer's Disease (AD). However, changes in MRI technologies and acquisition protocols often yield limited data, leading to potential overfitting. This study explores Transfer Learning (TL) approaches to enhance AD diagnosis using a Baseline model consisting of a 3D-Convolutional Neural Network trained on 80 3T MRI scans. Two scenarios are explored: (A) utilizing historical data to address changes in MRI acquisitions (from 1.5T to 3T MRI), and (B) adapting 2D models pre-trained on ImageNet (ResNet18, ResNet50, ResNet101) for 3D image processing when historical data is unavailable. In both scenarios, two modeling approaches are tested. The General Approach involves distinct feature extraction and classification steps, using Radiomic features and TL-based features evaluated with six classifiers. The Deep Approach integrates these steps by fine-tuning the pre-trained models for AD diagnosis. In scenario (A), TL significantly boosts the Baseline's accuracy from 63% to 99%. In scenario (B), Radiomic features better represents 3D MRI than TL-features in the General Approach. Nonetheless, fine-tuning models pre-trained on natural images can increase the Baseline's accuracy by up to 12 percentage points, achieving an overall accuracy of 83%.

Development and validation of a CT-based radiomic nomogram for predicting surgical resection risk in patients with adhesive small bowel obstruction

BACKGROUND

Adhesive small bowel obstruction (ASBO) is a common emergency that requires prompt medical attention, and the timing of surgical intervention poses a considerable challenge. Although computed tomography (CT) is widely used, its effectiveness in accurately identifying bowel strangulation is limited. The potential of radiomics models to predict the necessity for surgical resection in ASBO cases is not yet fully explored.

OBJECTIVES

The aim of this study is to identify risk factors for surgical resection in patients with ASBO and to develop a predictive model that integrates radiomic features with clinical data. This model designed to estimate the likelihood of surgical intervention and aid in clinical decision-making for acute ASBO cases.

METHODS

From January 2019 to February 2022, we enrolled 188 ASBO patients from our hospital, dividing them randomly into a training cohort (n = 131) and a test cohort (n = 57) using a 7:3 ratio. We collected baseline clinical data and extracted radiomic features from CT images to compute a radiomic score (Rad-score). A nomogram was developed that combines clinical characteristics and Rad-score. The performance of clinical, radiomic, and combined nomogram models was evaluated in both cohorts.

RESULTS

Of the 188 patients, 92 underwent surgical resection, while 96 did not. The nomogram integrated factors such as white blood cell count, duration of obstruction, and preoperative infection indicators (fever, tachycardia, peritonitis), along with CT findings (elevated wall density, thickened wall, mesenteric fluid, ascites, bowel wall gas, small bowel feces, and hyperdensity of mesenteric fat) (p < 0.1). This combined model accurately predicted the need for surgical resection, with area under the curve (AUC) values of 0.761 (95% CI, 0.628-0.893) for the test cohort. Calibration curves showed strong agreement between predicted and observed outcomes, and decision curve analysis validated the model's utility for acute ASBO cases.

CONCLUSION

We developed and validated a CT-based nomogram that combines radiomic features with clinical data to predict the risk of surgical resection in ASBO patients. This tool offers valuable support for treatment planning and decision-making in emergent situations.

Myocardial perfusion SPECT radiomic features reproducibility assessment: Impact of image reconstruction and harmonization

BACKGROUND

Coronary artery disease (CAD) has one of the highest mortality rates in humans worldwide. Single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) provides clinicians with myocardial metabolic information non-invasively. However, there are some limitations to interpreting SPECT images performed by physicians or automatic quantitative approaches. Radiomics analyzes images objectively by extracting quantitative features and can potentially reveal biological characteristics that the human eye cannot detect. However, the reproducibility and repeatability of some radiomic features can be highly susceptible to segmentation and imaging conditions.

PURPOSE

We aimed to assess the reproducibility of radiomic features extracted from uncorrected MPI-SPECT images reconstructed with 15 different settings before and after ComBat harmonization, along with evaluating the effectiveness of ComBat in realigning feature distributions.

MATERIALS AND METHODS

A total of 200 patients (50% normal and 50% abnormal) including rest and stress (without attenuation and scatter corrections) MPI-SPECT images were included. Images were reconstructed using 15 combinations of filter cut-off frequencies, filter orders, filter types, reconstruction algorithms, number of iterations and subsets resulting in 6000 images. Image segmentation was performed on the left ventricle in the first reconstruction for each patient and applied to 14 others. A total of 93 radiomic features were extracted from the segmented area, and ComBat was used to harmonize them. The intraclass correlation coefficient (ICC) and overall concordance correlation coefficient (OCCC) tests were performed before and after ComBat to examine the impact of each parameter on feature robustness and to assess harmonization efficiency. The ANOVA and the Kruskal-Wallis tests were performed to evaluate the effectiveness of ComBat in correcting feature distributions. In addition, the Student's t-test, Wilcoxon rank-sum, and signed-rank tests were implemented to assess the significance level of the impacts made by each parameter of different batches and patient groups (normal vs. abnormal) on radiomic features.

RESULTS

Before applying ComBat, the majority of features (ICC: 82, OCCC: 61) achieved high reproducibility (ICC/OCCC ≥ 0.900) under every batch except Reconstruction. The largest and smallest number of poor features (ICC/OCCC < 0.500) were obtained by IterationSubset and Order batches, respectively. The most reliable features were from the first-order (FO) and gray-level co-occurrence matrix (GLCM) families. Following harmonization, the minimum number of robust features increased (ICC: 84, OCCC: 78). Applying ComBat showed that Order and Reconstruction were the least and the most responsive batches, respectively. The most robust families, in a descending order, were found to be FO, neighborhood gray-tone difference matrix (NGTDM), GLCM, gray-level run length matrix (GLRLM), gray-level size zone matrix (GLSZM), and gray-level dependence matrix (GLDM) under Cut-off, Filter, and Order batches. The Wilcoxon rank-sum test showed that the number of robust features significantly differed under most batches in the Normal and Abnormal groups.

CONCLUSION

The majority of radiomic features show high levels of robustness across different OSEM reconstruction parameters in uncorrected MPI-SPECT. ComBat is effective in realigning feature distributions and enhancing radiomic features reproducibility.

Robustness of radiomics among photon-counting detector CT and dual-energy CT systems: a texture phantom study

OBJECTIVES

To evaluate the robustness of radiomics features among photon-counting detector CT (PCD-CT) and dual-energy CT (DECT) systems.

METHODS

A texture phantom consisting of twenty-eight materials was scanned with one PCD-CT and four DECT systems (dual-source, rapid kV-switching, dual-layer, and sequential scanning) at three dose levels twice. Thirty sets of virtual monochromatic images at 70 keV were reconstructed. Regions of interest were delineated for each material with a rigid registration. Ninety-three radiomics were extracted per PyRadiomics. The test-retest repeatability between repeated scans was assessed by Bland-Altman analysis. The intra-system reproducibility between dose levels, and inter-system reproducibility within the same dose level, were evaluated by intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC). Inter-system variability among five scanners was assessed by coefficient of variation (CV) and quartile coefficient of dispersion (QCD).

RESULTS

The test-retest repeatability analysis presented that 97.1% of features were repeatable between scan-rescans. The mean ± standard deviation ICC and CCC were 0.945 ± 0.079 and 0.945 ± 0.079 for intra-system reproducibility, respectively, and 86.0% and 85.7% of features were with ICC > 0.90 and CCC > 0.90, respectively, between different dose levels. The mean ± standard deviation ICC and CCC were 0.157 ± 0.174 and 0.157 ± 0.174 for inter-system reproducibility, respectively, and none of the features were with ICC > 0.90 or CCC > 0.90 within the same dose level. The inter-system variability suggested that 6.5% and 12.8% of features were with CV < 10% and QCD < 10%, respectively, among five CT systems.

