Non-invasive Imaging Biomarkers Research Articles

Repeatability, Reproducibility, and Observer Variability of Cortical T1 Mapping for Renal Tissue Characterization.

The global rise in kidney diseases underscores the need for reliable, noninvasive imaging biomarkers. Among these, renal cortical T1 has shown promise but further technical validation is still required. To evaluate the repeatability, reproducibility, and observer variability of kidney cortical T1 mapping in human volunteers without known renal disease. Prospective. Three cohorts without renal disease: 1) 25 volunteers (median age 38 [interquartile range, IQR: 28-42] years, female N = 11) for scan-rescan assessments on GE 1.5 T and Siemens 1.5 T; 2) 29 volunteers (median age 29 [IQR: 24-40] years, female N = 15) for scan-rescan assessments on Siemens 3 T; and 3) 16 volunteers (median age 34 [IQR: 31-42] years, female N = 8) for cross-scanner reproducibility. 1.5 T and 3 T, a modified Look-Locker imaging (MOLLI) sequence with a balanced steady-state free precession (bSSFP) readout. Kidney cortical T1 data was acquired on GE 1.5 T scanner, Siemens 1.5 T and 3 T scanners. Within-scanner repeatability and inter/intra-observer variability: GE 1.5 T and Siemens 1.5 T, and cross-scanner manufacturer reproducibility: Siemens 1.5 T-GE 1.5 T. Bland Altman analysis, coefficient of variation (CoV), intra-class coefficient (ICC), and repeatability coefficient (RC). Renal cortical T1 mapping showed high repeatability and reliability across scanner field strengths and manufacturers (repeatability: CoV 1.9%-2.8%, ICC 0.79-0.88, pooled RC 73 msec; reproducibility: CoV 3.0%, ICC 0.75, RC 90 msec). The method also showed robust observer variability (CoV 0.6%-1.4%, ICC 0.93-0.98, RC 22-48 msec). Kidney cortical T1 mapping is a highly repeatable and reproducible method across MRI manufacturers, field strengths, and observer conditions. 2 TECHNICAL EFFICACY: Stage 2.

Metabolic dysfunction-associated steatotic liver disease and cardiovascular risk: a comprehensive review.

Metabolic dysfunction-associated steatotic liver disease (MASLD), previously termed nonalcoholic fatty liver disease (NAFLD), poses a significant global health challenge due to its increasing prevalence and strong association with cardiovascular disease (CVD). This comprehensive review summarizes the current knowledge on the MASLD-CVD relationship, compares analysis of how different terminologies for fatty liver disease affect cardiovascular (CV) risk assessment using different diagnostic criteria, explores the pathophysiological mechanisms connecting MASLD to CVD, the influence of MASLD on traditional CV risk factors, the role of noninvasive imaging techniques and biomarkers in the assessment of CV risk in patients with MASLD, and the implications for clinical management and prevention strategies. By incorporating current research and clinical guidelines, this review provides a comprehensive overview of the complex interplay between MASLD and cardiovascular health.

An MRI-based radiomics nomogram to predict progression-free survival in patients with endometrial cancer.

Aim: This study aimed to explore the importance of an MRI-based radiomics nomogram in predicting the progression-free survival (PFS) of endometrial cancer.Methods: Based on clinicopathological and radiomic characteristics, we established three models (clinical, radiomics and combined model) and developed a nomogram for the combined model. The Kaplan-Meier method was utilized to evaluate the association between nomogram-based risk scores and PFS.Results: The nomogram had a strong predictive ability in calculating PFS with areas under the curve (ROC) of 0.905 and 0.901 at 1 and 3years, respectively. The high-risk groups identified by the nomogram-based scores had shorter PFS compared with the low-risk groups.Conclusion: The radiomics nomogram has the potential to serve as a noninvasive imaging biomarker for predicting individual PFS of endometrial cancer.

Imaging evaluation focused on microstructural tissue changes using tensor-valued diffusion encoding in breast cancers after neoadjuvant chemotherapy: is it a promising way forward?

