Current Medical Imaging - Volume 20, Issue 1, 2024

There are variations in prognosis and therapeutic approach for renal cell carcinoma among different histological subtypes. This study aims to determine the relationship between radiologically detected peritumoral neovascularization and the histological subtypes of Renal Cell Carcinoma (RCC) and to assess whether extratumoral neovascularization characteristics detected via imaging can contribute to distinguishing these subtypes alongside tumor size and T-stage.

104 renal tumors from 104 cases consisting of 31 females (29.8%) and 73 males (70.2%) who underwent abdominal CT or MRI and received a histopathological renal cell carcinoma diagnosis were included. Out of 104 cases, 45 (43.27%) cases had a preoperative CT, 52 (50%) cases had a preoperative MRI, and 7 (6.73%) cases had both preoperative CT and MR images. The cases were categorized according to the histopathologic subtypes. The presence of the radiologically visible peritumoral vascularity and its diameter was noted in order to compare with the histopathological subtypes and other morphologic or histopathological findings, including size, presence of cystic component, T score, and Fuhrman grade of the tumor.

104 unilateral renal tumors (median size 5 cm; range 2-26 cm) were included in this study, of which 71 (68.3%) were clear cell, 20 (9.2%) were papillary and 13 (12.5%) were chromophobe renal cell carcinomas. Although the presence of peritumoral neovascularization was observed to a lesser degree in papillary carcinomas than clear cell and chromophobe carcinomas, there was no statistically significant difference among histological subtypes and between clear cell and non-clear cell carcinomas according to the frequency of peritumoral neovascularization (p = 0.16 and p = 0.084). The presence of peritumoral neovascularization was significantly associated with tumor size for all tumors and within histological subtypes (p < 0.0001). As the diameter of the tumor increased, the presence of peritumoral neovascularization increased. T stage of tumors was significantly associated with both the presence of peritumoral neovascularization and the largest peritumoral vessel diameter (p < 0.01 and p = 0.002).

No statistically significant association between the histological subtype of tumors and the frequency of peritumoral neovascularization was found in this study. The frequency of peritumoral neovascularization increased with the size and T stage of the tumor. Additionally, the largest peritumoral vessel diameter increased with the T stage of the tumor. There was no statistically significant relationship between peritumoral vascularity and Fuhrman grade.

This study aimed to develop an automated method for segmenting spleen computed tomography (CT) images using a deep learning model to address the limitations of manual segmentation, which is known to be susceptible to inter-observer variability. Subsequently, a prediction model for gastric cancer (GC) differentiation was constructed alongside radiomics, and a nomogram was generated to investigate its clinical guiding significance.

This study enrolled 262 patients with pathologically confirmed GC. We employed a deep learning model, U-Mamba, to achieve fully automated segmentation of the spleen CT images. Subsequently, radiomic features were extracted from the entire spleen CT images, and significant features were identified through dimensionality reduction. The clinical and radiomic features were permuted and combined to create three predictive models: the CL model, the RA model, and the CR model. Finally, the model with superior performance was represented as a nomogram.

A total of 30 radiomic features and 1 clinical feature were considered valuable through dimensionality reduction and selection. The RA model demonstrated greater discriminative power than both the CR model and the CL model. A nomogram based on the logistic clinical model was created to facilitate the application and validation of the clinical model.

The radiomic features obtained through the automated segmentation of the spleen using deep learning demonstrate efficacy in predicting the degree of differentiation in GC. These features offer valuable guidance for clinical decision-making in the form of a nomogram.

Castleman disease (CD) is a rare lymphoproliferative disorder, with intracranial involvement being exceedingly rare. Unicentric Castleman disease (UCD) is typically benign and localized, but its presentation can mimic other intracranial pathologies, complicating diagnosis.

We reported a 52-year-old woman who presented with progressive headaches and language disturbances. Imaging, including MRI and CT, revealed an extra-axial left frontotemporal lesion initially diagnosed as an en plaque meningioma. Surgical resection of the lesion was performed. Histopathological examination revealed UCD with plasma cell predominance, characterized by lymphoid hyperplasia and concentric germinal centers. Immunohistochemical staining confirmed the diagnosis, with positive markers including CD20, CD3, and CD16.

