Asunto(s)

Asunto(s)

RESUMEN

PURPOSE: Herein, we assessed the accuracy of large language models (LLMs) in generating responses to questions in clinical radiology practice. We compared the performance of ChatGPT, GPT-4, and Google Bard using questions from the Japan Radiology Board Examination (JRBE). MATERIALS AND METHODS: In total, 103 questions from the JRBE 2022 were used with permission from the Japan Radiological Society. These questions were categorized by pattern, required level of thinking, and topic. McNemar's test was used to compare the proportion of correct responses between the LLMs. Fisher's exact test was used to assess the performance of GPT-4 for each topic category. RESULTS: ChatGPT, GPT-4, and Google Bard correctly answered 40.8% (42 of 103), 65.0% (67 of 103), and 38.8% (40 of 103) of the questions, respectively. GPT-4 significantly outperformed ChatGPT by 24.2% (p < 0.001) and Google Bard by 26.2% (p < 0.001). In the categorical analysis by level of thinking, GPT-4 correctly answered 79.7% of the lower-order questions, which was significantly higher than ChatGPT or Google Bard (p < 0.001). The categorical analysis by question pattern revealed GPT-4's superiority over ChatGPT (67.4% vs. 46.5%, p = 0.004) and Google Bard (39.5%, p < 0.001) in the single-answer questions. The categorical analysis by topic revealed that GPT-4 outperformed ChatGPT (40%, p = 0.013) and Google Bard (26.7%, p = 0.004). No significant differences were observed between the LLMs in the categories not mentioned above. The performance of GPT-4 was significantly better in nuclear medicine (93.3%) than in diagnostic radiology (55.8%; p < 0.001). GPT-4 also performed better on lower-order questions than on higher-order questions (79.7% vs. 45.5%, p < 0.001). CONCLUSION: ChatGPTplus based on GPT-4 scored 65% when answering Japanese questions from the JRBE, outperforming ChatGPT and Google Bard. This highlights the potential of using LLMs to address advanced clinical questions in the field of radiology in Japan.

Asunto(s)

RESUMEN

PURPOSE: To present a novel clinical application of silicon-photomultiplier-based positron emission tomography (SiPM-based PET)/computed tomography (CT), detecting a type II endoleak 5 years after endovascular aneurysm repair (EVAR). TECHNIQUE: SiPM-based PET/CT scans with a standard whole-body protocol were performed for a 73-year-old man with a past medical history of abdominal aortic aneurysms treated with EVAR and currently under investigation of his duodenal papillary carcinoma. The PET/CT demonstrated 18F-fluorodeoxyglucose (FDG) accumulation outside the stent graft in the native sac of the aneurysm. The site of accumulation corresponded to that of the contrast enhancement depicted in the CT angiography taken 1 month earlier. Another CT scan performed 3 months later revealed enlargement of the aneurysm. CONCLUSION: SiPM-based PET/CT, with its superior sensitivity and spatial resolution over conventional PET/CT, can detect type II low-flow endoleaks. CLINICAL IMPACT: Abnormal intra-aneurysmal FDG activity incidentally detected on SiPM-based PET/CT is worthy of attention because it may be indicative of endoleaks. Additional imaging using different modalities should be considered so that the patient would not miss the additional treatment opportunity upon observing sac enlargement. For patients with contraindications for iodine CT contrast media, SiPM-based PET/CT would serve as a suitable alternative.

RESUMEN

ABSTRACT: A 37-year-old man with previous heart transplantation for dilated cardiomyopathy underwent screening for malignancy under posttransplantation immunosuppression. 18 F-FDG PET/CT revealed uptake in 2 peritoneal sites of the pericardium that corresponded to the insertion sites of a left ventricular assist device that was used before transplantation. Additional abnormal uptake in the right axillary artery, aortic arch, and left femoral artery corresponded to the insertion sites for arterial inflow during cardiopulmonary bypass. Knowledge that FDG accumulation may occur at the insertion sites of an extracorporeal-circulation device enables unnecessary tests to be avoided.

Asunto(s)

RESUMEN

In recent years, a paradigm shift from single-photon-emitting radionuclide radiotracers toward positron-emission tomography (PET) radiotracers has occurred in nuclear oncology. Although PET-based molecular imaging of the kidneys is still in its infancy, such a trend has emerged in the field of functional renal radionuclide imaging. Potentially allowing for precise and thorough evaluation of renal radiotracer urodynamics, PET radionuclide imaging has numerous advantages including precise anatomical co-registration with CT images and dynamic three-dimensional imaging capability. In addition, relative to scintigraphic approaches, PET can allow for significantly reduced scan time enabling high-throughput in a busy PET practice and further reduces radiation exposure, which may have a clinical impact in pediatric populations. In recent years, multiple renal PET radiotracers labeled with 11C, 68Ga, and 18F have been utilized in clinical studies. Beyond providing a precise non-invasive read-out of renal function, such radiotracers may also be used to assess renal inflammation. This manuscript will provide an overview of renal molecular PET imaging and will highlight the transformation of conventional scintigraphy of the kidneys toward novel, high-resolution PET imaging for assessing renal function. In addition, future applications will be introduced, e.g. by transferring the concept of molecular image-guided diagnostics and therapy (theranostics) to the field of nephrology.

Asunto(s)

RESUMEN

Patients with pancreatic cancer have a poor prognosis, therefore identifying particular tumor characteristics associated with prognosis is important. This study aims to investigate the utility of radiomics with machine learning using 18F-fluorodeoxyglucose (FDG)-PET in patients with pancreatic cancer. We enrolled 161 patients with pancreatic cancer underwent pretreatment FDG-PET/CT. The area of the primary tumor was semi-automatically contoured with a threshold of 40% of the maximum standardized uptake value, and 42 PET features were extracted. To identify relevant PET parameters for predicting 1-year survival, Gini index was measured using random forest (RF) classifier. Twenty-three patients were censored within 1 year of follow-up, and the remaining 138 patients were used for the analysis. Among the PET parameters, 10 features showed statistical significance for predicting overall survival. Multivariate analysis using Cox HR regression revealed gray-level zone length matrix (GLZLM) gray-level non-uniformity (GLNU) as the only PET parameter showing statistical significance. In RF model, GLZLM GLNU was the most relevant factor for predicting 1-year survival, followed by total lesion glycolysis (TLG). The combination of GLZLM GLNU and TLG stratified patients into three groups according to risk of poor prognosis. Radiomics with machine learning using FDG-PET in patients with pancreatic cancer provided useful prognostic information.

Asunto(s)

RESUMEN

We present a case of a 38-year-old woman who complained with cough, fever, and back pain with a weight loss. F-FDG PET/CT to search fever origin revealed uptake in the tracheobronchial and the left auricular cartilage and wall of the thoracic aorta. She underwent biopsy of the left auricle and was diagnosed with relapsing polychondritis (RP) complicating vasculitis. After steroid therapy, FDG PET/CT demonstrated regression of inflammation, showing decreases in the uptakes. Vasculitis should be considered in case of RP with systemic manifestations. Our case demonstrated the utility of FDG PET/CT in evaluation of RP lesions including aortitis.

Asunto(s)