Single diffusion encoding is a widely used, noninvasive technique for probing the tissue microstructure in breast tumors. However, it does not provide detailed information about the microenvironmental complexity. This study investigated the clinical utility of tensor-valued diffusion encoding for evaluating microstructural changes in breast cancer after neoadjuvant chemotherapy (NAC). We retrospectively included patients underwent chemotherapy for histologically proven invasive breast cancer between July 2020 and June 2023 and monitored the tumor response with breast magnetic resonance imaging (MRI), including tensor-valued diffusion encoding. We reviewed pre- and post-NAC MRIs regarding chemotherapy in 23 breast cancers. Q-space trajectory imaging (QTI) parameters were estimated at each time-point, and were compared with histopathological parameters. The mean total mean kurtosis (MKT), anisotropic mean kurtosis (MKA), and microscopic fractional anisotropy (µFA) were significantly decreased on post-NAC MRI compared with pre-NAC MRI, with the large effect size (ES) in MKA and µFA (0.81±0.41 vs. 0.99±0.33, ES: 0.48, P=0.03; 0.48±0.30 vs. 0.73±0.27, ES: 0.88, P<0.001; 0.58±0.14 vs. 0.68±0.11, ES: 0.79, P=0.003; respectively). Regarding prognostic factors, tumors with high Ki-67 expression showed significantly lower pre-NAC mean diffusivity (MD) and higher pre-NAC µFA compared to tumors with low Ki-67 expression (0.98±0.09 vs. 1.25±0.20, P=0.002; and 0.72±0.07 vs. 0.57±0.10, P=0.005; respectively). And negative progesterone receptor (PR) group revealed significantly lower MKT, MKA, and isotropic mean kurtosis than positive PR group on the post-NAC MRI (0.60±0.31 vs. 1.03±0.40, P=0.008; 0.36±0.21 vs. 0.61±0.33, P=0.04; and 0.23±0.17 vs. 0.42±0.25, P=0.046; respectively). QTI parameters reflected the microstructural changes in breast cancer treated with NAC and can be used as noninvasive imaging biomarkers correlated with prognostic factors.

Diagnostic accuracy of an uncorrected native T1 mapping sequence for liver fibrosis and inflammation in autoimmune hepatitis: a prospective study using histopathology as reference standard.

There is an unmet clinical need for non-invasive imaging biomarkers that could replace liver biopsy in the management of patients with autoimmune hepatitis (AIH). In this study, we sought to evaluate the diagnostic accuracy of a simple uncorrected, non-contrast T1 mapping for detecting fibrosis and inflammation in AIH patients using histopathology as a reference standard. Over 3years, 33 patients with AIH were prospectively studied using a multiparametric liver MRI protocol which included T1 mapping. Biopsies were performed up to 3months before imaging, and a standardized histopathological score for fibrosis (F0-F4) and inflammatory activity (PPA0-4) was used as a reference. Statistical analysis included independent t test, Mann-Whitney U-test, and ROC (receiver operating characteristic) analysis. T1 mapping values were significantly higher in patients with advanced fibrosis (F0-2 vs. F3-4; p < 0.015), significant fibrosis (F0-1 vs. F2-4; p < 0.005), and significant inflammatory activity (PPA 0-1 vs. PPA 2-4 p = 0.048). Moreover, the technique demonstrated a good diagnostic performance in detecting significant (AUC 0.856) and advanced fibrosis (AUC 0.835), as well as significant inflammatory activity (AUC 0.763). A rapid, simple, uncorrected, non-contrast T1 mapping sequence showed satisfactory diagnostic performance in comparison with histopathology for detecting significant tissue inflammation and fibrosis in AIH patients, being a potential non-invasive imaging biomarker for monitoring disease activity in such individuals.

CD206+ macrophages are relevant non-invasive imaging biomarkers and therapeutic targets in experimental lung fibrosis

BackgroundInterstitial lung diseases (ILDs) include a large number of diseases associated with progressive pulmonary fibrosis (PPF), including idiopathic pulmonary fibrosis (IPF). Despite the rarity of each of the fibrotic ILDs...

The Use of Apparent Diffusion Coefficient Values for Differentiating Bevacizumab-Related Cytotoxicity from Tumor Recurrence and Radiation Necrosis in Glioblastoma.

