Diagnostic and interventional radiology最新文献

{"title":"Efficacy and safety of percutaneous thermal ablation in Bosniak III and IV cystic renal masses: a systematic review and meta-analysis.","authors":"Hans-Jonas Meyer, Timo Christian Meine, Manuel Florian Struck, Silke Zimmermann","doi":"10.4274/dir.2025.253263","DOIUrl":"https://doi.org/10.4274/dir.2025.253263","url":null,"abstract":"<p><strong>Purpose: </strong>Local thermal ablation is considered a standard treatment for small kidney masses. However, few studies have investigated the efficacy and safety of thermal ablation for cystic kidney masses.</p><p><strong>Methods: </strong>The MEDLINE library, Cochrane, and SCOPUS databases were screened for studies investigating the efficacy of thermal ablation for cystic renal masses, comprising studies between 1995 and February 2024. In total, seven studies were deemed suitable and included in the present analysis.</p><p><strong>Results: </strong>The studies included a total of 113 participants with 134 cystic renal masses. The sample sizes ranged from 5 to 38 participants. There were 76 men (67.2%) and 37 women (32.8%), with a mean age of 64.7 years (range: 50 to 75.4 years). Overall, 55 cystic masses were classified as Bosniak III (41%) and 79 as Bosniak IV (59%). Technical success of local thermal ablation was reported in 133 cystic masses (99.2%). The pooled meta-analytic technical success rate was 100% [95% confidence interval (CI): 96%-100%, I² = 0.0%]. Complications were reported in 9 cases (6.7%). According to the Society of Interventional Radiology classification system, there were 3 major complications (2.6%) and 6 minor complications (5.3%). The pooled meta-analytic complication rate was 10% (95% CI: 5%-20%, I² = 40%). No tumor recurrence was reported during follow-up.</p><p><strong>Conclusion: </strong>Local thermal ablation can be considered a highly effective and safe procedure for cystic kidney masses. Most studies were performed using radiofrequency ablation, underscoring the need for further studies on alternative ablation techniques such as microwave ablation and cryoablation.</p><p><strong>Clinical significance: </strong>Local thermal ablation is an effective and safe procedure for treating cystic kidney masses.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.7,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144871931","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reply: letter to editor: dual-energy computed tomography-based volumetric thyroid iodine quantification: correlation with thyroid hormonal status, pathologic diagnosis, and phantom validation.","authors":"Younghen Lee","doi":"10.4274/dir.2025.253518","DOIUrl":"https://doi.org/10.4274/dir.2025.253518","url":null,"abstract":"","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.7,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144871935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Emergency radiology in displaced populations: imaging and practical challenges.","authors":"Vefa Çakmak, Pınar Çakmak, Duygu Herek, Sonay Aydın","doi":"10.4274/dir.2025.253297","DOIUrl":"https://doi.org/10.4274/dir.2025.253297","url":null,"abstract":"<p><p>Migrants and refugees tend to use emergency departments as their primary source of care, leading to a substantial increase in emergency radiological imaging. Migrants and refugees have risk factors, such as low vaccination rates, poor hygiene, malnutrition, and inadequate self-care, which increase their risk of contracting infectious diseases. For example, the prevalence of tuberculosis among refugee populations in host countries is increasing, and imaging findings related to tuberculosis are frequently observed by radiologists. Strengthening screening programs in host countries for tuberculosis and other infectious diseases among migrant populations can help mitigate the risk of transmission within migrant communities. Another condition, cystic echinococcosis, is more common among refugees and migrants from the Middle East and Afghanistan. For radiologists working in host countries, echinococcosis involving the liver and lungs should be considered in the differential diagnosis. Both intentional (e.g., violence, assault) and unintentional (e.g., workplace injuries, accidents) traumas are frequently encountered in emergency radiology, particularly among refugees and immigrants. Workplace injuries are four times more common among migrants and refugees than among the local population due to their work in high-risk industries, such as construction and heavy industry, and emergency radiology frequently encounters radiological findings of organ injuries due to falls from height. In addition, healthcare professionals in emergency radiology face various challenges when dealing with migrant and refugee patients, such as communication barriers, social security problems, and psychological distress.