Current problems in diagnostic radiology最新文献

Artificial intelligence has emerged as a transformative technology in lung cancer diagnosis, with multiple large-scale meta-analyses demonstrating its significant clinical potential. Based on extensive research encompassing over 7 systematic reviews and meta-analyses from 2023 to 2025, AI models consistently achieve high diagnostic accuracy across various imaging modalities and clinical applications.

This study aimed to examine the experience of patients who often attempt to understand the results of a radiologic exam without physician assistance. This experience is a common scenario that occurs when patients access radiology reports through an electronic medical record (EMR). We conducted twenty qualitative interviews with participants aged 20 to 79. Each participant was provided with an electronic copy of a coronary calcium CT scan report and asked to read it, as if it were their own. They were then asked to reflect on that experience as if they were the patient receiving the report. Additionally, they were asked to provide suggestions for improving the reporting process. The notable sample report findings included a coronary calcium score of zero with an incidental finding of pulmonary nodule. Interview transcripts were independently coded by two MD-MPH students. Directed content analysis was used to identify themes in participant responses. When presented with the sample report, eighty-five percent of the participants shared that if it were their report, they would be concerned or overwhelmed by the information. Many participants were unable to explain the findings and about half shared that if the report was theirs, they would feel "fear". Scoring systems and medical jargon were a major cause of confusion. Commonly suggested solutions included: adding a lay summary statement, including visual explanations and removing citations. The overwhelming majority of participants felt confused or distressed after receiving the sample report. Many participants felt unsure of the next steps and wanted guidance from a provider regarding follow-up. Noting that in a real-world scenario receiving emotionally charged results in this manner would cause distress. These study findings underscore the need for patient centered reporting in radiology and provide suggestions to improve report clarity.

Background: Fetal Magnetic Resonance Imaging (MRI) is a vital tool for characterizing abnormalities detected on prenatal ultrasound, but receiving results can be emotionally challenging for patients. Patients typically receive imaging results from the ordering provider, however, in our practice radiologists traditionally conducted immediate face-to-face image review sessions with patients. The aim of this project was to improve patient-centered delivery of image results while optimizing the efficiency of the fetal MRI service workflow by removing direct radiologist review of images and results with patients. After the intervention was implemented, there was a significant reduction in average time radiologists spent per patient exam by 32 % and average report turnaround time by 55 %. Patient surveys indicated that most patients preferred receiving results from their maternal-fetal medicine specialist or obstetrician rather than the radiologist. Eliminating radiologist-patient image review sessions in the fetal MRI service improved efficiency while prioritizing patient-centered care.

Objective: To review the utilisation of radiology administrative assistants in supporting radiologists with daily tasks that are not directly related to image interpretation.

Methods: Records of all tasks delegated to radiology administrative assistants in a multi-centre radiological institution between 2021 and 2022 were collected. The number of tasks, types of task performed, and duration spent on each task were reviewed.

Results: 5455 tasks were assigned to and completed by radiology administrative assistants over the course of 12 months. 50 % of the tasks pertained to conveying significant or unexpected findings to clinicians, whilst the remaining 50 % accounted for tasks not related to results-communication, such as sourcing external imaging.

Conclusion: A substantial amount of radiologist's time is spent on ancillary tasks that are not related to reporting. Employing the help of radiology administrative assistants helps mitigate this, facilitates closed-loop communication and an overall improved radiologist workflow.

Purpose: Assessing tumour-vessel contact in pancreatic adenocarcinoma on CT is challenging for trainees and time-intensive for educators. Semi-automating feedback on this task may optimize radiologist time and standardize resident education. We hypothesized that residents who reviewed expert annotations of tumour-vessel contact would outperform those without feedback on an independent test set.