Glioblastomas (GBM) are the most common primary invasive neoplasms of the brain. Distinguishing between lesion recurrence and different types of treatment related changes in patients with GBM remains challenging using conventional MRI imaging techniques. Therefore, accurate and precise differentiation between true progression or pseudoresponse is crucial in deciding on the appropriate course of treatment. This retrospective study investigated the potential of apparent diffusion coefficient (ADC) map values derived from diffusion-weighted imaging (DWI) as a noninvasive method to increase diagnostic accuracy in treatment response. A cohort of 21 glioblastoma patients (mean age: 59.2 ± 11.8, 12 Male, 9 Female) that underwent treatment with bevacizumab were selected. The ADC values were calculated from the DWI images obtained from a standardized brain protocol across 1.5-T and 3-T MRI scanners. Ratios were calculated for rADC values. Lesions were classified as bevacizumab-induced cytotoxicity based on characteristic imaging features (well-defined regions of restricted diffusion with persistent diffusion restriction over the course of weeks without tissue volume loss and absence of contrast enhancement). The rADC value was compared to these values in radiation necrosis and recurrent lesions, which were concluded in our prior study. The nonparametric Wilcoxon signed rank test with p < 0.05 was used for significance. The mean ± SD age of the selected patients was 59.2 ± 11.8. ADC values and corresponding mean rADC values for bevacizumab-induced cytotoxicity were 248.1 ± 67.2 and 0.39 ± 0.10, respectively. These results were compared to the ADC values and corresponding mean rADC values of tumor progression and radiation necrosis. Significant differences between rADC values were observed in all three groups (p < 0.001). Bevacizumab-induced cytotoxicity had statistically significant lower ADC values compared to both tumor recurrence and radiation necrosis. The study demonstrates the potential of ADC values as noninvasive imaging biomarkers for differentiating recurrent glioblastoma from radiation necrosis and bevacizumab-induced cytotoxicity.

Perivascular fat attenuation index value and plaque volume increased in non-target lesions of coronary arteries after stenting.

Progression of non-target lesions (NTLs) after stenting has been reported and is associated with the triggering of an inflammatory response. The perivascular fat attenuation index (FAI) may be used as a novel imaging biomarker for the direct quantification of coronary inflammation. To investigate whether FAI values can help identify changes in inflammation status in patients undergoing stent implantation, especially in NTLs. Patients who underwent pre- and post-stenting coronary computed tomography angiography (CCTA) examination between January 2015 and February 2021 were consecutively enrolled. The pre- and post-stenting FAIs of the full coronary arteries were compared in both the non- and stent-implanted coronary arteries. Moreover, local FAI values were measured and compared between the NTLs and target lesions in the stent implantations. We also compared changes in plaque type and volume in NTLs before and after stenting. A total of 89 patients (mean age 61 years; male 59) were enrolled. The perivascular FAI values in the full coronary arteries decreased after stenting in both the non- and stent-implanted coronary arteries, similar to those in the target lesions. Conversely, the perivascular FAI values in the NTLs increased after stenting (p < 0.05). In addition, the plaque volumes significantly increased in the NTLs after stenting, regardless of whether they were non-calcified, mixed, or calcified (p < 0.05). Perivascular FAI values and plaque volumes increased in the NTLs after stenting. Perivascular FAI can be a promising imaging biomarker for monitoring coronary inflammation after stenting and facilitate long-term monitoring in clinical settings. Perivascular fat attenuation index, a non-invasive imaging biomarker, may help identify coronary arteries with high inflammation in non-target lesions and facilitate long-term monitoring, potentially providing an opportunity for more targeted treatment. • Perivascular fat attenuation index (FAI) values and plaque volumes increased in the non-target lesions (NTLs) after stenting, suggesting potential focal inflammation progression after stenting. However, stenting along with anti-inflammatory treatment ameliorated inflammation in the full coronary arteries. • Perivascular FAI, a non-invasive imaging biomarker, may help identify coronary arteries with high inflammation in NTLs and facilitate long-term monitoring, potentially providing an opportunity for more targeted treatment.

Decision-Tree Models Indicative of Microvascular Invasion on MRI Predict Survival in Patients with Hepatocellular Carcinoma Following Tumor Ablation.