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.7,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144871932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparative analysis of tumor and mesorectum radiomics in predicting neoadjuvant chemoradiotherapy response in locally advanced rectal cancer.","authors":"Ali Cantürk, Raif Can Yarol, Ali Samet Tasak, Hakan Gülmez, Kenan Kadirli, Tayfun Bişgin, Berke Manoğlu, Selman Sökmen, İlhan Öztop, İlknur Görken Bilkay, Özgül Sağol, Sülen Sarıoğlu, Funda Barlık","doi":"10.4274/dir.2025.253270","DOIUrl":"https://doi.org/10.4274/dir.2025.253270","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;Neoadjuvant chemoradiotherapy (CRT) is known to increase sphincter preservation rates and decrease the risk of postoperative recurrence in patients with locally advanced rectal tumors. However, the response to CRT in patients with locally advanced rectal cancer (LARC) varies significantly. The objective of this study was to compare the performance of models based on radiomics features of the tumor alone, the mesorectum alone, and a combination of both in predicting tumor response to neoadjuvant CRT in LARC.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;This retrospective study included 101 patients with LARC. Patients were categorized as responders (modified Ryan score 0-1) and non-responders (modified Ryan score 2-3). Pre-CRT magnetic resonance imaging evaluations included tumor-T2 weighted imaging (T2WI), tumor-diffusion weighted imaging (DWI), tumor-apparent diffusion coefficient (ADC) maps, and mesorectum-T2WI. The first radiologist segmented the tumor and mesorectum from T2-weighted images, and the second radiologist performed tumor segmentation using DWI and ADC maps. Feature reproducibility was assessed by calculating the intraclass correlation coefficient (ICC) using a two-way mixed-effects model with absolute agreement for single measurements [ICC(3,1)]. Radiomic features with ICC values &lt;0.60 were excluded from further analysis. Subsequently, the least absolute shrinkage and selection operator method was applied to select the most relevant radiomic features. The top five features with the highest coefficients were selected for model training. To address class imbalance between groups, the synthetic minority over-sampling technique was applied exclusively to the training folds during cross-validation. Thereafter, classification learner models were developed using 10-fold cross-validation to achieve the highest performance. The performance metrics of the final models, including accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC), were calculated to evaluate the classification performance.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Among the 101 patients, 36 were classified as responders and 65 as non-responders. A total of 25 radiomic features from the tumor and 20 from the mesorectum were found to be statistically significant (&lt;i&gt;P&lt;/i&gt; &lt; 0.05). The AUC values for predicting treatment response were 0.781 for the tumor-only model (random forest), 0.726 for the mesorectum-only model (logistic regression), and 0.837 for the combined model (logistic regression).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusion: &lt;/strong&gt;Radiomic features derived from both the tumor and mesorectum demonstrated complementary prognostic value in predicting treatment response. The inclusion of mesorectal features substantially improved model performance, with the combined model achieving the highest AUC value. These findings highlight the added predictive contribution of the mesorectum as a key peritumoral structure in radiomics-ba","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.7,"publicationDate":"2025-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144820905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Findings of suspicious calcifications on contrast-enhanced mammography and their pathological correlation.","authors":"Dilşah Oral, İhsan Şebnem Örgüç, Hanife Seda Mavili, Teoman Coşkun","doi":"10.4274/dir.2025.253352","DOIUrl":"https://doi.org/10.4274/dir.2025.253352","url":null,"abstract":"<p><strong>Purpose: </strong>This study aimed to determine the performance of contrast-enhanced mammography (CEM) in evaluating suspicious calcifications not associated with a mass.