European Journal of Radiology Open最新文献

Objectives

To investigate the correlations between IMPeTUs-based 18 F-FDG PET/CT parameters and clinical features in patients with newly diagnosed multiple myeloma (MM).

Materials and methods

PET/CT were analysed according to the IMPeTUs criteria. We correlated these PET/CT parameters with known clinically relevant features, bone marrow plasma cell (BMPC) infiltration rate and the presence of cytogenetic abnormalities.

Results

A total of 149 patients (86 males, 63 females; mean age, 59.9 ± 9.7 years) were included. Bone marrow metabolic state correlated with the most clinical features including hemoglobin (rho=-0.23, p=0.004), FLC ratio (rho=0.24, p=0.005), β2 M (rho=0.28, p=0.001), CRP (rho=0.25, p=0.003), serum calcium (rho=0.22, p=0.02), serum creatinine (rho=0.24, p=0.004) and BMPC (rho=0.21, p=0.003). Besides, the level of hemoglobin was significant lower (0.043), and the levels of FLC ratio (0.037), β2 M (p=0.024), CRP (p=0.05), and BMPC (p=0.043) were significant higher in patients having hypermetabolism in limbs and ribs. Hottest bone lesion Deauville criteria had a moderate correlation with CRP (rho=0.27, p=0.001) and serum calcium (rho=0.25, p=0.01).

Conclusion

Several IMPeTUs-based PET/CT parameters showed significant correlations with clinical features reflecting disease burden and biology, suggesting that these new criteria can be used in the risk stratification in MM patients.

Purpose

To assess radiomics and deep learning (DL) methods in identifying symptomatic Carotid Artery Disease (CAD) from carotid CT angiography (CTA) images. We further compare the performance of these novel methods to the conventional calcium score.

Methods

Carotid CT angiography (CTA) images from symptomatic patients (ischaemic stroke/transient ischaemic attack within the last 3 months) and asymptomatic patients were analysed. Carotid arteries were classified into culprit, non-culprit and asymptomatic. The calcium score was assessed using the Agatston method. 93 radiomic features were extracted from regions-of-interest drawn on 14 consecutive CTA slices. For DL, convolutional neural networks (CNNs) with and without transfer learning were trained directly on CTA slices. Predictive performance was assessed over 5-fold cross validated AUC scores. SHAP and GRAD-CAM algorithms were used for explainability.

Results

132 carotid arteries were analysed (41 culprit, 41 non-culprit, and 50 asymptomatic). For asymptomatic vs symptomatic arteries, radiomics attained a mean AUC of 0.96(± 0.02), followed by DL 0.86(± 0.06) and then calcium 0.79(± 0.08). For culprit vs non-culprit arteries, radiomics achieved a mean AUC of 0.75(± 0.09), followed by DL 0.67(± 0.10) and then calcium 0.60(± 0.02). For multi-class classification, the mean AUCs were 0.95(± 0.07), 0.79(± 0.05), and 0.71(± 0.07) for radiomics, DL and calcium, respectively. Explainability revealed consistent patterns in the most important radiomic features.

Conclusions

Our study highlights the potential of novel image analysis techniques in extracting quantitative information beyond calcification in the identification of CAD. Though further work is required, the transition of these novel techniques into clinical practice may eventually facilitate better stroke risk stratification.

Purpose

Sex-based differences in lumbar spine's fat content in adults are minimal, but significant variations exist in diffusion-weighted imaging (DWI) signal characteristics. This study aimed to investigate fat content’s impact on DWI performance in lumbar spine and potential sex differences.

Methods

A retrospective analysis was conducted on upper abdominal MRI examinations in asymptomatic adult. The lumbar 1 vertebral apparent diffusion coefficient (ADC) values and fat fraction were measured. Using DWI images (b = 800 s/mm²), the lumbar 1 vertebral signal was categorized into high and iso-low signal groups. A univariate and multivariate analysis was conducted to investigate the influence of fat fraction on DWI performance. Finally, the participants were divided into three groups to analyze sex differences in the effect of fat content on DWI performance.