Histological microvascular invasion (MVI) is a risk factor for poor survival and early recurrence in hepatocellular carcinoma (HCC) after surgery. Its prognostic value in the setting of locoregional therapies (LRT), where no tissue samples are obtained, remains unknown. This study aims to establish CT-derived indices indicative of MVI on liver MRI with superior soft tissue contrast and evaluate their association with patient survival after ablation via interstitial brachytherapy (iBT) versus iBT combined with prior conventional transarterial chemoembolization (cTACE). Ninety-five consecutive patients, who underwent ablation via iBT alone (n = 47) or combined with cTACE (n = 48), were retrospectively included between 01/2016 and 12/2017. All patients received contrast-enhanced MRI prior to LRT. Overall (OS), progression-free survival (PFS), and time-to-progression (TTP) were assessed. Decision-tree models to determine Radiogenomic Venous Invasion (RVI) and Two-Trait Predictor of Venous Invasion (TTPVI) on baseline MRI were established, validated on an external test set (TCGA-LIHC), and applied in the study cohorts to investigate their prognostic value for patient survival. Statistics included Fisher's exact and t-test, Kaplan-Meier and cox-regression analysis, area under the receiver operating characteristic curve (AUC-ROC) and Pearson's correlation. OS, PFS, and TTP were similar in both treatment groups. In the external dataset, RVI showed low sensitivity but relatively high specificity (AUC-ROC = 0.53), and TTPVI high sensitivity but only low specificity (AUC-ROC = 0.61) for histological MVI. In patients following iBT alone, positive RVI and TTPVI traits were associated with poorer OS (RVI: p < 0.01; TTPVI: p = 0.08), PFS (p = 0.04; p = 0.04), and TTP (p = 0.14; p = 0.03), respectively. However, when patients with combined cTACE and iBT were stratified by RVI or TTPVI, no differences in OS (p = 0.75; p = 0.55), PFS (p = 0.70; p = 0.43), or TTP (p = 0.33; p = 0.27) were observed. The study underscores the role of non-invasive imaging biomarkers indicative of MVI to identify patients, who would potentially benefit from embolotherapy via cTACE prior to ablation rather than ablation alone.

Protocol and rationale of the Australian multicentre registry for serial cardiac computed tomography angiography (ARISTOCRAT): a prospective observational study of the natural history of pericoronary adipose tissue attenuation and radiomics.

Vascular inflammation plays a crucial role in the development of atherosclerosis and atherosclerotic plaque rupture resulting in acute coronary syndrome (ACS). Pericoronary adipose tissue (PCAT) attenuation quantified from routine coronary computed tomography angiography (CCTA) has emerged as a promising non-invasive imaging biomarker of coronary inflammation. However, a detailed understanding of the natural history of PCAT attenuation is required before it can be used as a surrogate endpoint in trials of novel therapies targeting coronary inflammation. This article aims to explore the natural history of PCAT attenuation and its association with changes in plaque characteristics. The Australian natuRal hISTOry of periCoronary adipose tissue attenuation, RAdiomics and plaque by computed Tomographic angiography (ARISTOCRAT) registry is a multi-centre observational registry enrolling patients undergoing clinically indicated serial CCTA in 9 centres across Australia. CCTA scan parameters will be matched across serial scans. Quantitative analysis of plaque and PCAT will be performed using semiautomated software. The primary endpoint is to explore temporal changes in patient-level and lesion-level PCAT attenuation by CCTA and their associations with changes in plaque characteristics. Secondary endpoints include evaluating: (I) impact of statin therapy on PCAT attenuation and plaque characteristics; and (II) changes in PCAT attenuation and plaque characteristics in specific subgroups according to sex and risk factors. ARISTOCRAT will further our understanding of the natural history of PCAT attenuation and its association with changes in plaque characteristics. This study has been prospectively registered with the Australia and New Zealand Clinical Trials Registry (ACTRN12621001018808).

Dual-source dual-energy CT and deep learning for equivocal lymph nodes on CT images for thyroid cancer.