</p><p><strong>Methods: </strong>Patients with suspicious calcifications detected on CEM performed at our center between February 2021 and December 2023 were included in the study. Retrospectively, the morphology, distribution, and longest axis length of the calcifications were assessed on low-energy images, whereas contrast enhancement intensity, pattern, longest axis length, and enhancement curves were analyzed on recombined images. The pathological diagnosis, grade, Ki-67 index, and (if available) the longest lesion length in the surgical specimen were recorded. Using pathology as the gold standard, various CEM parameters were evaluated for their performance in assessing this group of calcifications. Primary and secondary analyses were performed based on combined low or no enhancement and no enhancement alone, respectively.</p><p><strong>Results: </strong>Our study includes 132 lesions in 114 patients,18 of whom had bilateral calcifications. Of the 132 lesions included in the study, 78 were benign, and 54 were malignant. Sensitivity, specificity, positive predictive value, and negative predictive value were determined as follows: 72.2%, 62.8%, 57.3%, and 76% in low-energy images; 79.6%, 80.8%, 74.1%, and 85.1% in the primary analysis of recombined images; and 98.2%, 47.4%, 56.4%, and 97.4% in the secondary analysis. Contrast enhancement intensity was identified as a significant parameter influencing malignancy risk. A strong statistical correlation was observed between lesion length measurements in both low-energy and recombined images compared with pathology (r = 0.733 and r = 0.879, <i>P</i> < 0.001 for both), with mean differences of -4.75 mm and +4.45 mm. No statistically significant relationship was found between contrast enhancement intensity and the distinction between invasive and <i>in situ</i> carcinoma (<i>P</i> = 0.698) or the differentiation of ductal carcinoma <i>in situ</i> grade (<i>P</i> = 0.336). A significant correlation was detected between pathology and dynamic contrast enhancement types adapted from magnetic resonance imaging (MRI) (<i>P</i> = 0.019). Although no statistically significant linear correlation was found between the Ki-67 index and contrast enhancement intensity, the <i>P</i> value was close to significance (<i>P</i> = 0.057).</p><p><strong>Conclusion: </strong>CEM demonstrates strong performance in the assessment of suspicious calcifications by combining the morphological and distributional features of digital mammography with enhancement characteristics similar to MRI.</p><p><strong>Clinical significance: </strong>The findings support that CEM exhibits effective performance in evaluating suspicious calcifications not associated with a mass and may have a potential role in routine clinical practice.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.4,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144674147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Non-pulmonary postoperative complications of cardiothoracic surgery.","authors":"Furkan Ufuk, İclal Ocak, Lydia Chelala, Luis Landeras","doi":"10.4274/dir.2025.253354","DOIUrl":"https://doi.org/10.4274/dir.2025.253354","url":null,"abstract":"<p><p>Cardiothoracic surgery, including coronary artery bypass grafting, valve replacement, and transplantation, has considerably advanced, improving patient survival and outcomes. However, non-pulmonary postoperative complications remain a major concern, contributing to morbidity and mortality. These complications encompass cardiovascular events, vascular injuries, infections, and device-related issues that can severely impact recovery. Early diagnosis and timely intervention are crucial to mitigating risks and improving patient outcomes. Advanced imaging modalities such as computed tomography, magnetic resonance imaging, and echocardiography play a pivotal role in identifying and characterizing complications before clinical deterioration occurs. This review highlights the spectrum of acute non-pulmonary complications following cardiothoracic surgery, emphasizing the diagnostic value of imaging in guiding clinical decision-making. By improving the awareness of imaging findings associated with postoperative complications, radiologists and clinicians can facilitate early detection, enabling prompt surgical or medical interventions. A multidisciplinary approach that integrates imaging surveillance with clinical assessment is essential for optimizing patient care and reducing long-term morbidity.