Medical ultrasonography最新文献

Aim: To evaluate the effectiveness of two large language models, ChatGPT-4 and Claude 3, in improving the accuracy of question responses by senior sonologist and junior sonologist.

Material and methods: A senior and a junior sonologist were given a practice exam. After answering the questions, they reviewed the responses and explanations provided by ChatGPT-4 and Claude 3. The accuracy and scores before and after incorporating the models' input were analyzed to compare their effectiveness.

Results: No statistically significant differences were found between the two models' responses scores for each section (all p>0.05). For junior sonologist, both ChatGPT-4 (p=0.039) and Claude 3 (p=0.039) significantly improved scores in basic knowledge. The responses provided by ChatGPT-4 also significantly improved scores in relevant professional knowledge (p=0.038), though their explanations did not (p=0.077). For all exam sections, both models' responses and explanations significantly improved scores (all p<0.05). For senior sonologist, both ChatGPT-4's responses (p=0.022) and explanations (p=0.034) improved scores in basic knowledge, as did Claude 3's explanations (p=0.003). Across all sections, Claude 3's explanations significantly improved scores (p=0.041).

Conclusion: ChatGPT-4 and Claude 3 significantly improved sonologist' examination performance, particularly in basic knowledge.

Aims: To investigate the multimodal ultrasonography (US) features of extramedullary plasmacytomas (EMPs) in patients with multiple myeloma (MM).

Material and methods: Nine patients with MM who were diagnosed EMPs and underwent multimodal ultrasonography were enrolled in this study. Clinical information and multimodal US results were retrospectively analyzed. The multimodal US with a standardized sonographic protocol including US, color Doppler flow imaging (CDFI), microvascular US, ultrasound elastography (UE) and contrast-enhanced ultrasound (CEUS) was performed on the enrolled patients.

Results: This study included both males (4 of 9) and females (5 of 9), with ages ranging from 48 to 72 years. The nine lesions were located in the skin/muscle (7 of 9) and pleura (2 of 9). On US images, all lesions showed inhomogeneous solid hypoechogenicity without calcification. Vascularity was abundant (hypervascularization). CEUS images confirmed this vascular pattern (9 of 9) and presented hyperenhancement with enlarged range. On elastographic images, the lesions presented with high elastic score (3-4).

Conclusions: An inhomogeneous hypoechoic soft tissue or visceral mass occurring in MM with increased vascularity, low elasticity, and hyperenhancement along with enlarged range can indicate the diagnosis of EMP on multimodal US images.

Aim: Accurate prediction of preoperative molecular subtypes of breast cancer is crucial for treatment planning and prognosis evaluation of patients. This systematic review aims to investigate the capacity of ultrasound radiomics in accurately identifying the molecular subtypes of breast cancer.

Material and methods: We conducted a thorough search of  PubMed, Embase, and Cochrane databases to identify relevant research up until May 2024. We pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, diagnostic odds ratio, and the area under the curve (AUC) to summarize the ability of ultrasound radiomics to predict Luminal type and triple negative breast cancer (TNBC) type in patient with breast cancer. Results: Our meta-analysis found that nine studies provided ultrasound radiomics predictions for TNBC. Pooled sensitivity, specificity, and AUC were 0.64, 0.89, and 0.86, respectively. Four studies have provided the ultrasound radiomics prediction of Luminal molecular subtype breast cancer. Pooled sensitivity, specificity, and AUC were 0.89, 0.79, and 0.90, respectively. There are two studies that predict Luminal A, Luminal B and Her 2+molecular subtype by ultrasound radiology, but they cannot be synthesized quantitatively because of the small number of studies.

Conclusion: Ultrasound radiomics has a good diagnostic performance in predicting molecular subtypes in breast cancer.

Aim: Transcranial ultrasonography, including transcranial doppler (TCD) and transcranial colour-coded duplex/sonography (TCCD/TCCS), offers potential benefits in the evaluation of cerebrovascular patency, but its accuracy remains under scrutiny. Hence, this study aims to investigate the diagnostic accuracy of transcranial ultrasonography for detecting stenosis in acute ischaemic stroke patients.

Material and methods: We systematically reviewed the databases such as CINAHL, SCOPUS, EMBASE, MEDLINE, Cochrane Library, Google Scholar, and ScienceDirect from inception till June 2023. Studies were eligible if they compared the diagnostic role of transcranial ultrasonography with established techniques, and reported accuracy parameters such as sensitivity, specificity, positive and negative likelihood ratios. Subgroup analysis was performed for TCD and TCCD/TCCS.

Results: Seventeen studies involving 1496 participants were included. The pooled sensitivity and specificity of transcranial ultrasonography were 90% and 95% respectively. For TCD, the sensitivity and specificity were 87% and 93% respectively. TCCD/TCCS showed sensitivity 98% and specificity 97%. High clinical value was evident in both TCD and TCCD/TCCS, with heterogeneity across studies but no observable publication bias.

Conclusion: Transcranial ultrasonography, including both TCD and TCCD/TCCS, demonstrates high diagnostic accuracy in detecting stenosis in patients with acute ischaemic stroke. These findings support the utility of these methods as reliable, non-invasive tools for cerebrovascular assessment.

Aim: This study aimed to evaluate the reproducibility of Ultrasound Attenuation (USAT), the Liver Texture Index (LTI) and the HepatoRenal Index (HRI).

Material and methods: Between May 2023 to December 2023, 279 patients who underwent USAT, LTI and HRI tests at the Ultrasound Department of The Second Affiliated Hospital, School of Medicine, The Chinese University of Hong Kong, Shenzhen were included. All patients were examined by a physician with over 4 years of ultrasound experience. The intra-class correlation coefficient (ICC) was utilized to assess the reproducibility of the three quantitative parameters in all patients. Additionally, 63 cases were examined by another physician with six months of experience to evaluate inter-operator reproducibility.

Results: The intra-operator reproducibility of USAT, LTI, and HRI were 0.879 (95% CI:0.858-0.898), 0.919 (95% CI:0.904-0.933), and 0.836 (95% CI:0.809-0.862), respectively. The inter-operator reproducibility among different operators were 0.790 (95% CI:0.604-0.894), 0.782 (95% CI:0.591-0.890), and 0.631 (95% CI:0.356-0.806). After 30 operations, the inter-operator reproducibility improved to 0.858 (95% CI:0.731-0.927), 0.909 (95% CI:0.823-0.954), and 0.796 (95% CI:0.587-0.899).

Conclusions: USAT, LTI, and HRI have shown good intra-operator reproducibility, and inter-operator reproducibility between the two operators reached a comparable level after 30 training cases.