Results

202 subjects, 99 men were included. Fat content significantly influenced lumbar spine DWI signal in both sexes (p < 0.05). The effect on ADC values was significant only in women (p < 0.001). Women demonstrated a significantly higher proportion of high DWI signal than men in the low (p = 0.002) and middle (p = 0.012) fat content groups. Additionally, women had higher ADC values in the low fat group (p = 0.004) but lower values in the high fat group (p = 0.004).

Conclusion

Fat content significantly impacts the DWI signal of lumbar spine, with a slight sex difference observed. These sex differences suggest that DWI signals may provide valuable information about the bone marrow beyond fat content.

Purpose

To compare image quality and detection accuracy of renal stones between deep learning image reconstruction (DLIR) and Adaptive Statistical Iterative Reconstruction-Veo (ASIR-V) reconstructed virtual non-contrast (VNC) images and true non-contrast (TNC) images in spectral CT Urography (CTU).

Methods

A retrospective analysis was conducted on images of 70 patients who underwent abdominal-pelvic CTU in TNC phase using non-contrast scan and contrast-enhanced corticomedullary phase (CP) and excretory phase (EP) using spectral scan. The TNC scan was reconstructed using ASIR-V70 % (TNC-AR70), contrast-enhanced scans were reconstructed using AR70, DLIR medium-level (DM), and high-level (DH) to obtain CP-VNC-AR70/DM/DH and EP-VNC-AR70/DM/DH image groups, respectively. CT value, image quality and kidney stones quantification accuracy were measured and compared among groups. The subjective evaluation was independently assessed by two senior radiologists using the 5-point Likert scale for image quality and lesion visibility.

Results

DH images were superior to AR70 and DM images in objective image quality evaluation. There was no statistical difference in the liver and spleen (both P > 0.05), or within 6HU in renal and fat in CT value between VNC and TNC images. EP-VNC-DH had the lowest image noise, highest SNR, and CNR, and VNC-AR70 images had better noise and SNR performance than TNC-AR70 images (all p < 0.05). EP-VNC-DH had the highest subjective image quality, and CP-VNC-DH performed the best in lesion visibility. In stone CT value and volume measurements, there was no statistical difference between VNC and TNC (P > 0.05).

Conclusion

The DLIR-reconstructed VNC images in CTU provide better image quality than the ASIR-V reconstructed TNC images and similar quantification accuracy for kidney stones for potential dose savings.

The study highlights that deep learning image reconstruction (DLIR)-reconstructed virtual non-contrast (VNC) images in spectral CT Urography (CTU) offer improved image quality compared to traditional true non-contrast (TNC) images, while maintaining similar accuracy in kidney stone detection, suggesting potential dose savings in clinical practice.

Purpose

The purpose of this study was to evaluate the efficacy of radiomics derived from slice-reduced CT (srCT) scans versus full-chest CT (fcCT) for diagnosing and staging of interstitial lung disease (ILD) in systemic sclerosis (SSc), considering the potential to reduce radiation exposure.

Material and methods

The fcCT corresponded to a standard high-resolution full-chest CT whereas the srCT consisted of nine axial slices. 1451 radiomic features in two dimensions from srCT and 1375 features in three dimensions from fcCT scans were extracted from 166 SSc patients. The study included first- and second-order features from original and wavelet-transformed images. We assessed the predictive performance of quantitative CT (qCT)-based logistic regression (LR) models relying on preselected features and machine learning workflows involving LR and extra-trees classifiers with data-driven feature selection. The area under the receiver operating characteristic curve (AUC) was used to estimate model performance.