CONCLUSION

The radiomics features were non-reproducible with significant variability in values among different CT techniques.

CLINICAL RELEVANCE STATEMENT

Radiomics features are non-reproducible with significant variability in values among photon-counting detector CT and dual-energy CT systems, necessitating careful attention to improve the cross-system generalizability of radiomic features before implementation of radiomics analysis in clinical routine.

KEY POINTS

CT radiomics stability should be guaranteed before the implementation in the clinical routine. Radiomics robustness was on a low level among photon-counting detectors and dual-energy CT techniques. Limited inter-system robustness of radiomic features may impact the generalizability of models.

Robustness of radiomics within photon-counting detector CT: impact of acquisition and reconstruction factors

OBJECTIVES

To assess the impact of acquisition and reconstruction factors on the robustness of radiomics within photon-counting detector CT (PCD-CT).

METHODS

A phantom with twenty-eight texture materials was scanned with different acquisition and reconstruction factors including reposition, scan mode (standard vs high-pitch), tube voltage (120 kVp vs 140 kVp), slice thickness (1.0 mm vs 0.4 mm), radiation dose level (0.5 mGy, 1.0 mGy, 3.0 mGy, 5.0 mGy, vs 10.0 mGy), quantum iterative reconstruction level (0/4, 2/4, vs 4/4), and reconstruction kernel (Qr40, Qr44, vs Qr48). Thirteen sets of virtual monochromatic images at 70-keV were reconstructed. The regions of interest were drawn with rigid registrations. Ninety-three radiomics features were extracted from each material. The reproducibility of radiomics features was evaluated using the intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC). The variability of radiomics features was assessed by coefficient of variation (CV) and quartile coefficient of dispersion (QCD).

RESULTS

The percentage of features with ICC > 0.90 and CCC > 0.90 were high when repositioned (88.2% and 88.2%) and tube voltage was changed (87.1% and 87.1%), but none of the features with ICC > 0.90 and CCC > 0.90 when high-pitch scan and different slice thickness were used. The percentage of features with CV < 10% and QCD < 10% were high when repositioned (47.3% and 68.8%) and tube voltage was changed (64.2% and 71.0%), but that with CV < 10% and QCD < 10% were low between standard and high-pitch scans (16.1% and 26.9%) and slice thickness (19.4% and 29.0%).

CONCLUSIONS

The PCD-CT radiomics was robust to tube voltage, radiation dose, reconstruction strength level, and kernel, but brittle to high-pitch scan and slice thickness.

KEY POINTS

Question The stability of radiomics features against acquisition and reconstruction factors within PCD-CT should be fully determined before academic research and clinical application. Findings The radiomics features are robust against tube voltage, radiation dose, reconstruction strength level, and kernel within PCD-CT but brittle to high-pitch scan and slice thickness. Clinical relevance The high-pitch scan and slice thickness that influence voxel size should be set with careful attention within PCD-CT, to allow a higher robustness of radiomics features before the implementation of radiomics analysis in clinical routine.

Can CT Image Reconstruction Parameters Impact the Predictive Value of Radiomics Features in Grading Pancreatic Neuroendocrine Neoplasms?

The WHO grading of pancreatic neuroendocrine neoplasms (PanNENs) is essential in patient management and an independent prognostic factor for patient survival. Radiomics features from CE-CT images hold promise for the outcome and tumor grade prediction. However, variations in reconstruction parameters can impact the predictive value of radiomics. 127 patients with histopathologically confirmed PanNENs underwent CT scans with filtered back projection (B20f) and iterative (I26f) reconstruction kernels. 3190 radiomic features were extracted from tumors and pancreatic volumes. Wilcoxon paired tests assessed the impact of reconstruction kernels and ComBat harmonization efficiency. SVM models were employed to predict tumor grade using the entire set of radiomics features or only those identified as harmonizable. The models' performance was assessed on an independent dataset of 36 patients. Significant differences, after correction for multiple testing, were observed in 69% of features in the pancreatic volume and 51% in the tumor volume with B20f and I26f kernels. SVM models demonstrated accuracy ranging from 0.67 (95%CI: 0.50-0.81) to 0.83 (95%CI: 0.69-0.94) in distinguishing grade 1 cases from higher grades. Reconstruction kernels alter radiomics features and iterative kernel models trended towards higher performance. ComBat harmonization mitigates kernel impacts but addressing this effect is crucial in studies involving data from different kernels.

Assessing Glioblastoma Treatment Response Using Machine Learning Approach Based on Magnetic Resonance Images Radiomics: An Exploratory Study

Background and Objectives

Assessing treatment response in glioblastoma multiforme (GBM) tumors necessitates developing more objective and quantitative approaches. A machine learning-based approach is presented in this exploratory study for GBM patients' treatment response assessment based on radiomics extracted from magnetic resonance (MR) images.

Methods

MR images from 77 GBM patients were acquired at two post-surgery stages and preprocessed. From these images, 107 radiomics were extracted from the segmented tumoral cavities. The most informative features for training machine learning (ML) classifiers were identified using the Spearman correlation analysis of features retained by the forward sequential and LASSO algorithms. Applied machine learning models included support vector machine (SVM), random forest (RF), K-nearest neighbors (KNN), AdaBoost, categorical boosting (CatBoost), light gradient boosting machine (LightGBM), extreme gradient boosting (XGBoost), Naïve Bayes (NB) and logistic regression (LR). Ten-fold cross-validation was used to validate the models. Statistical analysis was conducted using SPSS version 27; p-value < 0.05 was considered significant.

Results

The Naïve Bayes classifier demonstrated the highest performance among the trained models, achieving an AUC (area under the receiver operating characteristic curve) of 0.86 ± 0.13 when trained on the seven features selected by the forward sequential algorithm and an AUC of 0.84 ± 0.14 when trained using the five features chosen by the LASSO algorithm. The second-best performance was observed with the KNN classifier, which achieved an AUC of 0.80 ± 0.17 when trained on the features selected by the forward sequential algorithm.

Conclusion

Findings demonstrated that MRI-based radiomics could be used as distinctive features to train ML models for GBM patients' treatment response assessment. Trained ML classifiers based on these features serve as aiding tools to expedite the quantitative assessment of GBM patients' treatment response besides qualitative evaluations.

From images to clinical insights: an educational review on radiomics in lung diseases

Radiological imaging is a cornerstone in the clinical workup of lung diseases. Radiomics represents a significant advancement in clinical lung imaging, offering a powerful tool to complement traditional qualitative image analysis. Radiomic features are quantitative and computationally describe shape, intensity, texture and wavelet characteristics from medical images that can uncover detailed and often subtle information that goes beyond the visual capabilities of radiological examiners. By extracting this quantitative information, radiomics can provide deep insights into the pathophysiology of lung diseases and support clinical decision-making as well as personalised medicine approaches. In this educational review, we provide a step-by-step guide to radiomics-based medical image analysis, discussing the technical challenges and pitfalls, and outline the potential clinical applications of radiomics in diagnosing, prognosticating and evaluating treatment responses in respiratory medicine.

Artificial intelligence-enhanced magnetic resonance imaging-based pre-operative staging in patients with endometrial cancer

OBJECTIVE

Evaluation of prognostic factors is crucial in patients with endometrial cancer for optimal treatment planning and prognosis assessment. This study proposes a deep learning pipeline for tumor and uterus segmentation from magnetic resonance imaging (MRI) images to predict deep myometrial invasion and cervical stroma invasion and thus assist clinicians in pre-operative workups.

METHODS

Two experts consensually reviewed the MRIs and assessed myometrial invasion and cervical stromal invasion as per the International Federation of Gynecology and Obstetrics staging classification, to compare the diagnostic performance of the model with the radiologic consensus. The deep learning method was trained using sagittal T2-weighted images from 142 patients and tested with a 3-fold stratified test with 36 patients in each fold. Our solution is based on a segmentation module, which employed a 2-stage pipeline for efficient uterus in the whole MRI volume and then tumor segmentation in the uterus predicted region of interest.