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.4,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144674148","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Short to mid-term outcomes of flow re-direction endoluminal device X (FRED^TM X) in the management of intracranial aneurysms: a meta-analysis.","authors":"Alperen Elek, Gülfem Nur Dindar, Sidar Karagöz, Semra Yücel, Eda Teomete, Celal Çınar, Mahmut Küsbeci, Egemen Öztürk, İsmail Oran","doi":"10.4274/dir.2025.253309","DOIUrl":"https://doi.org/10.4274/dir.2025.253309","url":null,"abstract":"<p><p>This meta-analysis evaluates the clinical and angiographic outcomes of the flow re-direction endoluminal device X (FRED^TM X) in treating intracranial aneurysms. A systematic review was performed across Medline, Scopus, and Web of Science databases from inception to March 2025. Eligible studies included those reporting clinical and angiographic results of FRED X treatment. Favorable outcomes were defined as those stated explicitly in the studies or a modified Rankin scale score of 0-2. Pooled estimates were calculated using a random-effects model in R. A total of nine studies encompassing 780 patients with 869 aneurysms were included. The weighted mean age was 56.28 years, with 19.1% of patients being men. Most aneurysms were saccular (85.7%), unruptured (92.52%), and located in the anterior circulation (73.6%), primarily in the internal carotid artery. The average aneurysm size was 13.12 mm. All studies employed dual antiplatelet therapy, with antiplatelet response testing performed in eight studies. The mean clinical follow-up period was 9.27 months. The meta-analysis demonstrated favorable neurological outcomes in 97.71% of cases and complete or near-complete occlusion in 86.9%. Procedure-related complications were reported in 9.28% of cases, while in-stent thrombosis or intimal hyperplasia occurred in 4.29%. Overall mortality was low at 0.60%. Subgroup analysis revealed that unruptured aneurysms had a 100% rate of favorable neurological outcomes and an 84.76% rate of complete or near-complete occlusion. Complication and mortality rates were 7.76% and 0.25%, respectively. In addition, favorable outcomes were seen in 100% of ruptured aneurysm cases; however, complete occlusion was achieved in only 59.65%, and the mortality rate was higher at 9.19%. Therefore, FRED X demonstrated high efficacy and procedural safety in the treatment of intracranial aneurysms, offering improved outcomes compared with earlier-generation flow diverters.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.4,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144674110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Computed tomography-guided irreversible electroporation for a pubic bone metastasis: a technical note and early experience.","authors":"Claudio Pusceddu, Eliodoro Faiella, Claudio Cau, Pierluigi Rinaldi, Paolo Cossu, Salvatore Marsico","doi":"10.4274/dir.2025.253402","DOIUrl":"10.4274/dir.2025.253402","url":null,"abstract":"<p><p>Bone metastases are common in advanced solid tumors and often require local control strategies in addition to systemic therapy. Although thermal ablation is an established method for selected lesions, its use may be limited in anatomically complex or heat-sensitive locations. Irreversible electroporation, a non-thermal ablative technique widely used in visceral oncology, has not yet been reported for the treatment of bone metastases.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.4,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144575058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence in radiology examinations: a psychometric comparison of question generation methods.","authors":"Emre Emekli, Betül Nalan Karahan","doi":"10.4274/dir.2025.253407","DOIUrl":"https://doi.org/10.4274/dir.2025.253407","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;This study aimed to evaluate the usability of artificial intelligence (AI)-based question generation methods-Chat Generative Pre-trained Transformer (ChatGPT)-4o (a non-template-based large language model) and a template-based automatic item generation (AIG) method-in the context of radiology education. The primary objective was to compare the psychometric properties, perceived quality, and educational applicability of generated multiple-choice questions (MCQs) with those written by a faculty member.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;Fifth-year medical students who participated in the radiology clerkship at Eskişehir Osmangazi University were invited to take a voluntary 15-question examination covering musculoskeletal and rheumatologic imaging. The examination included five MCQs from each of three sources: a radiologist educator, ChatGPT-4o, and the template-based AIG method. Student responses were evaluated in terms of difficulty and discrimination indices. Following the examination, students rated each question using a Likert scale based on clarity, difficulty, plausibility of distractors, and alignment with learning goals. Correlations between students' examination performance and their theoretical/practical radiology grades were analyzed using Pearson's correlation method.