Results

The best models for diagnosis and staging ILD achieved AUC=0.85±0.08 and AUC=0.82±0.08 with srCT, and AUC=0.83±0.06 and AUC=0.76±0.08 with fcCT, respectively. srCT-based models showed slightly superior performance over fcCT-based models, particularly in 2D-radiomic analyses when interpolation resolution closely matched the original in-plane resolution. For diagnosis, the LR outperformed qCT-models, whereas for staging, the best results were obtained with a qCT-based model.

Conclusions

Radiomics from srCT is an effective and preferable alternative to fcCT for diagnosing and staging SSc-ILD. This approach not only enhances predictive accuracy but also minimizes radiation exposure risks, offering a promising avenue for improved treatment decision support in SSc-ILD management.

Purpose

To analyze the role of qualitative and quantitative 3 T MR imaging assessment as a non-invasive method for the evaluation of disease severity in patients with primary sclerosing cholangitis (PSC).

Methods

A series of 26 patients, with histological diagnosis of PSC undergoing 3 T MRI and hepatological evaluation, was retrospectively enrolled. All MR examinations included diffusion-weighted imaging (DWI), T2-weighted (T2w) and T1-weighted (T1w) sequences, before and after administration of Gd-EOB-DTPA with the acquisition of both dynamic and hepato-biliary phase (HBP). Qualitative analysis was performed by assessment of liver parenchyma and biliary tract changes, also including biliary excretion of gadoxetic acid on HBP. Quantitative evaluation was conducted on liver parenchyma by measurement of apparent diffusion coefficient (ADC) and relative enhancement (RE) on 3-minute delayed phase and on HBP. Results of blood tests (ALT, ALP, GGT, total and direct bilirubin, albumin, and platelets) and transient elastography-derived liver stiffness measurements (TE-LSM) were collected and correlated with qualitative and quantitative MRI findings.

Results

Among qualitative and quantitative findings, fibrosis visual assessment and RE had the best performance in estimating disease severity, showing a statistically significant correlation with both biomarkers of cholestasis and TE-LSM. Statistical analysis also revealed a significant correlation of gadoxetic acid biliary excretion with ALT and direct bilirubin, as well as of ADC with total bilirubin.

Conclusion

Qualitative and quantitative 3 T MR evaluation is a promising non-invasive method for the assessment of disease severity in patients with PSC.

Background

Artificial intelligence (AI) has been proven useful for the assessment of tubes and lines on chest radiographs of general patients. However, validation on intensive care unit (ICU) patients remains imperative.

Methods

This retrospective case-control study evaluated the performance of deep learning (DL) models for tubes and lines classification on both an external public dataset and a local dataset comprising 303 films randomly sampled from the ICU database. The endotracheal tubes (ETTs), central venous catheters (CVCs), and nasogastric tubes (NGTs) were classified into “Normal,” “Abnormal,” or “Borderline” positions by DL models with and without rule-based modification. Their performance was evaluated using an experienced radiologist as the standard reference.

Results

The algorithm showed decreased performance on the local ICU dataset, compared to that of the external dataset, decreasing from the Area Under the Curve of Receiver (AUC) of 0.967 (95 % CI 0.965–0.973) to the AUC of 0.70 (95 % CI 0.68–0.77). Significant improvement in the ETT classification task was observed after modifications were made to the model to allow the use of the spatial relationship between line tips and reference anatomy with the improvement of the AUC, increasing from 0.71 (95 % CI 0.70 – 0.75) to 0.86 (95 % CI 0.83 – 0.94)

Conclusions

The externally trained model exhibited limited generalizability on the local ICU dataset. Therefore, evaluating the performance of externally trained AI before integrating it into critical care routine is crucial. Rule-based algorithm may be used in combination with DL to improve results.

