Current Medical Imaging Reviews最新文献

Background: Colorectal tumors are common, and magnetic resonance imaging (MRI) with diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), and contrast-enhanced sequences can assist in distinguishing benign from malignant lesions. However, certain atypical cases exist in which malignancies do not exhibit restricted water molecule diffusion.

Case presentation: We present the case of a 66-year-old male with moderately differentiated rectal adenocarcinoma showing heterogeneous diffusion restriction on MRI. The tumor surface demonstrated diffusion restriction (hyperintense on DWI, hypointense on ADC), whereas the base did not. Quantitative ADC analysis showed mean values of 0.82 × 10-3 mm2/s at the surface and 1.36 × 10-3 mm2/s at the base.

Conclusion: This case highlights that regions with lower cellularity in moderately differentiated adenocarcinoma may lack diffusion restriction. Correlating imaging findings with histopathological results is essential to prevent diagnostic misinterpretation.

Background: Different subtypes of renal cell carcinoma (RCC) exhibit distinct tumor vascular morphological features. However, current imaging techniques still have limitations in delineating the vascular morphology of RCC.

Objective: This study aimed to evaluate the value of high-frame-rate contrast-enhanced ultrasound (H-CEUS) in assessing the vascular morphology of RCC.

Methods: A retrospective analysis was conducted on 163 RCC patients who were classified postoperatively into clear cell RCC (ccRCC) and non-clear cell RCC (nccRCC) groups. All patients underwent conventional US, conventional CEUS (C-CEUS), and H-CEUS preoperatively. Vascular morphology during the early phase of CEUS perfusion was categorized into five types (I-V). Differences in vascular morphology were compared using the x2 test or Fisher's exact test, and inter-observer agreement was evaluated using the Kappa coefficient.

Results: A significant difference in CDFI was observed between the ccRCC and nccRCC groups (x2=11.755, P=0.0088). Type III was the predominant vascular morphology in ccRCC on both C-CEUS and H-CEUS, with proportions significantly higher on H-CEUS (62.6% vs. 45.8%; x2=6.099, P=0.014). Type IV was the most common vascular morphology in nccRCC, with no significant difference between C-CEUS (58.9%) and H-CEUS (44.6%) (x2=2.289, P=0.130).

Discussion: H-CEUS revealed significant vascular morphological differences in ccRCC ≤4 cm (x2=9.307, P=0.038), but not in nccRCC ≤4 cm or in any tumors >4 cm. Inter-observer agreement for vascular morphology evaluation was substantial for both C-CEUS (κ=0.751) and H-CEUS (κ=0.657) (both P<0.001).

Conclusion: H-CEUS shows superior visualization capabilities in evaluating vascular morphology for ccRCC lesions ≤4 cm, which provides valuable insights into its potential as a non-invasive imaging modality for differentiating RCC subtypes and tailoring treatment strategies.

Introduction: The objective of this study is to develop an advanced and interpretable diagnostic framework combining deep learning and machine learning for the initial diagnosis of pneumonia with high accuracy in pediatric patients to overcome the critical limitations of existing diagnostic procedures.

Methods: We introduce a hybrid model, XceptRF-Net, that integrates deep feature learning of the Xception convolutional neural network and the probabilistic modelling power of the Random Forest for the classification of Fisher's iris dataset. The first stage of the model considers the highlevel spatial features from the chest X-rays, which are extracted using Xception. Followed by that, they are subsequently mapped to a probabilistic feature space using Random Forest, contributing to the feature representation and the classification robustness. The discriminative capability of the engineered features was tested by different machine learning classifiers such as Logistic Regression (LR), K-Nearest Neighbours (KNN), and Multi-Layer Perceptron (MLP). Fine-tuning and k-fold cross-validation were also performed for generalization purposes and to speed up computation.

Results: The proposed XceptRF-Net framework, established from an experimental study on one dataset with 5,863 pediatric CXRs, has been objectively shown to benefit over conventional methods. Logistic Regression achieved the highest diagnostic accuracy of 98%, which is a validation of the spatial and probabilistic feature learning integration.

Discussion: The effectiveness of the XceptRF-Net model highlights the value of combining deep feature extraction with probabilistic modeling to enhance clinical decision-making.