Intracranial UCD is a rare and challenging differential diagnosis for extra-axial lesions, often resembling meningiomas. Accurate diagnosis requires a combination of imaging and histopathology, with immunohistochemistry playing a crucial role. Complete surgical resection is the optimal treatment for localized UCD.

This study aimed to establish a multimodal deep-learning network model to enhance the diagnosis of benign and malignant pulmonary ground glass nodules (GGNs).

Retrospective data on pulmonary GGNs were collected from multiple centers across China, including North, Northeast, Northwest, South, and Southwest China. The data were divided into a training set and a validation set in an 8:2 ratio. In addition, a GGN dataset was also obtained from our hospital database and used as the test set. All patients underwent chest computed tomography (CT), and the final diagnosis of the nodules was based on postoperative pathological reports. The Residual Network (ResNet) was used to extract imaging data, the Word2Vec method for semantic information extraction, and the Self Attention method for combining imaging features and patient data to construct a multimodal classification model. Then, the diagnostic efficiency of the proposed multimodal model was compared with that of existing ResNet and VGG models and radiologists.

The multicenter dataset comprised 1020 GGNs, including 265 benign and 755 malignant nodules, and the test dataset comprised 204 GGNs, with 67 benign and 137 malignant nodules. In the validation set, the proposed multimodal model achieved an accuracy of 90.2%, a sensitivity of 96.6%, and a specificity of 75.0%, which surpassed that of the VGG (73.1%, 76.7%, and 66.5%) and ResNet (78.0%, 83.3%, and 65.8%) models in diagnosing benign and malignant nodules. In the test set, the multimodal model accurately diagnosed 125 (91.18%) malignant nodules, outperforming radiologists (80.37% accuracy). Moreover, the multimodal model correctly identified 54 (accuracy, 80.70%) benign nodules, compared to radiologists' accuracy of 85.47%. The consistency test comparing radiologists' diagnostic results with the multimodal model's results in relation to postoperative pathology showed strong agreement, with the multimodal model demonstrating closer alignment with gold standard pathological findings (Kappa=0.720, P<0.01).

The multimodal deep learning network model exhibited promising diagnostic effectiveness in distinguishing benign and malignant GGNs and, therefore, holds potential as a reference tool to assist radiologists in improving the diagnostic accuracy of GGNs, potentially enhancing their work efficiency in clinical settings.

To investigate factors influencing the effectiveness and safety of super-selective embolization in patients with high-grade gross hematuria.

This retrospective, single-center study included 19 consecutive cancer patients (12 men and 7 women, mean age of 72.3 years) who had undergone TAE for intractable hematuria between January 2008 and February 2024. Factors such as technical and clinical success rates, embolized vessels, embolic agents used, and complications were evaluated. This study specifically focused on patients with severe hematuria (grade 3 or above) and examined the effects of super-selective embolization, hematuria grade, and embolic agents on patient outcomes.

Technical success was achieved in all 23 angiography procedures performed, with a clinical success rate of 56.5%. Clinical success was significantly correlated with hematuria grade, super-selectivity of the procedure, and type of embolic agent used. Multivariate logistic regression analysis revealed that the embolic material, specifically tris-acryl gelatin microspheres (TAGM), was an independent factor that significantly affected clinical outcomes. No major complications were reported.

TAGM for supers-elective embolization in patients with massive gross hematuria is both effective and safe. However, the effectiveness of TAE may decrease in patients with severe hematuria, highlighting the need for combination therapies.

This paper presents a multichannel deep-learning method for detecting lung diseases using chest X-ray images. Using EfficientNetB0 through EfficientNetB7 pretrained models, the methodology offers improved performance in classifying COVID-19, viral pneumonia, and normal chest X-rays.

The EfficientNetB2 model was customized by incorporating Squeeze-and-Excitation (SE) blocks and the Convolutional Block Attention Module (CBAM) to improve the model's attention mechanisms. Additional convolutional layers were added for improved feature extraction, and multi-scale feature fusion was implemented to capture features at different scales.