RESULTS

A total of 178 patients were included. For deep myometrial invasion prediction, the model achieved an average balanced test accuracy over 3-folds of 0.702, while experts reached an average accuracy of 0.769. For cervical stroma invasion prediction, our model demonstrated an average balanced accuracy of 0.721 on the 3-fold test set, while experts achieved an average balanced accuracy of 0.859. Additionally, the accuracy rates for uterus and tumor segmentation, measured by the Dice score, were 0.847 and 0.579 respectively.

CONCLUSION

Despite the current challenges posed by variations in data, class imbalance, and the presence of artifacts, our fully automatic approach holds great promise in supporting in pre-operative staging. Moreover, it demonstrated a robust capability to segment key regions of interest, specifically the uterus and tumors, highlighting the positive impact our solution can bring to health care imaging.

Multi-modality artificial intelligence-based transthyretin amyloid cardiomyopathy detection in patients with severe aortic stenosis

PURPOSE

Transthyretin amyloid cardiomyopathy (ATTR-CM) is a frequent concomitant condition in patients with severe aortic stenosis (AS), yet it often remains undetected. This study aims to comprehensively evaluate artificial intelligence-based models developed based on preprocedural and routinely collected data to detect ATTR-CM in patients with severe AS planned for transcatheter aortic valve implantation (TAVI).

METHODS

In this prospective, single-center study, consecutive patients with AS were screened with [99mTc]-3,3-diphosphono-1,2-propanodicarboxylic acid ([99mTc]-DPD) for the presence of ATTR-CM. Clinical, laboratory, electrocardiogram, echocardiography, invasive measurements, 4-dimensional cardiac CT (4D-CCT) strain data, and CT-radiomic features were used for machine learning modeling of ATTR-CM detection and for outcome prediction. Feature selection and classifier algorithms were applied in single- and multi-modality classification scenarios. We split the dataset into training (70%) and testing (30%) samples. Performance was assessed using various metrics across 100 random seeds.

RESULTS

Out of 263 patients with severe AS (57% males, age 83 ± 4.6years) enrolled, ATTR-CM was confirmed in 27 (10.3%). The lowest performances for detection of concomitant ATTR-CM were observed in invasive measurements and ECG data with area under the curve (AUC) < 0.68. Individual clinical, laboratory, interventional imaging, and CT-radiomics-based features showed moderate performances (AUC 0.70-0.76, sensitivity 0.79-0.82, specificity 0.63-0.72), echocardiography demonstrated good performance (AUC 0.79, sensitivity 0.80, specificity 0.78), and 4D-CT-strain showed the highest performance (AUC 0.85, sensitivity 0.90, specificity 0.74). The multi-modality model (AUC 0.84, sensitivity 0.87, specificity 0.76) did not outperform the model performance based on 4D-CT-strain only data (p-value > 0.05). The multi-modality model adequately discriminated low and high-risk individuals for all-cause mortality at a mean follow-up of 13 months.

CONCLUSION

Artificial intelligence-based models using collected pre-TAVI evaluation data can effectively detect ATTR-CM in patients with severe AS, offering an alternative diagnostic strategy to scintigraphy and myocardial biopsy.

Advances and challenges in precision imaging

Technological innovations in genomics and related fields have facilitated large sequencing efforts, supported new biological discoveries in cancer, and spawned an era of liquid biopsy biomarkers. Despite these advances, precision oncology has practical constraints, partly related to cancer's biological diversity and spatial and temporal complexity. Advanced imaging technologies are being developed to address some of the current limitations in early detection, treatment selection and planning, drug delivery, and therapeutic response, as well as difficulties posed by drug resistance, drug toxicity, disease monitoring, and metastatic evolution. We discuss key areas of advanced imaging for improving cancer outcomes and survival. Finally, we discuss practical challenges to the broader adoption of precision imaging in the clinic and the need for a robust translational infrastructure.

Challenges in clinical translation of cardiac magnetic resonance imaging radiomics in non-ischemic cardiomyopathy: a narrative review

Background and Objective

Radiomics is an emerging technology that facilitates the quantitative analysis of multi-modal cardiac magnetic resonance imaging (MRI). This study aims to introduce a standardized workflow for applying radiomics to non-ischemic cardiomyopathies, enabling clinicians to comprehensively understand and implement this technology in clinical practice.

Methods

A computerized literature search (up to August 1, 2024) was conducted using PubMed to identify relevant studies on the roles and workflows of radiomics in non-ischemic cardiomyopathy. Expert discussions were also held to ensure the accuracy and relevance of the findings. Only English-language publications were reviewed.

Key Content and Findings

The paper details the essential processes of radiomics, including feature extraction, feature engineering, model construction, and data analysis. It emphasizes the role of MRI in assessing cardiac structure and function and demonstrates how MRI-based radiomics can aid in diagnosing and differentiating non-ischemic cardiomyopathies such as hypertrophic cardiomyopathy, dilated cardiomyopathy, and myocarditis. The study also investigates various cardiac MRI sequences to enhance the clinical application of radiomics.

Conclusions

The standardized radiomics workflow presented in this study aims to assist clinicians in effectively utilizing MRI-based radiomics for the diagnosis and management of non-ischemic cardiomyopathies, thereby improving clinical decision-making.

Enhancing the Opportunistic Bone Status Assessment Using Radiomics Based on Dual-Energy Spectral CT Material Decomposition Images

The dual-energy spectral CT (DEsCT) employs material decomposition (MD) technology, opening up novel avenues for the opportunistic assessment of bone status. Radiomics, a powerful tool for elucidating the structural and textural characteristics of bone, aids in the detection of mineral loss. Therefore, this study aims to compare the efficacy of bone status assessment using both bone density measurements and radiomics models derived from MD images and to further explore the clinical value of radiomics models.

METHODS

Retrospective data were collected from 307 patients who underwent both quantitative computed tomography (QCT) and full-abdomen DEsCT scans at our institution. Based on QCT measurements, patients were divided into three categories: normal bone mineral density (BMD), osteopenia, and osteoporosis. Using the abdominal DEsCT data, six types of MD images were reconstructed, including HAP (Water), HAP (Fat), Ca (Water), Ca (Fat), Fat (Ca), and Fat (HAP). Patients were randomly divided into a training cohort (n = 214) and a validation cohort (n = 93) at a ratio of 7:3. Focusing on the L1 to L3 vertebrae, density values from the six MD images were measured. Six density value models and six radiomics models were constructed using a random forest (RF) classifier. The performance of these models in assessing bone status was evaluated using the receiver operating characteristic (ROC) curves, and the DeLong test was employed to compare performance differences between the models.

RESULTS

The macro-area under the curve (AUC) values for the density value models based on HAP (Water), HAP (Fat), Ca (Water), and Ca (Fat) MD images were 0.870, 0.870, 0.847, and 0.765, respectively, which outperformed those of Fat (Ca) (AUC = 0.623) and Fat (HAP) (AUC = 0.618) density value models. In the comparison of radiomics models, the trends of model performance were consistent with the density value models across the six MD images. However, the models based on HAP (Water), Ca (Water), HAP (Fat), Ca (Fat), Fat (Ca), and Fat (HAP) images exhibited superior performance than those of the density value models with the corresponding MD images, with values of 0.946, 0.941, 0.934, 0.926, 0.831, and 0.824, respectively.

CONCLUSIONS

Bone status assessment can be accurately conducted using density values from HAP (Water), HAP (Fat), Ca (Water), and Ca (Fat) MD images. However, radiomics models derived from MD images surpass traditional density measurement methods in evaluating bone status, highlighting their superior diagnostic potential.

CT-based radiomics features for the differential diagnosis of nodular goiter and papillary thyroid carcinoma: an analysis employing propensity score matching

Purpose

This study aims to evaluate the effectiveness of CT-based radiomics features in discriminating between nodular goiter (NG) and papillary thyroid carcinoma (PTC).