This study investigated the diagnostic performance of dual-energy computed tomography (CT) and deep learning for the preoperative classification of equivocal lymph nodes (LNs) on CT images in thyroid cancer patients. In this prospective study, from October 2020 to March 2021, 375 patients with thyroid disease underwent thin-section dual-energy thyroid CT at a small field of view (FOV) and thyroid surgery. The data of 183 patients with 281 LNs were analyzed. The targeted LNs were negative or equivocal on small FOV CT images. Six deep-learning models were used to classify the LNs on conventional CT images. The performance of all models was compared with pathology reports. Of the 281 LNs, 65.5% had a short diameter of less than 4 mm. Multiple quantitative dual-energy CT parameters significantly differed between benign and malignant LNs. Multivariable logistic regression analyses showed that the best combination of parameters had an area under the curve (AUC) of 0.857, with excellent consistency and discrimination, and its diagnostic accuracy and sensitivity were 74.4% and 84.2%, respectively (p < 0.001). The visual geometry group 16 (VGG16) based model achieved the best accuracy (86%) and sensitivity (88%) in differentiating between benign and malignant LNs, with an AUC of 0.89. The VGG16 model based on small FOV CT images showed better diagnostic accuracy and sensitivity than the spectral parameter model. Our study presents a noninvasive and convenient imaging biomarker to predict malignant LNs without suspicious CT features in thyroid cancer patients. Our study presents a deep-learning-based model to predict malignant lymph nodes in thyroid cancer without suspicious features on conventional CT images, which shows better diagnostic accuracy and sensitivity than the regression model based on spectral parameters. Many cervical lymph nodes (LNs) do not express suspicious features on conventional computed tomography (CT). Dual-energy CT parameters can distinguish between benign and malignant LNs. Visual geometry group 16 model shows superior diagnostic accuracy and sensitivity for malignant LNs.

Radiomics of pericoronary adipose tissue on computed tomography angiography predicts coronary heart disease in patients with type 2 diabetes mellitus

BackgroundDiabetes is a common chronic metabolic disease. The progression of the disease promotes vascular inflammation and the formation of atherosclerosis, leading to cardiovascular disease. The coronary artery perivascular adipose tissue attenuation index based on CCTA is a new noninvasive imaging biomarker that reflects the spatial changes in perivascular adipose tissue attenuation in CCTA images and the inflammation around the coronary arteries. In this study, a radiomics approach is proposed to extract a large number of image features from CCTA in a high-throughput manner and combined with clinical diagnostic data to explore the predictive ability of vascular perivascular adipose imaging data based on CCTA for coronary heart disease in diabetic patients.MethodsR language was used for statistical analysis to screen the variables with significant differences. A presegmentation model was used for CCTA vessel segmentation, and the pericoronary adipose region was screened out. PyRadiomics was used to calculate the radiomics features of pericoronary adipose tissue, and SVM, DT and RF were used to model and analyze the clinical data and radiomics data. Model performance was evaluated using indicators such as PPV, FPR, AAC, and ROC.ResultsThe results indicate that there are significant differences in age, blood pressure, and some biochemical indicators between diabetes patients with and without coronary heart disease. Among 1037 calculated radiomic parameters, 18.3% showed significant differences in imaging omics features. Three modeling methods were used to analyze different combinations of clinical information, internal vascular radiomics information and pericoronary vascular fat radiomics information. The results showed that the dataset of full data had the highest ACC values under different machine learning models. The support vector machine method showed the best specificity, sensitivity, and accuracy for this dataset.ConclusionsIn this study, the clinical data and pericoronary radiomics data of CCTA were fused to predict the occurrence of coronary heart disease in diabetic patients. This provides information for the early detection of coronary heart disease in patients with diabetes and allows for timely intervention and treatment.

Application of TSPO-Specific Positron Emission Tomography Radiotracer as an Early Indicator of Acute Liver Failure Induced by Propacetamol, a Prodrug of Paracetamol.