The rapid evolution of artificial intelligence (AI) in healthcare, particularly in radiology, underscores a transformative era marked by a potential for enhanced diagnostic precision, increased patient engagement, and streamlined clinical workflows. Amongst the key developments at the heart of this transformation are Large Language Models like the Generative Pre-trained Transformer 4 (GPT-4), whose integration into radiological practices could potentially herald a significant leap by assisting in the generation and summarization of radiology reports, aiding in differential diagnoses, and recommending evidence-based treatments. This review delves into the multifaceted potential applications of Large Language Models within radiology, using GPT-4 as an example, from improving diagnostic accuracy and reporting efficiency to translating complex medical findings into patient-friendly summaries. The review acknowledges the ethical, privacy, and technical challenges inherent in deploying AI technologies, emphasizing the importance of careful oversight, validation, and adherence to regulatory standards. Through a balanced discourse on the potential and pitfalls of GPT-4 in radiology, the article aims to provide a comprehensive overview of how these models have the potential to reshape the future of radiological services, fostering improvements in patient care, educational methodologies, and clinical research.

Background

Human epidermal growth factor receptor 2 (HER2) is a tumor biomarker with significant prognostic and therapeutic implications for invasive ductal breast carcinoma (IDC).

Objective

This study aimed to explore the effectiveness of a multisequence magnetic resonance imaging (MRI)-based machine learning radiomics model in classifying the expression status of HER2, including HER2-positive, HER2-low, and HER2 completely negative (HER2-zero), among patients with IDC.

Methods

A total of 402 female patients with IDC confirmed through surgical pathology were enrolled and subsequently divided into a training group (n = 250, center I) and a validation group (n = 152, center II). Radiomics features were extracted from the preoperative MRI. A simulated annealing algorithm was used for key feature selection. Two classification tasks were performed: task 1, the classification of HER2-positive vs. HER2-negative (HER2-low and HER2-zero), and task 2, the classification of HER2-low vs. HER2-zero. Logistic regression, random forest (RF), and support vector machine were conducted to establish radiomics models. The performance of the models was evaluated using the area under the curve (AUC) of the operating characteristics (ROC).

Results

In total, 4506 radiomics features were extracted from multisequence MRI. A radiomics model for prediction of expression state of HER2 was successfully developed. Among the three classification algorithms, RF achieved the highest performance in classifying HER2-positive from HER2-negative and HER2-low from HER2-zero, with AUC values of 0.777 and 0.731, respectively.

Conclusions

Machine learning-based MRI radiomics may aid in the non-invasive prediction of the different expression status of HER2 in IDC.

Introduction

Peroneus brevis split rupture poses a diagnostic challenge, often requiring magnetic resonance imaging (MRI), yet splits are missed in initial radiological reports. However, the frequency of reported peroneus brevis split rupture in clinical MRI examinations is unknown.

Aim

This study aimed to investigate underreporting frequency of peroneus brevis split rupture in patients with lateral ankle pain.

Methods

We re-evaluated 143 consecutive MRI examinations of the ankle joint, conducted in 2021 in our region, for patients experiencing ankle pain persisting for more than 8 months. Two musculoskeletal radiologists, with 12 and 8 years of experience respectively, assessed the presence of peroneus brevis split rupture. Patients with recent ankle trauma, fractures, postoperative changes, or MRI artifacts were excluded. The radiologists evaluated each MRI for incomplete or complete peroneus brevis split rupture. The consensus between the raters was used as the reference standard. Additionally, raters reviewed the original clinical radiological reports to determine if the presence of peroneus brevis split rupture was noted. Agreement between raters' assessments, consensus, and initial reports was evaluated using Gwet’s AC1 coefficients.

Results

Initial radiological reports indicated 23 cases (52.3 %) of peroneus brevis split rupture, meaning 21 cases (47.7 %) were underreported. The Gwet’s AC1 coefficients showed that the agreement between raters and initial reports was 0.401 (standard error 0.070), 95 % CI (0.261, 0.541), p<.001, while the agreement between raters in the study was 0.716 (standard error 0.082), 95 % CI (0.551, 0.881), p<.001.

Conclusion

Peroneus brevis split rupture is underreported on MRI scans of patients with lateral ankle pain.