Methods

A retrospective cohort comprising 228 patients with nodular goiter (NG) and 227 patients with papillary thyroid carcinoma (PTC) diagnosed between January 2018 and December 2022 was consecutively enrolled. Propensity score matching (PSM) was applied to align patients with NG and PTC. A total of 851 radiomics features were extracted from CT images acquired during the arterial phase for each individual. Feature selection was carried out utilizing the least absolute shrinkage and selection operator (LASSO) logistic regression algorithm to generate the radiomics score (Rad-score). Subsequently, the Rad-score was incorporated into a multivariate logistic regression analysis to construct a radiomics nomogram for visual representation.

Results

Following PSM implementation, 101 patients diagnosed with NG were matched with an equivalent number of patients diagnosed with PTC. The developed radiomics score exhibited excellent predictive performance in distinguishing between NG and PTC, with high values of AUC, sensitivity, and specificity in both the training cohort (AUC = 0.823, accuracy = 0.759, sensitivity = 0.794, specificity = 0.740) and validation cohort (AUC = 0.904, accuracy = 0.820, sensitivity = 0.758, specificity = 0.964).

Conclusion

The utilization of CT-based radiomics analysis following PMS offers a quantitative and data-driven approach to enhance the accuracy of distinguishing between nodular goiter (NG) and papillary thyroid carcinoma (PTC).

The Challenge of External Generalisability: Insights from the Bicentric Validation of a [68Ga]Ga-PSMA-11 PET Based Radiomics Signature for Primary Prostate Cancer Characterisation Using Histopathology as Reference

Background: PSMA PET radiomics is a promising tool for primary prostate cancer (PCa) characterisation. However, small single-centre studies and lack of external validation hinder definitive conclusions on the potential of PSMA PET radiomics in the initial workup of PCa. We aimed to validate a radiomics signature in a larger internal cohort and in an external cohort from a separate centre. Methods: One hundred and twenty-seven PCa patients were retrospectively enrolled across two independent hospitals. The first centre (IRCCS San Raffaele Scientific Institute, Centre 1) contributed 62 [68Ga]Ga-PSMA-11 PET scans, 20 patients classified as low-grade (ISUP grade < 4), and 42 as high-grade (ISUP grade ≥ 4). The second centre (Stanford University Hospital, Centre 2) provided 65 [68Ga]Ga-PSMA-11 PET scans, and 49 low-grade and 16 high-grade patients. A radiomics model previously generated in Centre 1 was tested on the two cohorts separately and afterward on the entire dataset. Then, we evaluated whether the radiomics features selected in the previous investigation could generalise to new data. Several machine learning (ML) models underwent training and testing using 100-fold Monte Carlo cross-validation, independently at both Centre 1 and Centre 2, with a 70-30% train-test split. Additionally, models were trained in one centre and tested in the other, and vice versa. Furthermore, data from both centres were combined for training and testing using Monte Carlo cross-validation. Finally, a new radiomics signature built on this bicentric dataset was proposed. Several performance metrics were computed. Results: The previously generated radiomics signature resulted in an area under the receiver operating characteristic curve (AUC) of 80.4% when tested on Centre 1, while it generalised poorly to Centre 2, where it reached an AUC of 62.7%. When the whole cohort was considered, AUC was 72.5%. Similarly, new ML models trained on the previously selected features yielded, at best, an AUC of 80.9% for Centre 1 and performed at chance for Centre 2 (AUC of 49.3%). A new signature built on this bicentric dataset reached, at best, an average AUC of 91.4% in the test set. Conclusions: The satisfying performance of radiomics models when used in the original development settings, paired with the poor performance otherwise observed, emphasises the need to consider centre-specific factors and dataset characteristics when developing radiomics models. Combining radiomics datasets is a viable strategy to reduce such centre-specific biases, but external validation is still needed.

Fully Automated Region-Specific Human-Perceptive-Equivalent Image Quality Assessment: Application to 18 F-FDG PET Scans

INTRODUCTION

We propose a fully automated framework to conduct a region-wise image quality assessment (IQA) on whole-body 18 F-FDG PET scans. This framework (1) can be valuable in daily clinical image acquisition procedures to instantly recognize low-quality scans for potential rescanning and/or image reconstruction, and (2) can make a significant impact in dataset collection for the development of artificial intelligence-driven 18 F-FDG PET analysis models by rejecting low-quality images and those presenting with artifacts, toward building clean datasets.

PATIENTS AND METHODS

Two experienced nuclear medicine physicians separately evaluated the quality of 174 18 F-FDG PET images from 87 patients, for each body region, based on a 5-point Likert scale. The body regisons included the following: (1) the head and neck, including the brain, (2) the chest, (3) the chest-abdomen interval (diaphragmatic region), (4) the abdomen, and (5) the pelvis. Intrareader and interreader reproducibility of the quality scores were calculated using 39 randomly selected scans from the dataset. Utilizing a binarized classification, images were dichotomized into low-quality versus high-quality for physician quality scores ≤3 versus >3, respectively. Inputting the 18 F-FDG PET/CT scans, our proposed fully automated framework applies 2 deep learning (DL) models on CT images to perform region identification and whole-body contour extraction (excluding extremities), then classifies PET regions as low and high quality. For classification, 2 mainstream artificial intelligence-driven approaches, including machine learning (ML) from radiomic features and DL, were investigated. All models were trained and evaluated on scores attributed by each physician, and the average of the scores reported. DL and radiomics-ML models were evaluated on the same test dataset. The performance evaluation was carried out on the same test dataset for radiomics-ML and DL models using the area under the curve, accuracy, sensitivity, and specificity and compared using the Delong test with P values <0.05 regarded as statistically significant.

RESULTS

In the head and neck, chest, chest-abdomen interval, abdomen, and pelvis regions, the best models achieved area under the curve, accuracy, sensitivity, and specificity of [0.97, 0.95, 0.96, and 0.95], [0.85, 0.82, 0.87, and 0.76], [0.83, 0.76, 0.68, and 0.80], [0.73, 0.72, 0.64, and 0.77], and [0.72, 0.68, 0.70, and 0.67], respectively. In all regions, models revealed highest performance, when developed on the quality scores with higher intrareader reproducibility. Comparison of DL and radiomics-ML models did not show any statistically significant differences, though DL models showed overall improved trends.

CONCLUSIONS

We developed a fully automated and human-perceptive equivalent model to conduct region-wise IQA over 18 F-FDG PET images. Our analysis emphasizes the necessity of developing separate models for body regions and performing data annotation based on multiple experts' consensus in IQA studies.

Radiomics in radiology: What the radiologist needs to know about technical aspects and clinical impact

Radiomics represents the science of extracting and analyzing a multitude of quantitative features from medical imaging, revealing the quantitative potential of radiologic images. This scientific review aims to provide radiologists with a comprehensive understanding of radiomics, emphasizing its principles, applications, challenges, limits, and prospects. The limitations of standardization in current scientific production are analyzed, along with possible solutions proposed by some of the referenced papers. As the continuous evolution of medical imaging is ongoing, radiologists must be aware of new perspectives to play a central role in patient management.

A Tumor Volume Segmentation Algorithm Based on Radiomics Features in FDG-PET in Lung Cancer Patients, Validated Using Surgical Specimens

BACKGROUND

Although the integration of positron emission tomography into radiation therapy treatment planning has become part of clinical routine, the best method for tumor delineation is still a matter of debate. In this study, therefore, we analyzed a novel, radiomics-feature-based algorithm in combination with histopathological workup for patients with non-small-cell lung cancer.