In this study, 99.3% of the unseen chest X-ray images were identified using the proposed model. It demonstrated superior performance, surpassing existing techniques and highlighting its robustness and generalizability on unseen data samples.

Moreover, visualization techniques were used to inspect the intermediate layers of the model, providing deeper insights into its processing and interpretation of medical images. The proposed method offers healthcare radiologists a valuable tool for rapid and accurate point of care diagnoses.

Budd-Chiari syndrome is a rare entity that is caused by an obstruction of the flow in the hepatic veins or inferior vena cava.

Herein, we report a rare case of iatrogenic Budd-Chiari syndrome. A 52-year-old woman with chronic renal failure under hemodialysis, presented to our hospital for dyspnea caused by a large pleural effusion. After the placement of the central right jugular vein catheter, she suffered from right upper quadrant acute abdominal pain along with elevation of liver function enzymes in blood tests. An abdominal computed tomography with contrast revealed obstruction of the right hepatic vein by the catheter tip with concomitant thrombosis, thus the diagnosis of Budd-Chiari syndrome was confirmed. Removal of the catheter and anticoagulant therapy were successfully utilized to treat the patient.

Knowledge of the full spectrum of adverse effects of such a procedure is crucial for their early identification and treatment, often with a multidisciplinary approach.

This study aims to optimize the downgrading of BI-RADS class 4a nodules by combining various sectional elastography techniques with age.

We performed conventional ultrasonography, strain elastography (SE), and shear wave elastography (SWE) on patients. Quantitative parameters recorded included age, cross-sectional and longitudinal area ratios (C-EI/B, L-EI/B), strain rate ratios (C-SR, L-SR), overall average elastic modulus values (C-Emean1, L-Emean1), five-point average elastic modulus values (C-Emean2, L-Emean2), and maximum elastic modulus values (C-Emax, L-Emax).

Histopathological evaluations showed that out of 230 lesions, 45 were malignant, and 185 were benign. The sensitivity and specificity of conventional ultrasonography were 100% and 0%, respectively. In contrast, SE and SWE exhibited higher specificity but lower sensitivity. Cross-sectional parameters (C-EI/B, C-SR, C-Emean1, C-Emean2, and C-Emax) outperformed their longitudinal counterparts, with C-SR and C-Emax showing the highest specificity (72.43% and 73.51%) and satisfactory sensitivity (80.00% and 88.89%). Combining age with C-SR and C-Emax significantly improved diagnostic efficiency, achieving a sensitivity of 97.78% and a specificity of 77.30%.

Integrating age with C-SR and C-Emax effectively reduces unnecessary biopsies for most BI-RADS 4a benign lesions while maintaining a very low misdiagnosis rate.

It remains unknown whether the parameters obtained using the Stretched Exponential Model (SEM) and Fractional Order Calculus (FROC) models can help distinguish Hepatocellular Carcinoma (HCC) from Intrahepatic Cholangiocarcinoma (ICC).

This study aimed to evaluate the application value of the parameters of the 3.0T Magnetic Resonance Imaging (MRI) high-order SEM and FROC diffusion model in differentiating HCC and ICC.

Patients with pathologically confirmed HCC and ICC were prospectively enrolled. Diffusion-weighted imaging scans with multiple b-values were acquired 2 weeks before the surgery. The original MRI images were fitted using the mono-exponential model, SEM, and FROC, and several parameters were obtained for the analysis.

In total, 74 patients with HCC and 21 with ICC were included in the study. Significant differences between the HCC and ICC groups were noted in the Apparent Diffusion Coefficient (ADC: p = 0.007), Distributed Diffusion Coefficient (DDC: p < 0.001), and Diffusion coefficient (D: p < 0.001), as each value was significantly lower in the HCC than in the ICC group. The area under the receiver operating characteristic curve of ADC, DDC, and D was 0.694, 0.812, and 0.825, respectively, and the most effective corresponding cut-off values were 1.135 μm²/ms, 1.477 μm²/ms, and 1.104 μm²/ms, respectively.