Conclusion: The findings highlight the theoretical superiority of the integration of convolutional deep features with ensemble learning and the generation of probabilistic features for medical image analysis. The proposed method provides a stable and explainable framework for clinical decision support and has high potential for practical use in real-world systems for pediatric pneumonia screening and diagnosis.

Introduction: MR-HSG offers radiation-free multiplanar visualization for evaluating fallopian tube patency, but persistent challenges in achieving consistent high-quality visualization, particularly in smaller-diameter segments, limit clinical adoption. Previous studies modifying contrast viscosity achieved only partial improvements. This study aimed to use standardized phantom models simulating the female pelvis to systematically explore factors affecting MR-HSG image quality and provide evidence-based guidance for protocol optimization.

Method: Nine standardized phantoms were constructed using agar gel with embedded tubes of three inner diameters (4mm, 2mm, and 1.4mm) representing ampullary, isthmic, and interstitial segments. Three contrast agents with different viscosities were tested: gadolinium-saline (2.6 mPa·s), gadolinium-iopromide (7.9 mPa·s), and gadolinium-iodixanol (8.7 mPa·s) at 37°C. T1-weighted 3D-mDIXON sequences with keyhole technology were employed on a 1.5T MRI scanner. Signal intensity measurements and qualitative assessment (good/fair/poor) were performed by blinded evaluators.

Results: No significant differences in signal intensity were found between contrast agents of different viscosities (P>0.05). However, tube diameter significantly affected imaging quality (P<0.001). The 4 mm tubes showed the highest SI (5554.49±1042) with 100% good imaging, the 2 mm tubes showed intermediate SI (733.65±78.76) with 100% fair imaging, and the 1.4 mm tubes showed the lowest SI (444.55±34.70) with 100% poor imaging across all contrast agents.

Discussion: Fallopian tube inner diameter is the primary determinant of MR-HSG imaging quality, while contrast agent viscosity (within 2.6-8.7 mPa·s range) shows no significant effect under controlled conditions. This study provides foundational data for understanding physical factors affecting MRHSG quality and suggests that anatomical factors may be more critical than contrast properties for clinical protocol optimization, potentially reducing procedure costs while maintaining diagnostic quality.

Conclusion: Optimizing spatial resolution to minimize partial volume effects may be more beneficial than modifying contrast agent properties for improving visualization of narrow fallopian tube segments. Clinical validation studies are warranted to confirm these findings in the complex in vivo environment.

Introduction: Breast cancer is a disease in which abnormal breast cells grow uncontrollably and develop into a tumor. It is one of the most common cancers that affects women around the world. The most often used imaging tool for identifying breast cancer is mammography. Early and accurate identification of tumors can be crucial for effective treatment and planning, which reduces mortality rates. This work proposes a novel hybrid deep learning-based automated framework for early and accurate breast cancer detection using mammographic images.

Methods: The methodology integrates multiple components into a hybrid model. Initially, mammographic images from the Curated Breast Imaging Subset of the Digital Database for Screening Mammography (CBIS-DDSM) undergo a preprocessing step, uses a guided filter to remove noise and enhance the visibility of regions of interest (ROI). Modified Dingo Optimization (MDO) algorithm is used to segment the tumour-affected regions, not only to identify abnormalities localized in a single region of the breast but also to effectively detect multiple abnormal areas distributed across different tissue regions. Deep features are then extracted using a pretrained U-Net architecture. The Search and Rescue Optimization (SRO) algorithm was utilized for feature optimization to select the most relevant deep features, reducing dimensionality and enhancing the model's diagnostic accuracy. A Dual Stage Spiking Convolutional Neural Network (DSS-CNN) is implemented for classification and enhancing the model's ability.

Results: The proposed hybrid deep learning model achieves outstanding performance, with an accuracy of 98.598%, precision of 97.343%, recall of 97.514%, and an F-measure of 96.89%. Comparative analysis confirms that the approach significantly reduces false positive and false negative rates, outperforming existing state-of-the-art techniques.

Conclusion: The proposed robust end-to-end system for early and accurate breast cancer detection demonstrates the efficacy of the framework in improving diagnostic accuracy, precision, recall, and F-measure, offering valuable support in clinical decision-making.