METHODS

A total of 20 patients with biopsy-proven lung cancer who underwent [18F]fluorodeoxyglucose positron emission/computed tomography (FDG-PET/CT) examination before tumor resection were included. Tumors were segmented in positron emission tomography (PET) data using previously reported algorithms based on three different radiomics features, as well as a threshold-based algorithm. To obtain gold-standard results, lesions were measured after resection. Pathological volumes and maximal diameters were then compared with the results of the segmentation algorithms.

RESULTS

A total of 20 lesions were analyzed. For all algorithms, segmented volumes correlated well with pathological volumes. In general, the threshold-based volumes exhibited a tendency to be smaller than the radiomics-based volumes. For all lesions, conventional threshold-based segmentation produced coefficients of variation which corresponded best with pathologically based volumes; however, for lesions larger than 3 ccm, the algorithm based on Local Entropy performed best, with a significantly better coefficient of variation (p = 0.0002) than the threshold-based algorithm.

CONCLUSIONS

We found that, for small lesions, results obtained using conventional threshold-based segmentation compared well with pathological volumes. For lesions larger than 3 ccm, the novel algorithm based on Local Entropy performed best. These findings confirm the results of our previous phantom studies. This algorithm is therefore worthy of inclusion in future studies for further confirmation and application.

Image biomarkers and explainable AI: handcrafted features versus deep learned features

Feature extraction and selection from medical data are the basis of radiomics and image biomarker discovery for various architectures, including convolutional neural networks (CNNs). We herein describe the typical radiomics steps and the components of a CNN for both deep feature extraction and end-to-end approaches. We discuss the curse of dimensionality, along with dimensionality reduction techniques. Despite the outstanding performance of deep learning (DL) approaches, the use of handcrafted features instead of deep learned features needs to be considered for each specific study. Dataset size is a key factor: large-scale datasets with low sample diversity could lead to overfitting; limited sample sizes can provide unstable models. The dataset must be representative of all the "facets" of the clinical phenomenon/disease investigated. The access to high-performance computational resources from graphics processing units is another key factor, especially for the training phase of deep architectures. The advantages of multi-institutional federated/collaborative learning are described. When large language models are used, high stability is needed to avoid catastrophic forgetting in complex domain-specific tasks. We highlight that non-DL approaches provide model explainability superior to that provided by DL approaches. To implement explainability, the need for explainable AI arises, also through post hoc mechanisms. RELEVANCE STATEMENT: This work aims to provide the key concepts for processing the imaging features to extract reliable and robust image biomarkers. KEY POINTS: The key concepts for processing the imaging features to extract reliable and robust image biomarkers are provided. The main differences between radiomics and representation learning approaches are highlighted. The advantages and disadvantages of handcrafted versus learned features are given without losing sight of the clinical purpose of artificial intelligence models.

Automatic segmentation-based multi-modal radiomics analysis of US and MRI for predicting disease-free survival of breast cancer: a multicenter study

BACKGROUND

Several studies have confirmed the potential value of applying radiomics to predict prognosis of breast cancer. However, the tumor segmentation in these studies depended on delineation or annotation of breast cancer by radiologist, which is often laborious, tedious, and vulnerable to inter- and intra-observer variability. Automatic segmentation is expected to overcome this difficulty. Herein, we aim to investigate the value of automatic segmentation-based multi-modal radiomics signature and magnetic resonance imaging (MRI) features in predicting disease-free survival (DFS) of patients diagnosed with invasive breast cancer.

METHODS

This retrospective multicenter study included a total of 643 female patients with invasive breast cancer who underwent preoperative ultrasound (US) and MRI for prognostic analysis. Data (n = 480) from center 1 were divided into training and internal testing sets, while data (n = 163) from centers 2 and 3 were analyzed as the external testing set. We developed automatic segmentation frameworks for tumor segmentation by deep learning. Then, Least absolute shrinkage and selection operator Cox regression was used to select features to construct radiomics signature, and corresponding radiomics score (Rad-score) was calculated. Finally, six models for predicting DFS were constructed by using Cox regression and assessed in terms of discrimination, calibration, and clinical usefulness.

RESULTS

The multi-modal radiomics signature combining intra- and peri-tumoral radiomics signatures of US and MRI achieved a higher C-index in the internal (0.734) and external (0.708) testing sets than most other radiomics signatures in predicting DFS, and successfully stratified patients into low- and high-risk groups. The multi-modal clinical imaging model combining the multi-modal Rad-score and clinical traditional MRI model-score resulted in a higher C-index (0.795) than other models in the external testing set, and it had a better calibration and higher clinical benefit.

CONCLUSIONS

This study demonstrates that the multi-modal radiomics signature derived from automatic segmentations of US and MRI is a promising risk stratification biomarker for breast cancer, and highlights that the appropriate combination of multi-modal radiomics signature, clinical characteristics, and MRI feature can improve performance of individualized DFS prediction, which might assist in guiding decision-making related to breast cancer.

Artificial Intelligence for Response Assessment in Neuro Oncology (AI-RANO), part 2: recommendations for standardisation, validation, and good clinical practice

Technological advancements have enabled the extended investigation, development, and application of computational approaches in various domains, including health care. A burgeoning number of diagnostic, predictive, prognostic, and monitoring biomarkers are continuously being explored to improve clinical decision making in neuro-oncology. These advancements describe the increasing incorporation of artificial intelligence (AI) algorithms, including the use of radiomics. However, the broad applicability and clinical translation of AI are restricted by concerns about generalisability, reproducibility, scalability, and validation. This Policy Review intends to serve as the leading resource of recommendations for the standardisation and good clinical practice of AI approaches in health care, particularly in neuro-oncology. To this end, we investigate the repeatability, reproducibility, and stability of AI in response assessment in neuro-oncology in studies on factors affecting such computational approaches, and in publicly available open-source data and computational software tools facilitating these goals. The pathway for standardisation and validation of these approaches is discussed with the view of trustworthy AI enabling the next generation of clinical trials. We conclude with an outlook on the future of AI-enabled neuro-oncology.

Reproducible and Interpretable Machine Learning-Based Radiomic Analysis for Overall Survival Prediction in Glioblastoma Multiforme

PURPOSE

To develop and validate an MRI-based radiomic model for predicting overall survival (OS) in patients diagnosed with glioblastoma multiforme (GBM), utilizing a retrospective dataset from multiple institutions.

MATERIALS AND METHODS

Pre-treatment MRI images of 289 GBM patients were collected. From each patient's tumor volume, 660 radiomic features (RFs) were extracted and subjected to robustness analysis. The initial prognostic model with minimum RFs was subsequently enhanced by including clinical variables. The final clinical-radiomic model was derived through repeated three-fold cross-validation on the training dataset. Performance evaluation included assessment of concordance index (C-Index), integrated area under curve (iAUC) alongside patient stratification into low and high-risk groups for overall survival (OS).

RESULTS

The final prognostic model, which has the highest level of interpretability, utilized primary gross tumor volume (GTV) and one MRI modality (T2-FLAIR) as a predictor and integrated the age variable with two independent, robust RFs, achieving moderately good discriminatory performance (C-Index [95% confidence interval]: 0.69 [0.62-0.75]) with significant patient stratification (p = 7 × 10-5) on the validation cohort. Furthermore, the trained model exhibited the highest iAUC at 11 months (0.81) in the literature.

CONCLUSION

We identified and validated a clinical-radiomic model for stratification of patients into low and high-risk groups based on OS in patients with GBM using a multicenter retrospective dataset. Future work will focus on the use of deep learning-based features, with recently standardized convolutional filters on OS tasks.

Quantifying the reproducibility and longitudinal repeatability of radiomics features in magnetic resonance Image-Guide accelerator Imaging: A phantom study

OBJECTIVE

This study aimed to quantitatively evaluate the inter-platform reproducibility and longitudinal acquisition repeatability of MRI radiomics features in Fluid-Attenuated Inversion Recovery (FLAIR), T2-weighted (T2W), and T1-weighted (T1W) sequences on MR-Linac systems using an American College of Radiology (ACR) phantom.