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;A total of 115 students participated. Faculty-written questions had the highest mean correct response rate (2.91 ± 1.34), followed by template-based AIG (2.32 ± 1.66) and ChatGPT-4o (2.3 ± 1.14) questions (&lt;i&gt;P&lt;/i&gt; &lt; 0.001). The mean difficulty index was 0.58 for faculty, and 0.46 for both template- based AIG and ChatGPT-4o. Discrimination indices were acceptable (≥0.2) or very good (≥0.4) for template-based AIG questions. In contrast, four of the ChatGPT-generated questions were acceptable, and three were very good. Student evaluations of questions and the overall examination were favorable, particularly regarding question clarity and content alignment. Examination scores showed a weak correlation with practical examination performance (&lt;i&gt;P&lt;/i&gt; = 0.041), but not with theoretical grades (&lt;i&gt;P&lt;/i&gt; = 0.652).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusion: &lt;/strong&gt;Both the ChatGPT-4o and template-based AIG methods produced MCQs with acceptable psychometric properties. While faculty-written questions were most effective overall, AI-generated questions- especially those from the template-based AIG method-showed strong potential for use in radiology education. However, the small number of items per method and the single-institution context limit the robustness and generalizability of the findings. These results should be regarded as exploratory, and further validation in larger, multicenter studies is required.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Clinical significance: &lt;/strong&gt;AI-based question generation may potentially support educators by enhancing efficiency and consistency in assessment item creation. These methods may complement traditional approaches to h","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.4,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144674142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence in radiology: diagnostic sensitivity of ChatGPT for detecting hemorrhages in cranial computed tomography scans.","authors":"Olga Bayar-Kapıcı, Erman Altunışık, Feyza Musabeyoğlu, Şeyda Dev, Ömer Kaya","doi":"10.4274/dir.2025.253456","DOIUrl":"https://doi.org/10.4274/dir.2025.253456","url":null,"abstract":"<p><strong>Purpose: </strong>Chat Generative Pre-trained Transformer (ChatGPT)-4V, a large language model developed by OpenAI, has been explored for its potential application in radiology. This study assesses ChatGPT-4V's diagnostic performance in identifying various types of intracranial hemorrhages in non-contrast cranial computed tomography (CT) images.</p><p><strong>Methods: </strong>Intracranial hemorrhages were presented to ChatGPT using the clearest 2D imaging slices. The first question, \"Q1: Which imaging technique is used in this image?\" was asked to determine the imaging modality. ChatGPT was then prompted with the second question, \"Q2: What do you see in this image and what is the final diagnosis?\" to assess whether the CT scan was normal or showed pathology. For CT scans containing hemorrhage that ChatGPT did not interpret correctly, a follow-up question-\"Q3: There is bleeding in this image. Which type of bleeding do you see?\"-was used to evaluate whether this guidance influenced its response.</p><p><strong>Results: </strong>ChatGPT accurately identified the imaging technique (Q1) in all cases but demonstrated difficulty diagnosing epidural hematoma (EDH), subdural hematoma (SDH), and subarachnoid hemorrhage (SAH) when no clues were provided (Q2). When a hemorrhage clue was introduced (Q3), ChatGPT correctly identified EDH in 16.7% of cases, SDH in 60%, and SAH in 15.6%, and achieved 100% diagnostic accuracy for hemorrhagic cerebrovascular disease. Its sensitivity, specificity, and accuracy for Q2 were 23.6%, 92.5%, and 57.4%, respectively. These values improved substantially with the clue in Q3, with sensitivity rising to 50.9% and accuracy to 71.3%. ChatGPT also demonstrated higher diagnostic accuracy in larger hemorrhages in EDH and SDH images.</p><p><strong>Conclusion: </strong>Although the model performs well in recognizing imaging modalities, its diagnostic accuracy substantially improves when guided by additional contextual information.</p><p><strong>Clinical significance: </strong>These findings suggest that ChatGPT's diagnostic performance improves with guided prompts, highlighting its potential as a supportive tool in clinical radiology.</p>","PeriodicalId":11341,"journal":{"name":"Diagnostic and interventional radiology","volume":" ","pages":""},"PeriodicalIF":1.4,"publicationDate":"2025-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144674143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}