Methods: We retrospectively reviewed pre-operative staging CTs from 60 patients who underwent upfront surgical resection for pancreatic adenocarcinoma. Two resident groups (control and test) independently annotated tumour contact with the superior mesenteric artery. The test group received feedback-annotations from three expert radiologists-for the first 30 cases; the control group received none. Resident performance on the remaining 30 cases was compared against both surgical pathology and expert annotations.

Results: Test group residents demonstrated higher sensitivity than control group residents (mean sensitivity = 93 % vs. 79 %), with comparable specificity and accuracy relative to surgical pathology. While both groups performed similarly relative to expert consensus, the test group showed greater consistency in sensitivity (mean variation = 29 % vs. 46 %).

Conclusion: Virtual expert feedback improved resident sensitivity in identifying tumour-vessel contact without compromising specificity or accuracy. These findings support the use of semi-automated educational tools to enhance radiology training efficiency and effectiveness.

Background: Core needle biopsy (CNB) has emerged as a less invasive alternative to open biopsy for diagnosing soft tissue sarcomas (STS). However, its accuracy in determining tumor grade and distinguishing between benign and malignant lesions remains a subject of ongoing research.

Objective: To assess the diagnostic accuracy of image-guided CNB for soft tissue sarcomas, focusing on grade determination and benign vs. malignant distinction.

Methods: This retrospective study analysed 83 patients who underwent both CNB and surgical excision for soft tissue tumors between 2020 and 2024. CNB results were compared with final histopathology findings. Diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. ROC curve analysis was performed to evaluate diagnostic performance.

Results: The overall accuracy of CNB was 95.18 % (95 % CI: 88.11 % - 98.71 %) for distinguishing between benign and malignant lesions, with a sensitivity of 96.34 % (95 % CI: 89.68 % - 99.24 %) and PPV of 98.75 % (95 % CI: 93.23 % - 99.97 %). For grade determination, CNB showed an accuracy of 86.75 % (95 % CI: 77.52 % - 93.19 %), with sensitivity of 85.92 % (95 % CI: 75.67 % - 93.03 %) and specificity of 91.67 % (95 % CI: 61.52 % - 99.79 %) for high-grade tumors. ROC curve analysis demonstrated excellent discriminatory ability with AUC of 0.982 for benign vs. malignant distinction and 0.888 for grade determination. Diagnostic accuracy varied across sarcoma subtypes, with some rare types showing perfect accuracy and more common types demonstrating moderate to good sensitivity and high specificity.

Conclusion: Image-guided CNB demonstrates high overall accuracy in diagnosing and grading soft tissue sarcomas, particularly in identifying malignant lesions. While performance varies across sarcoma subtypes, these findings support the use of CNB as a reliable diagnostic tool in the management of suspected soft tissue sarcomas.

Pulmonary embolism (PE) remains a diagnostic challenge due to its nonspecific clinical presentation and overlapping imaging features with a wide array of conditions. Computed tomography pulmonary angiography (CTPA) is the gold standard for diagnosing PE, but technical limitations and mimicking entities may lead to false-positive interpretations. This review highlights key radiologic mimics of acute and chronic PE, grouped into artifacts and slow-flow phenomena, tumoral mimics, pseudo-defects from adjacent structures, and chronic PE mimics. Particular focus is placed on common pitfalls such as motion artifacts, pulmonary artery opacification artifacts, and transient interruption of contrast. We outline distinctive imaging features and offer practical strategies to optimize acquisition protocols and interpretative accuracy, including the use of dual-source CT, ECG-gating, and attenuation-based criteria. Recognizing these mimics is essential to avoid misdiagnosis and ensure appropriate clinical management. This review aims to equip radiologists and trainees with a structured diagnostic framework to improve confidence and accuracy when interpreting CTPA in suspected PE.