Acetaminophen overdose is a leading cause of acute liver failure (ALF), and effective treatment depends on early prediction of disease progression. ALF diagnosis currently requires blood collection 24-72 h after APAP ingestion, necessitating repeated tests and hospitalization. Here, we assessed earlier ALF diagnosis using positron emission tomography (PET) imaging of translocator proteins (TSPOs), which are involved in molecular transport, oxidative stress, apoptosis, and energy metabolism, with the radiotracer [18F]GE180. We intraperitoneally administered propacetamol hydrochloride to male C57BL/6 mice to induce ALF. We performed in vivo PET/CT imaging 3 h later using the TSPO-specific radiotracer [18F]GE180 and quantitatively analyzed the PET images by determining the averaged standardized uptake value (SUVav) in the liver parenchyma. We assessed liver TSPO expression levels via real-time polymerase chain reaction, Western blotting, and immunohistochemistry. [18F]GE180 PET imaging 3 h after propacetamol administration (1500 mg/kg) significantly increased liver SUVav compared to controls (p = 0.001). Analyses showed a 10-fold and 4-fold increase in TSPO gene and protein expression, respectively, in the liver, 3 h after propacetamol induction compared to controls. [18F]GE180 PET visualized and quantified propacetamol-induced ALF through TSPO overexpression. These findings highlight TSPO PET's potential as a non-invasive imaging biomarker for early-stage ALF.

Quantitative Abdominal Arterial Calcification Correlates with Kidney Transplant Waitlist Mortality

Introduction: The scarcity of available organs for kidney transplantation has resulted in a substantial waiting time for patients with end-stage kidney disease. This prolonged wait contributes to an increased risk of cardiovascular mortality. Calcification of large arteries is a high-risk factor in the development of cardiovascular diseases, and it is common among candidates for kidney transplant. The aim of this study was to correlate abdominal arterial calcification (AAC) score value with mortality on the waitlist. Methods: We modified the coronary calcium score and used it to quantitate the AAC. We conducted a retrospective clinical study of all adult patients who were listed for kidney transplant, between 2005 and 2015, and had abdominal computed tomography scan. Patients were divided into two groups: those who died on the waiting list group and those who survived on the waiting list group. Results: Each 1,000 increase in the AAC score value of the sum score of the abdominal aorta, bilateral common iliac, bilateral external iliac, and bilateral internal iliac was associated with increased risk of death (HR 1.034, 95% CI: 1.013, 1.055) (p = 0.001). This association remained significant even after adjusting for various patient characteristics, including age, tobacco use, diabetes, coronary artery disease, and dialysis status. Conclusion: The study highlights the potential value of the AAC score as a noninvasive imaging biomarker for kidney transplant waitlist patients. Incorporating the AAC scoring system into routine imaging reports could facilitate improved risk assessment and personalized care for kidney transplant candidates.

Use of Non-Invasive Elastographic and Microvascularization Biomarkers in the Diagnosis and Follow-Up of Children with Severe Bacterial Pneumonia

Imaging tests such as Chest X-Ray (CXR) and Lung Ultrasound (LUS), and blood biomarkers like C-reactive protein and procalcitonin have demonstrated its role for Bacterial Pneumonia (BP) diagnosis. However, the implementation of novel biomarkers such as Lung Shear-Wave Elastography (LSWE) or the quantification of lung microvascularization ratio pattern by Superb Microvascular Imaging (SMI) has not yet been assessed. These biomarkers combined with the already described LUS findings, could be used to increase BP diagnosis accuracy, improving the sensitivity and specificity of its diagnosis. In this article, we describe the use of these non-invasive novel imaging biomarkers for the diagnosis approach in a pediatric patient with severe BP.

Progressive Age-Associated Blood-Brain Barrier Leak/Dysfunction-Nexus of Neurodegenerative Disease Using MRI Markers to Identify Preclinical Disease and Potential New Targets for Future Treatments.

This review article focuses on the upstream pertinent pathophysiology leading to neurodegenerative disease. Specifically, the nexus appears to be blood-brain barrier (BBB) leakiness resulting in a two-prong inflammatory disease spectrum damaging the microvasculature and corrupting protein synthesis and degradation with accumulating misfolded toxic proteins. The suboptimal results of removing misfolded proteins mean a new approach to disease in the preclinical state is required aimed at other targets. Validated noninvasive imaging and serologic biomarkers of early preclinical disease implemented in the high-risk patient cohort along with periodic surveillance once effective treatments are developed will be required. This review discusses the physiology and pathophysiology of the BBB, new MRI imaging techniques identifying the leak, and altered fluid dynamic effects in the preclinical state. The risk factors for disease development, preventative measures, and potential treatment targets are also discussed.