MATERIALS AND METHODS

This study used two MR-Linac systems (A and B) in different cancer centers. The ACR phantom was scanned on system A daily for 30 consecutive days to evaluate longitudinal repeatability. Additionally, retest data were collected after repositioning the phantom. Inter-platform reproducibility was assessed by conducting scans under identical conditions using system B. Regions of interest were delineated on the T1W sequence from system A and mapped to other sequences via rigid registration. Intra-observer and inter-observer comparisons were conducted. Repeatability and reproducibility were assessed using the intraclass correlation coefficient (ICC) and coefficient of variation (CV). Robust radiomics features were identified based on ICC>0.9 and CV<10 %.

RESULTS

Analysis showed that a higher proportion of radiomics features derived from longitudinal FLAIR sequence (51.65 %) met robustness criteria compared to T2W (48.35 %) and T1W (43.96 %). Additionally, more inter-platform features from the FLAIR sequence (62.64 %) were robust compared to T2W (42.86 %) and T1W (39.56 %). Test-retest and intra-observer repeatability were excellent across all sequences, with a median ICC of 0.99 and CV<5%. However, inter-observer reproducibility was inferior, especially for the T1W sequence.

CONCLUSIONS

Different sequences show variations in repeatability and reproducibility. The FLAIR sequence demonstrated advantages in both longitudinal repeatability and inter-platform reproducibility. Caution is warranted when interpreting data, particularly in longitudinal or multiplatform radiomics studies.

Radiogenomics: bridging the gap between imaging and genomics for precision oncology

Genomics allows the tracing of origin and evolution of cancer at molecular scale and underpin modern cancer diagnosis and treatment systems. Yet, molecular biomarker-guided clinical decision-making encounters major challenges in the realm of individualized medicine, consisting of the invasiveness of procedures and the sampling errors due to high tumor heterogeneity. By contrast, medical imaging enables noninvasive and global characterization of tumors at a low cost. In recent years, radiomics has overcomes the limitations of human visual evaluation by high-throughput quantitative analysis, enabling the comprehensive utilization of the vast amount of information underlying radiological images. The cross-scale integration of radiomics and genomics (hereafter radiogenomics) has the enormous potential to enhance cancer decoding and act as a catalyst for digital precision medicine. Herein, we provide a comprehensive overview of the current framework and potential clinical applications of radiogenomics in patient care. We also highlight recent research advances to illustrate how radiogenomics can address common clinical problems in solid tumors such as breast cancer, lung cancer, and glioma. Finally, we analyze existing literature to outline challenges and propose solutions, while also identifying future research pathways. We believe that the perspectives shared in this survey will provide a valuable guide for researchers in the realm of radiogenomics aiming to advance precision oncology.

Peritumoral radiomics increases the efficiency of classification of pure ground-glass lung nodules: a multicenter study

PURPOSE

We aimed to evaluate the efficiency of computed tomography (CT) radiomic features extracted from gross tumor volume (GTV) and peritumoral volumes (PTV) of 5, 10, and 15 mm to identify the tumor grades corresponding to the new histological grading system proposed in 2020 by the Pathology Committee of the International Association for the Study of Lung Cancer (IASLC).

METHODS

A total of 151 lung adenocarcinomas manifesting as pure ground-glass lung nodules (pGGNs) were included in this randomized multicenter retrospective study. Four radiomic models were constructed from GTV and GTV + 5/10/15-mm PTV, respectively, and compared. The diagnostic performance of the different models was evaluated using receiver operating characteristic curve analysis RESULTS: The pGGNs were classified into grade 1 (117), 2 (34), and 3 (0), according to the IASLC grading system. In all four radiomic models, pGGNs of grade 2 had significantly higher radiomic scores than those of grade 1 (P < 0.05). The AUC of the GTV and GTV + 5/10/15-mm PTV were 0.869, 0.910, 0.951, and 0.872 in the training cohort and 0.700, 0.715, 0.745, and 0.724 in the validation cohort, respectively.

CONCLUSIONS

The radiomic features we extracted from the GTV and PTV of pGGNs could effectively be used to differentiate grade-1 and grade-2 tumors. In particular, the radiomic features from the PTV increased the efficiency of the diagnostic model, with GTV + 10 mm PTV exhibiting the highest efficacy.

A mixed Mamba U-net for prostate segmentation in MR images

The diagnosis of early prostate cancer depends on the accurate segmentation of prostate regions in magnetic resonance imaging (MRI). However, this segmentation task is challenging due to the particularities of prostate MR images themselves and the limitations of existing methods. To address these issues, we propose a U-shaped encoder-decoder network MM-UNet based on Mamba and CNN for prostate segmentation in MR images. Specifically, we first proposed an adaptive feature fusion module based on channel attention guidance to achieve effective fusion between adjacent hierarchical features and suppress the interference of background noise. Secondly, we propose a global context-aware module based on Mamba, which has strong long-range modeling capabilities and linear complexity, to capture global context information in images. Finally, we propose a multi-scale anisotropic convolution module based on the principle of parallel multi-scale anisotropic convolution blocks and 3D convolution decomposition. Experimental results on two public prostate MR image segmentation datasets demonstrate that the proposed method outperforms competing models in terms of prostate segmentation performance and achieves state-of-the-art performance. In future work, we intend to enhance the model's robustness and extend its applicability to additional medical image segmentation tasks.

A reference framework for standardization and harmonization of CT radiomics features on cadaveric sample

Radiomics features (RFs) serve as quantitative metrics to characterize shape, density/intensity, and texture patterns in radiological images. Despite their promise, RFs exhibit reproducibility challenges across acquisition settings, thus limiting implementation into clinical practice. In this investigation, we evaluate the effects of different CT scanners and CT acquisition protocols (KV, mA, field-of-view, and reconstruction kernel settings) on RFs extracted from lumbar vertebrae of a cadaveric trunk. Employing univariate and multivariate Generalized Linear Models (GLM), we evaluated the impact of each acquisition parameter on RFs. Our findings indicate that variations in mA had negligible effects on RFs, while alterations in kV resulted in exponential changes in several RFs, notably First Order (94.4%), GLCM (87.5%), and NGTDM (100%). Moreover, we demonstrated that a tailored GLM model was superior to the ComBat algorithm in harmonizing CT images. GLM achieved R2 > 0.90 in 21 RFs (19.6%), contrasting ComBat's mean R2 above 0.90 in only 1 RF (0.9%). This pioneering study unveils the effects of CT acquisition parameters on bone RFs in cadaveric specimens, highlighting significant variations across parameters and scanner datasets. The proposed GLM model presents a robust solution for mitigating these differences, potentially advancing harmonization efforts in Radiomics-based studies across diverse CT protocols and vendors.

Comparative Analysis of Repeatability in CT Radiomics and Dosiomics Features under Image Perturbation: A Study in Cervical Cancer Patients

This study aims to evaluate the repeatability of radiomics and dosiomics features via image perturbation of patients with cervical cancer. A total of 304 cervical cancer patients with planning CT images and dose maps were retrospectively included. Random translation, rotation, and contour randomization were applied to CT images and dose maps before radiomics feature extraction. The repeatability of radiomics and dosiomics features was assessed using intra-class correlation of coefficient (ICC). Pearson correlation coefficient (r) was adopted to quantify the correlation between the image characteristics and feature repeatability. In general, the repeatability of dosiomics features was lower compared with CT radiomics features, especially after small-sigma Laplacian-of-Gaussian (LoG) and wavelet filtering. More repeatable features (ICC > 0.9) were observed when extracted from the original, Large-sigma LoG filtered, and LLL-/LLH-wavelet filtered images. Positive correlations were found between image entropy and high-repeatable feature number in both CT and dose (r = 0.56, 0.68). Radiomics features showed higher repeatability compared to dosiomics features. These findings highlight the potential of radiomics features for robust quantitative imaging analysis in cervical cancer patients, while suggesting the need for further refinement of dosiomics approaches to enhance their repeatability.