The expanding clinical application of immune checkpoint inhibitors (ICIs) across oncologic therapeutics has revealed a spectrum of pulmonary toxicities, with reported incidence rates ranging from 2 % to 19 % depending on the therapeutic regimen. Computed tomography (CT) imaging is the cornerstone for diagnostic evaluation, reliably identifying characteristic radiographic patterns such as organizing pneumonia, hypersensitivity pneumonitis-like reactions, and diffuse alveolar damage. The diagnostic challenge lies in distinguishing these immune-mediated pulmonary injuries from infectious etiologies, radiation-induced lung injury, and neoplastic progression, each requiring distinct therapeutic interventions. Contemporary management protocols employ glucocorticoid therapy dosed according to toxicity severity, while investigational approaches explore targeted immunomodulators. This review synthesizes current evidence regarding radiographic manifestations, diagnostic pathways, and therapeutic algorithms, offering radiologists a structured approach to evaluating ICI-associated pulmonary complications. We highlight recent advances, including quantitative CT analysis and serum biomarkers, that promise to refine early detection and risk stratification.

The primary aim of this study is to assess how well ultrasonography and 99mTc-sestamibi scintigraphy perform as diagnostic tools for hyperparathyroidism (HPT) when compared to the clinical measurements of parathyroid hormone (PTH) levels. This evaluation is intended to help formulate the most effective preoperative plan and guide clinical decision-making. A retrospective study of 350 HPT cases in King Abdul-Aziz University Hospital over a period spanning 2012 to 2023 was carried out. Sensitivity, specificity, and receiver operator characteristic curve AUC were used to determine the diagnostic performance of ultrasonography, 99mTc-sestamibi scintigraphy, and combined imaging to the standard of biochemical PTH levels. The statistical tests involved the McNemar test and logistic regression to evaluate the predictors such as chronic kidney disease (CKD). The combined imaging demonstrated diagnostic accuracy of 0.69 compared to 0.74 and 0.64 of scintigraphy and ultrasonography respectively. The scintigraphy had a total of 161 true positives and 73 false negative results whereas ultrasonography had a total of 139 true positives and 95 false negative results. CKD was also a good determinant in HPT (odds ratio = 1.988, p = 0.026). According to the McNemar test, there was no significant difference between ultrasonography and scintigraphy (p = 0.494). The diagnostic inaccuracy of ultrasonography is lower in diagnosing HPT as compared to scintigraphy, however using combined imaging may provide more reliability in diagnosis hence it can be used in preoperative planning, especially in patients with CKD.

Background and purpose: Accurate preoperative staging is essential for guiding surgical planning and optimizing outcomes in early-stage breast cancer. Magnetic resonance imaging (MRI) is considered the gold standard but is often limited by cost and availability. This study aimed to prospectively compare the diagnostic performance of full-field digital mammography (FFDM), digital breast tomosynthesis (DBT), contrast-enhanced mammography (CEM), and MRI for tumor detection and size estimation in patients eligible for upfront surgery.

Materials and methods: This single-center, prospective study included 46 women with histologically confirmed early-stage invasive breast cancer. All patients underwent FFDM, DBT, CEM, and MRI within one week prior to surgery. Histopathology served as the reference standard. Tumor size measurements were compared using Pearson's correlation coefficients (r), with concordance defined as a size difference within ±10 mm. Detection rates, size accuracy, and ability to identify multifocal lesions were evaluated.

Results: The mean patient age was 55.4 years. FFDM identified the primary tumor in 89.1% of cases, DBT in 97.8%, and both CEM and MRI in 100%. Tumor size correlation with pathology was highest for MRI (r=0.811), followed by CEM (r=0.660), DBT (r=0.636), and FFDM (r=0.314). Concordance with pathology was 80.4% for MRI, 71.7% for CEM and DBT, and 58.7% for FFDM. Multifocal disease was detected in 15.2% of cases by MRI, 8.7% by DBT, and 6.5% by CEM.

Conclusion: CEM and DBT showed strong diagnostic performance and may serve as accessible and cost-effective alternatives to MRI for preoperative staging in early-stage breast cancer. These modalities offer valuable imaging options in settings where MRI is limited or contraindicated.