The diffusion parameters DDC from the SEM and D from the FROC model have been found to be more effective in discriminating HCC and ICC than the ADC from the mono-exponential model. Combining these quantitative parameters can improve the MRI’s diagnostic accuracy, providing useful information for the preoperative differential diagnosis between HCC and ICC.

Pulmonary sclerosing haemangioma (PSH) is a relatively uncommon benign neoplasm that is often asymptomatic and predominantly affects young and middle-aged females. PSH often appears as a single nodule, whereas multiple lesions with a cavity are relatively rare and easily misdiagnosed.

In our study, we report a patient with separated nodules in the same lobe with a cavity and clinical manifestations of cough and sputum with a radiographic presentation similar to that of tuberculosis. The patient underwent percutaneous lung biopsy and thoracoscopic partial pneumonectomy and was diagnosed with multiple PSHs.

We report a rare case of multiple PSHs that were treated with a thoracoscopic partial resection of the left upper lobe. Postoperative pathology confirmed multiple PSHs. Due to the rarity of PSH, it is easily misdiagnosed in clinical practice as lung cancer, tuberculosis, or other diseases. The final diagnosis depends on the pathology, and surgery is considered to be an appropriate treatment that leads to a good prognosis.

Bruxism may cause increased alveolar bone thickness and density and irregular enlargement of the periodontal space.

This study aimed to evaluate the mandibular bone quality using radio-morphometric indices and Fractal Dimension (FD) analysis in orthopantomography (OPG).

OPGs of 100 patients, 50 bruxers and 50 non-bruxers, were included in this study. Values, such as mental index (MI), panoramic mandibular index (PMI), gonial index (GI), antegonial notch depth (AND), mandibular cortical index (MCI), and antegonial index (AI), were calculated in OPG. Eight bilateral areas of interest (ROI) were selected on ort for FD analysis: ROI 1, mandibular condyle; ROI 2, mandibular ramus; ROI 3, mandibular angulus; and ROI 4, mandibular mental area.

MI, PMI, and AND values were higher in bruxers than in the control group (p<0.05). MCI and AI values calculated on both sides were not statistically significantly related in bruxism and control group individuals (p>0.05). As a result of the calculations, the FD values of the left condyle (p=0.02) and left angulus (p=0.03) areas showed a statistically significant difference between individuals with and without bruxer. No statistically significant difference was found in the FD measurements calculated from the ramus and mental areas on the right and left sides (p>0.05). The relationship between FD values and gender in these areas was examined, and no statistically significant difference was found (p<0.05).

In dentistry, bruxism can be diagnosed and treated by measuring MI, PMI, and AND values. No difference was found in mandibular cortical bone thickness in bruxers and non-bruxers, according to AI and MCI. The mean GI measured on the right side differed between groups. FD values of the mandibular trabecular bone were affected by bruxism in the right condyle and right angulus areas.

Currently, most multimodal medical image fusion techniques focus solely on integrating the edge details of image features, often overlooking color preservation from the source images. Hence, this paper proposes a multi-channel fusion algorithm based on gradient domain-guided image filtering.

This study aims to enhance the color preservation of source images in multimodal medical image fusion algorithms.

Utilizing gradient field-guided image filters for image smoothing, the process involves constructing different image layers, decomposing using wavelet transforms, and downsampling. Various fusion rules are then applied before inverse wavelet transformation.

Regarding MSE, CCI, PSNR, SSIM, DD, SM, and other metrics, the proposed algorithm consistently ranks highest compared to alternative methods.

Through both subjective and objective analyses, experimental results substantiate the significant edge-preserving effects of the proposed fusion algorithm while effectively maintaining image fidelity and spectral integrity.

Perivascular Epithelioid Cell tumor (PEComa) is a rare mesenchymal neoplasm characterized by the co-expression of melanocytic and myoid markers. While PEComas can arise in diverse anatomical sites, gastric PEComas are exceedingly rare, with merely nine cases documented in the extant literature.