Rectal cancer (RC), one of the most common malignant tumors, has a high incidence rate and mortality rate worldwide. Radiomics turns medical images into high-dimensional mineable data through high-throughput extraction algorithms, where the methods include filter-based algorithms and texture analysis. All these features are then combined with machine learning or deep learning algorithms to provide objective evidence to facilitate accurate diagnosis, radiation staging, radiotherapy planning, or prognosis prediction. Multi-parametric magnetic resonance imaging has been considered as one of the best modalities for performing radiomics analysis on rectal cancer because it can capture most features about tumor heterogeneity and micro-environment information. In the past few years, magnetic resonance imaging (MRI)-based radiomics has shown great promise in a variety of fields, including tumor-node-metastasis staging, monitoring pathological high-risk factors, predicting genetic markers, neoadjuvant therapy response evaluation, and prognostic survival analysis in rectal cancer. In this paper, we provide an overview of the current state-of-the-art on MRI radiomics for rectal cancer and present a comparison between the available methods of feature extraction, and provide a critical discussion of current issues and possible developments that might be pursued in future research on this topic.

Introduction: Dedifferentiated liposarcoma is a rare type of mesenchymal sarcoma. Although most cases occur in the retroperitoneum or extremities, this report presents a rare case of dedifferentiated liposarcoma in the epididymis, highlighting the diverse sites where this disease may manifest and emphasizing the diagnostic challenges it poses, as well as the importance of comprehensive imaging and histopathological assessment.

Case presentation: A 70-year-old Chinese male patient presented to the urology department with progressive left scrotal enlargement, pain, and a palpable firm mass over the past month. After admission, initial ultrasound examination indicated a left spermatic cord mass. Subsequently, an enhanced MRI was performed, revealing no obvious fat signal within the lesion but significant enhancement, with thickening and compression of the spermatic cord. The findings ultimately suggested a malignant sarcoma of the left epididymis. Surgical resection and subsequent histopathological examination confirmed a spindle cell-predominant dedifferentiated liposarcoma of the epididymis encircling the spermatic cord. After 11 months of follow-up, no recurrence has been detected.

Conclusion: We report a surgically confirmed rare case of dedifferentiated liposarcoma originating from the epididymis, characterized by predominant spindle cells and significant imaging enhancement. This atypical presentation complicated differentiation from other spindle cell sarcomas, highlighting diagnostic challenges at rare sites.

With the increasing popularity and clinical adoption of deep learning in CT image reconstruction, two distinct approaches have emerged under the concept of 'deep' reconstruction: Deep Learning Reconstruction (DLR) and Deep Iterative Reconstruction (DIR). Despite falling under the umbrella of 'deep' reconstruction, DLR and DIR differ in technical principle, clinical applicability, and reconstruction performance. This review aims to provide a clinically oriented overview of these two methods, emphasizing their coexistence and differentiated roles on a 320-row CT scanner platform, offering radiologists insights for clinical practice as well as inspirations for future research. On this platform, DLR and DIR represent complementary strategies in clinical practice, where DLR is implemented as a cardiac-specific algorithm and DIR for other bodyparts. By summarizing representative clinical applications, we highlight the advantages of DLR in cardiac CT and strengths of DIR across chest, abdominal, vascular, and perfusion CT imaging. Quantitative evidence from recent studies demonstrates consistent improvements of both DIR and cardiac-specific DLR over routine Hybrid Iterative Reconstruction (HIR). Their complementary characteristics also suggest potential benefits when applied in multi-region CT imaging. In addition, the clinically valuable image features of DIR that merit further investigation, as well as other technical considerations relevant to 'deep' reconstructions are discussed.

Cerebral toxoplasmosis is one of the most common opportunistic infections among AIDS patients. Clinical and neuroimaging manifestations are diverse and non-specific, resulting in frequent delayed diagnosis and even misdiagnosis, leading to neurological impairment, coma, and death. In addition to clinical and serological examinations, multimodal neuroimaging is indispensable for early diagnosis and subsequent treatment evaluation. Indeed, functional magnetic resonance imaging technologies and positron emission tomography provide complementary information for early diagnosis and treatment, which can improve prognosis when combined with prevention strategies. Recent advances in vaccine development have provided new hope for the prevention of cerebral toxoplasmosis. This article reviews multimodal imaging evaluation strategies and other recent clinical advances for the prevention, diagnosis, and treatment of AIDS-related cerebral toxoplasmosis.