Deep-Learning-Based Predictive Imaging Biomarker Model for EGFR Mutation Status in Non-Small Cell Lung Cancer from CT Imaging.

The authors aimed to develop and validate deep-learning-based radiogenomic (DLR) models and radiomic signatures to predict the EGFR mutation in patients with NSCLC, and to assess the semantic and clinical features that can contribute to detecting EGFR mutations. Using 990 patients from two NSCLC trials, we employed an end-to-end pipeline analyzing CT images without precise segmentation. Two 3D convolutional neural networks segmented lung masses and nodules. The combined radiomics and DLR model achieved an AUC of 0.88 ± 0.03 in predicting EGFR mutation status, outperforming individual models. Semantic features further improved the model's accuracy, with an AUC of 0.88 ± 0.05. CT semantic features that were found to be significantly associated with EGFR mutations were pure solid tumours with no associated ground glass component (p < 0.03), the absence of peripheral emphysema (p < 0.03), the presence of pleural retraction (p = 0.004), the presence of fissure attachment (p = 0.001), the presence of metastatic nodules in both the tumour-containing lobe (p = 0.001) and the non-tumour-containing lobe (p = 0.001), the presence of ipsilateral pleural effusion (p = 0.04), and average enhancement of the tumour mass above 54 HU (p < 0.001). This AI-based radiomics and DLR model demonstrated high accuracy in predicting EGFR mutation, serving as a non-invasive and user-friendly imaging biomarker for EGFR mutation status prediction.

Arterial spin labeling combined with T1 mapping for assessment of kidney function and histopathology in patients with long-term renal transplant survival after kidney transplantation.

The long-term survival of kidney transplants is often influenced by various factors, among which renal allograft rejection is the most notable factor. A noninvasive and reliable imaging biomarker correlating with kidney function and histopathology would facilitate longitudinal long-term follow-up of renal allografts. The aim of the study is to investigate the value of arterial spin labeling (ASL) combined with T1 mapping for assessing kidney function in patients with long-term renal transplant survival, and to establish radiological and histopathologic correlations between the magnetic resonance imaging (MRI) measurements and kidney allograft biopsy findings. Kidney transplant recipients who were admitted to the Department of Urology in First Affiliated Hospital of Soochow University between January and December 2022 were prospectively consecutively recruited [group A, estimated glomerular filtration rate (eGFR) ≥60 mL/min/1.73 m2; group B, 30≤ eGFR <60 mL/min/1.73 m2; group C, eGFR <30 mL/min/1.73 m2], and part of them underwent biopsies. All patients underwent ASL and T1 mapping. MRI parameters were calculated and analyzed. A total of 63 patients (Group A, 30 cases; Group B, 20 cases; and Group C, 13 cases) were included in this cross-sectional study. Cortical T1 increased, whereas renal blood flow (RBF) and ΔT1 [100% × (cortical T1 - medullary T1)/cortical T1] decreased with the decrease of eGFR. The RBF, cortical T1, and ΔT1 values were moderately correlated with eGFR (r=0.569, -0.573, and 0.672, respectively). The MRI parameters were moderately correlated with Banff scores, which determined renal allograft rejection and chronicity. The area under the curve (AUC) for the discrimination of groups A versus B and groups A versus C were 0.740 [95% confidence interval (CI): 0.597-0.854, P=0.004] and 0.923 (95% CI: 0.800-0.982, P<0.001), respectively, using ASL; 0.873 (95% CI: 0.749-0.950, P<0.001) and 0.926 (95% CI: 0.803-0.983, P<0.001), respectively, using T1 mapping; and 0.892 (95% CI: 0.771-0.962, P<0.001) and 0.956 (95% CI: 0.846-0.995, P<0.001), respectively, using multi-parameter MRI. The AUC for discrimination between groups B and C was 0.729 (95% CI: 0.546-0.868, P=0.02) using ASL. The RBF, cortical T1, and ΔT1 can serve as new imaging biomarkers of kidney function and histopathological microstructure.