Adaptive Mish activation and ranger optimizer-based SEA-ResNet50 model with explainable AI for multiclass classification of COVID-19 chest X-ray images

A recent global health crisis, COVID-19 is a significant global health crisis that has profoundly affected lifestyles. The detection of such diseases from similar thoracic anomalies using medical images is a challenging task. Thus, the requirement of an end-to-end automated system is vastly necessary in clinical treatments. In this way, the work proposes a Squeeze-and-Excitation Attention-based ResNet50 (SEA-ResNet50) model for detecting COVID-19 utilizing chest X-ray data. Here, the idea lies in improving the residual units of ResNet50 using the squeeze-and-excitation attention mechanism. For further enhancement, the Ranger optimizer and adaptive Mish activation function are employed to improve the feature learning of the SEA-ResNet50 model. For evaluation, two publicly available COVID-19 radiographic datasets are utilized. The chest X-ray input images are augmented during experimentation for robust evaluation against four output classes namely normal, pneumonia, lung opacity, and COVID-19. Then a comparative study is done for the SEA-ResNet50 model against VGG-16, Xception, ResNet18, ResNet50, and DenseNet121 architectures. The proposed framework of SEA-ResNet50 together with the Ranger optimizer and adaptive Mish activation provided maximum classification accuracies of 98.38% (multiclass) and 99.29% (binary classification) as compared with the existing CNN architectures. The proposed method achieved the highest Kappa validation scores of 0.975 (multiclass) and 0.98 (binary classification) over others. Furthermore, the visualization of the saliency maps of the abnormal regions is represented using the explainable artificial intelligence (XAI) model, thereby enhancing interpretability in disease diagnosis.

Impact of harmonization on the reproducibility of MRI radiomic features when using different scanners, acquisition parameters, and image pre-processing techniques: a phantom study

This study investigated the impact of ComBat harmonization on the reproducibility of radiomic features extracted from magnetic resonance images (MRI) acquired on different scanners, using various data acquisition parameters and multiple image pre-processing techniques using a dedicated MRI phantom. Four scanners were used to acquire an MRI of a nonanatomic phantom as part of the TCIA RIDER database. In fast spin-echo inversion recovery (IR) sequences, several inversion durations were employed, including 50, 100, 250, 500, 750, 1000, 1500, 2000, 2500, and 3000 ms. In addition, a 3D fast spoiled gradient recalled echo (FSPGR) sequence was used to investigate several flip angles (FA): 2, 5, 10, 15, 20, 25, and 30 degrees. Nineteen phantom compartments were manually segmented. Different approaches were used to pre-process each image: Bin discretization, Wavelet filter, Laplacian of Gaussian, logarithm, square, square root, and gradient. Overall, 92 first-, second-, and higher-order statistical radiomic features were extracted. ComBat harmonization was also applied to the extracted radiomic features. Finally, the Intraclass Correlation Coefficient (ICC) and Kruskal-Wallis's (KW) tests were implemented to assess the robustness of radiomic features. The number of non-significant features in the KW test ranged between 0-5 and 29-74 for various scanners, 31-91 and 37-92 for three times tests, 0-33 to 34-90 for FAs, and 3-68 to 65-89 for IRs before and after ComBat harmonization, with different image pre-processing techniques, respectively. The number of features with ICC over 90% ranged between 0-8 and 6-60 for various scanners, 11-75 and 17-80 for three times tests, 3-83 to 9-84 for FAs, and 3-49 to 3-63 for IRs before and after ComBat harmonization, with different image pre-processing techniques, respectively. The use of various scanners, IRs, and FAs has a great impact on radiomic features. However, the majority of scanner-robust features is also robust to IR and FA. Among the effective parameters in MR images, several tests in one scanner have a negligible impact on radiomic features. Different scanners and acquisition parameters using various image pre-processing might affect radiomic features to a large extent. ComBat harmonization might significantly impact the reproducibility of MRI radiomic features.

Contrast enhancement boost improves the image quality of CT angiography derived from 80-kVp cerebral CT perfusion data

RATIONALE AND OBJECTIVE

To investigate the impact of the contrast enhancement boost (CE-boost) technique on the image quality of CT angiography (CTA) derived from 80-kVp cerebral CT perfusion (CTP) data, and to compare it with conventional CTApeak as well as other currently employed methods for enhancing CTA images, such as CTAtMIP and CTAtAve extracted from CTP.

MATERIALS AND METHODS

The data of forty-seven patients who underwent CTP at 80 kVp were retrospectively collected. Four sets of images: CTApeak, CTAtMIP, CTAtAve, and CE-boost images. The CTApeak image represents the arterial phase at its peak value, captured as a single time point. CTAtMIP and CTAtAve are 4D CTA images that provide maximum density projection and average images from the three most prominent time points. CE-boost is a postprocessing technique used to enhance contrast in the arterial phase at its peak value. We compared the average CT value, standard deviation (SD), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) of the internal carotid artery (ICA) and basilar artery (BA) among the four groups. Image quality was evaluated using a 5-point scale.

RESULTS

The CE-boost demonstrated and CNR in the ICA and BA (all p < 0.001). Compared with the other three CTA reconstructed images, the CE-boost images had the best subjective image quality, with the highest scores of 4.77 ± 0.43 and 4.87 ± 0.34 for each reader (all p < 0.001).

CONCLUSION

Compared with other currently used techniques,CE-boost enhances the image quality of CTA derived from 80-kVp CTP data, leading to improved visualization of intracranial arteries.

Value of fractional-order calculus (FROC) model diffusion-weighted imaging combined with simultaneous multi-slice (SMS) acceleration technology for evaluating benign and malignant breast lesions

BACKGROUND

This study explores the diagnostic value of combining fractional-order calculus (FROC) diffusion-weighted model with simultaneous multi-slice (SMS) acceleration technology in distinguishing benign and malignant breast lesions.

METHODS

178 lesions (73 benign, 105 malignant) underwent magnetic resonance imaging with diffusion-weighted imaging using multiple b-values (14 b-values, highest 3000 s/mm2). Independent samples t-test or Mann-Whitney U test compared image quality scores, FROC model parameters (D,, ), and ADC values between two groups. Multivariate logistic regression analysis identified independent variables and constructed nomograms. Model discrimination ability was assessed with receiver operating characteristic (ROC) curve and calibration chart. Spearman correlation analysis and Bland-Altman plot evaluated parameter correlation and consistency.

RESULTS

Malignant lesions exhibited lower D, and ADC values than benign lesions (P < 0.05), with higher values (P < 0.05). In SSEPI-DWI and SMS-SSEPI-DWI sequences, the AUC and diagnostic accuracy of D value are maximal, with D value demonstrating the highest diagnostic sensitivity, while value exhibits the highest specificity. The D and combined model had the highest AUC and accuracy. D and ADC values showed high correlation between sequences, and moderate. Bland-Altman plot demonstrated unbiased parameter values.

CONCLUSION

SMS-SSEPI-DWI FROC model provides good image quality and lesion characteristic values within an acceptable time. It shows consistent diagnostic performance compared to SSEPI-DWI, particularly in D and values, and significantly reduces scanning time.

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

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

Radiomics Beyond the Hype: A Critical Evaluation Toward Oncologic Clinical Use

Radiomics is a promising and fast-developing field within oncology that involves the mining of quantitative high-dimensional data from medical images. Radiomics has the potential to transform cancer management, whereby radiomics data can be used to aid early tumor characterization, prognosis, risk stratification, treatment planning, treatment response assessment, and surveillance. Nevertheless, certain challenges have delayed the clinical adoption and acceptability of radiomics in routine clinical practice. The objectives of this report are to (a) provide a perspective on the translational potential and potential impact of radiomics in oncology; (b) explore frequent challenges and mistakes in its derivation, encompassing study design, technical requirements, standardization, model reproducibility, transparency, data sharing, privacy concerns, quality control, as well as the complexity of multistep processes resulting in less radiologist-friendly interfaces; (c) discuss strategies to overcome these challenges and mistakes; and (d) propose measures to increase the clinical use and acceptability of radiomics, taking into account the different perspectives of patients, health care workers, and health care systems. Keywords: Radiomics, Oncology, Cancer Management, Artificial Intelligence © RSNA, 2024.