Herein, we have presented a case of gastric PEComa in a 65-year-old male patient who exhibited a 3-year history of epigastric pain, with notable exacerbation in the two months prior to diagnosis. For the initial evaluation of the patient's condition, Endoscopic Ultrasound-guided Fine Needle Aspiration (EUS-FNA) and Computed Tomography (CT) were employed, which enabled a preoperative diagnosis. Radiological assessment demonstrated a neoplasm exhibiting heterogeneous arterial enhancement, persistent delayed enhancement, and distinct margins. Subsequent to diagnosis, the patient underwent surgical resection and has maintained a disease-free status for one year postoperatively. This case report highlights the crucial role of EUS-FNA in facilitating preoperative histological diagnosis and optimizing surgical planning for gastric PEComa.

This case constitutes the tenth documented instance of gastric PEComa in the global literature. In this case, EUS-FNA facilitated a preoperative histopathological diagnosis, thereby enabling precise surgical planning. An accurate preoperative diagnosis is crucial for devising an optimal treatment strategy.

Various studies have demonstrated large variations in the annual occupational exposure of medical personnel working in interventional cardiology departments, ranging from 0.1 mSv to exceeding the annual effective occupational dose limit of 20 mSv.

The purpose of this study was to investigate the 17-year dynamics in the personal dosimetry records of the medical staff in one interventional cardiology department in Bulgaria.

The study was performed between 2007 and 2023 and included 31 interventional cardiologists. For each of them, data from all individual dosimetry control reports were analysed. The number and complexity of interventional procedures were analysed on an annual basis. A total number of 39639 procedures performed over 17 consecutive years were classified and analysed.

The results have suggested that when a newly formed team gains clinical experience, the focus shifts towards optimizing radiation exposure to patients, and it has been observed to change from 40 Gy.cm² in 2009 to 14.8 Gy.cm² in 2023 for diagnostic and from 146 Gy.cm² in 2009 to 51.2 Gy.cm² in 2023 for interventional procedures, and from 19.5 mSv/year under the lead apron in 2012 and 3.7 mSv/year in 2023 for one of the interventional cardiologists among the medical staff. The optimization process in the department has been found to be slow but consistent, starting with the routine application of basic methods to reduce the likelihood of skin injury. Any practical implementation of a methodology or process requires periodic training to raise awareness of the topic and the use of different strategies to put it into practice. Most of the reported values from individual dosimetry monitoring have been found to be in the range below 4 mSv/year, consistent with the summarised results from other studies.

The radiation protection awareness program introduced in 2014 has been found to result in between a 2- and 6-fold reduction in individual effective doses for some staff members and a 2-fold reduction in typical patient doses.

Controversies still exist regarding the mechanism formation of the blend sign, defined as hypodensity and hyperdensity regions, in Intracerebral Hemorrhage (ICH), and which region associated with bleeding remains unknown. Spot sign is an independent predictor of hematoma expansion (HE) and indicates persistent bleeding focus in the hematoma. Here, we sought to establish the relationship between the spot sign and the blend sign to gain insights into the formation of the blend sign.

Patients were categorized based on the spot sign location within the blend sign in patients with ICH from 2018 to 2023. subjects with a spot sign in the hypodensity region of the blend sign (hypo-spot sign group); subjects with a spot sign in the hyperdensity region of the blend sign (hyper-spot sign group). Subsequently, patients were stratified into two groups based on the presence or absence of HE. Also, we analyzed the relationship between the spot sign and the blend sign, as well as the association between the blend sign and HE.

A total of 205 patients were included, including 190 patients (92.7%) who had the spot sign in the hyper-spot sign and 15 patients (7.3%) who had the spot sign in the hypo-spot sign. HE was observed in 60 patients (29.3%), 59 (98.3%) of whom had the spot sign detected in the hyper-spot sign, while only one (1.7%) had the spot sign in the hypo-spot sign. Univariate logistic regression analysis revealed that the hyper-spot sign group (6.305, 1.810–49.072; P < 0.05) was an independent predictor of HE.

The hyperdensity area of the blend sign may represent fresh bleeding in ICH rather than the hypodensity area.

ClinicalTrials.gov, NCT05548530. Registered on September 21, 2022, Prognostic Analysis of Different Treatment Options for Cerebral Hemorrhage-Full Text View - ClinicalTrials.gov “retrospectively registered.”