Magnetic resonance imaging-based radiomics was used to evaluate the level of prognosis-related immune cell infiltration in breast cancer tumor microenvironment.

The tumor immune microenvironment is a valuable source of information for predicting prognosis in breast cancer (BRCA) patients. To identify immune cells associated with BRCA patient prognosis from the Cancer Genetic Atlas (TCGA), we established an MRI-based radiomics model for evaluating the degree of immune cell infiltration in breast cancer patients. CIBERSORT was utilized to evaluate the degree of infiltration of 22 immune cell types in breast cancer patients from the TCGA database, and both univariate and multivariate Cox regressions were employed to determine the prognostic significance of immune cell infiltration levels in BRCA patients. We identified independent prognostic factors for BRCA patients. Additionally, we obtained imaging features from the Cancer Imaging Archive (TCIA) database for 73 patients who underwent preoperative MRI procedures, and used the Least Absolute Shrinkage and Selection Operator (LASSO) to select the best imaging features for constructing an MRI-based radiomics model for evaluating immune cell infiltration levels in breast cancer patients. According to the results of Cox regression analysis, M2 macrophages were identified as an independent prognostic factor for BRCA patients (HR = 32.288, 95% CI: 3.100-357.478). A total of nine significant features were selected to calculate the radiomics-based score. We established an intratumoral model with AUCs (95% CI) of 0.662 (0.495-0.802) and 0.678 (0.438-0.901) in the training and testing cohorts, respectively. Additionally, a peritumoral model was created with AUCs (95% CI) of 0.826 (0.710-0.924) and 0.752 (0.525-0.957), and a combined model was established with AUCs (95% CI) of 0.843 (0.723-0.938) and 0.744 (0.491-0.965). The peritumoral model demonstrated the highest diagnostic efficacy, with an accuracy, sensitivity, and specificity of 0.773, 0.727, and 0.818, respectively, in its testing cohort. The MRI-based radiomics model has the potential to evaluate the degree of immune cell infiltration in breast cancer patients, offering a non-invasive imaging biomarker for assessing the tumor microenvironment in this disease.

Abstract WMP20: Deep Learning Radiomics Identifies White Matter Hyperintensity-Related Cognitive Decline Based on T2-FLAIR

Introduction: Early identification of white matter hyperintensity-related cognitive impairment (WMH-CI) through normal MRI images is of great significance for early clinical intervention to reduce the occurrence of dementia. The aim of this study was to develop a generalizable and interpretable deep learning model for screening WMH-related CI using radiomic features (RFs) from T2 fluid-attenuated inversion recovery (T2-FLAIR) images. Methods: A total of 783 subjects were enrolled from three medical centres. RFs for WMH lesions were extracted from T2-FLAIR images in each subject. A deep learning model with a hierarchical transformer architecture was used to leverage all extracted RFs, instead of feature selection, to develop and cross-validate the diagnostic model for cognitive dysfunction. Unsupervised domain adaptation was used to address the cross-centre data inconsistencies without labelling the outer centre data. A gradient-weighted class activation mapping approach was adopted to identify the RFs that contributed most to the diagnosis. Subgroup analysis, correlation analysis and mediation analysis were performed to confirm the clinical relevance of the model. Besides, we also verify the clinical importance of the model by microstructural pathology of WMH detected by diffusion tensor imaging. Results: The deep learning model showed robust diagnostic power for WMH related CI, with an area under the receiver operating curve (AUC) of 0.841±0.016 in the development cohort. The prediction accuracy, sensitivity, specificity, precision and recall were 0.793 ± 0.108, 0.798 ± 0.021, 0.800 ± 0.065, 0.716±0.055 and 0.793 ± 0.108, respectively. The model generalized well across different subgroups. It also performed well in other two external verification cohorts, with an AUC of 0.859 and 0.749, respectively. The visual representation showed that the most important features were textural features, which were also significantly correlated with clinical assessment scale and diffusion parameters. Conclusions: This study presents a non-invasive imaging biomarker that can identify WMH-CI patients and is applicable to all levels of hospitals since only conventional MRI images are needed.