Radiomics and Deep Features: Robust Classification of Brain Hemorrhages and Reproducibility Analysis Using a 3D Autoencoder Neural Network

This study evaluates the reproducibility of machine learning models that integrate radiomics and deep features (features extracted from a 3D autoencoder neural network) to classify various brain hemorrhages effectively. Using a dataset of 720 patients, we extracted 215 radiomics features (RFs) and 15,680 deep features (DFs) from CT brain images. With rigorous screening based on Intraclass Correlation Coefficient thresholds (>0.75), we identified 135 RFs and 1054 DFs for analysis. Feature selection techniques such as Boruta, Recursive Feature Elimination (RFE), XGBoost, and ExtraTreesClassifier were utilized alongside 11 classifiers, including AdaBoost, CatBoost, Decision Trees, LightGBM, Logistic Regression, Naive Bayes, Neural Networks, Random Forest, Support Vector Machines (SVM), and k-Nearest Neighbors (k-NN). Evaluation metrics included Area Under the Curve (AUC), Accuracy (ACC), Sensitivity (SEN), and F1-score. The model evaluation involved hyperparameter optimization, a 70:30 train-test split, and bootstrapping, further validated with the Wilcoxon signed-rank test and q-values. Notably, DFs showed higher accuracy. In the case of RFs, the Boruta + SVM combination emerged as the optimal model for AUC, ACC, and SEN, while XGBoost + Random Forest excelled in F1-score. Specifically, RFs achieved AUC, ACC, SEN, and F1-scores of 0.89, 0.85, 0.82, and 0.80, respectively. Among DFs, the ExtraTreesClassifier + Naive Bayes combination demonstrated remarkable performance, attaining an AUC of 0.96, ACC of 0.93, SEN of 0.92, and an F1-score of 0.92. Distinguished models in the RF category included SVM with Boruta, Logistic Regression with XGBoost, SVM with ExtraTreesClassifier, CatBoost with XGBoost, and Random Forest with XGBoost, each yielding significant q-values of 42. In the DFs realm, ExtraTreesClassifier + Naive Bayes, ExtraTreesClassifier + Random Forest, and Boruta + k-NN exhibited robustness, with 43, 43, and 41 significant q-values, respectively. This investigation underscores the potential of synergizing DFs with machine learning models to serve as valuable screening tools, thereby enhancing the interpretation of head CT scans for patients with brain hemorrhages.

Deep Learning for Alzheimer's Disease Prediction: A Comprehensive Review

Alzheimer's disease (AD) is a neurological disorder that significantly impairs cognitive function, leading to memory loss and eventually death. AD progresses through three stages: early stage, mild cognitive impairment (MCI) (middle stage), and dementia. Early diagnosis of Alzheimer's disease is crucial and can improve survival rates among patients. Traditional methods for diagnosing AD through regular checkups and manual examinations are challenging. Advances in computer-aided diagnosis systems (CADs) have led to the development of various artificial intelligence and deep learning-based methods for rapid AD detection. This survey aims to explore the different modalities, feature extraction methods, datasets, machine learning techniques, and validation methods used in AD detection. We reviewed 116 relevant papers from repositories including Elsevier (45), IEEE (25), Springer (19), Wiley (6), PLOS One (5), MDPI (3), World Scientific (3), Frontiers (3), PeerJ (2), Hindawi (2), IO Press (1), and other multiple sources (2). The review is presented in tables for ease of reference, allowing readers to quickly grasp the key findings of each study. Additionally, this review addresses the challenges in the current literature and emphasizes the importance of interpretability and explainability in understanding deep learning model predictions. The primary goal is to assess existing techniques for AD identification and highlight obstacles to guide future research.

Development and validation of a risk prediction model for invasiveness of pure ground-glass nodules based on a systematic review and meta-analysis

BACKGROUND

Assessing the aggressiveness of pure ground glass nodules early on significantly aids in making informed clinical decisions.

OBJECTIVE

Developing a predictive model to assess the aggressiveness of pure ground glass nodules in lung adenocarcinoma is the study's goal.

METHODS

A comprehensive search for studies on the relationship between computed tomography(CT) characteristics and the aggressiveness of pure ground glass nodules was conducted using databases such as PubMed, Embase, Web of Science, Cochrane Library, Scopus, Wanfang, CNKI, VIP, and CBM, up to December 20, 2023. Two independent researchers were responsible for screening literature, extracting data, and assessing the quality of the studies. Meta-analysis was performed using Stata 16.0, with the training data derived from this analysis. To identify publication bias, Funnel plots and Egger tests and Begg test were employed. This meta-analysis facilitated the creation of a risk prediction model for invasive adenocarcinoma in pure ground glass nodules. Data on clinical presentation and CT imaging features of patients treated surgically for these nodules at the Third Affiliated Hospital of Kunming Medical University, from September 2020 to September 2023, were compiled and scrutinized using specific inclusion and exclusion criteria. The model's effectiveness for predicting invasive adenocarcinoma risk in pure ground glass nodules was validated using ROC curves, calibration curves, and decision analysis curves.

RESULTS

In this analysis, 17 studies were incorporated. Key variables included in the model were the largest diameter of the lesion, average CT value, presence of pleural traction, and spiculation. The derived formula from the meta-analysis was: 1.16×the largest lesion diameter + 0.01 × the average CT value + 0.66 × pleural traction + 0.44 × spiculation. This model underwent validation using an external set of 512 pure ground glass nodules, demonstrating good diagnostic performance with an ROC curve area of 0.880 (95% CI: 0.852-0.909). The calibration curve indicated accurate predictions, and the decision analysis curve suggested high clinical applicability of the model.

CONCLUSION

We established a predictive model for determining the invasiveness of pure ground-glass nodules, incorporating four key radiological indicators. This model is both straightforward and effective for identifying patients with a high likelihood of invasive adenocarcinoma.

Does consensus contour improve robustness and accuracy in 18F-FDG PET radiomic features?

PURPOSE

The purpose of our study is to validate the robustness and accuracy of consensus contour in 2-deoxy-2-[18 F]fluoro-D-glucose (18 F-FDG) PET radiomic features.

METHODS

225 nasopharyngeal carcinoma (NPC) and 13 extended cardio-torso (XCAT) simulated data were enrolled. All segmentation were performed with four segmentation methods under two different initial masks, respectively. Consensus contour (ConSeg) was then developed using the majority vote rule. 107 radiomic features were extracted by Pyradiomics based on segmentation and the intraclass correlation coefficient (ICC) was calculated for each feature between masks or among segmentation, respectively. In XCAT ICC between segmentation and simulated ground truth were also calculated to access the accuracy.

RESULTS

ICC varied with the dataset, segmentation method, initial mask and feature type. ConSeg presented higher ICC for radiomic features in robustness tests and similar ICC in accuracy tests, compared with the average of four segmentation results. Higher ICC were also generally observed in irregular initial masks compared with rectangular masks in both robustness and accuracy tests. Furthermore, 19 features (17.76%) had ICC ≥ 0.75 in both robustness and accuracy tests for any of the segmentation methods or initial masks. The dataset was observed to have a large impact on the correlation relationships between radiomic features, but not the segmentation method or initial mask.

CONCLUSIONS

The consensus contour combined with irregular initial mask could improve the robustness and accuracy in radiomic analysis to some extent. The correlation relationships between radiomic features and feature clusters largely depended on the dataset, but not segmentation method or initial mask.