Medical ultrasonography最新文献

.

.

.

Aims: Ultrasound (US) is an essential diagnostic and educational tool in medical practice, and its effective implementation into medical curricula is critical. This study aimed to compare the efficacy of two disparate educational approaches-an optional semester course and a specifically curated intensive workshop-on the learning curve of medical students in abdominal ultrasonography.

Material and methods: Engaging fourth and fifth-year medical students, this study, incorporated both theoretical and practical elements of US, providing participants with hands-on experience and evaluative assessments pre- and post-training. Students were segregated into two groups: one experienced a 14-hour optional semester course and the other a 6-hour intensive workshop, both yielding distinct teaching methodologies yet aspiring for synonymous educational outcomes.

Results: Involving a total of 93 participants, findings elucidated that regardless of the educational method employed, post-training identification of US structures exhibited a significant enhancement compared to pre-training. Interestingly, no substantial disparities were discerned between the two educational approaches nor gender-based differences in learning outcomes.

Conclusions: This investigation provides pivotal insights into the versatile utility of different educational strategies in abdominal US training for medical students, affirming that varied pedagogical methods can achieve comparable augmentations in student proficiency. Further research is paramount to ascertain the optimal integration of US education into medicalcurricula, considering aspects such as duration, depth, and mode of delivery.

Aims: To develop a deep learning model, with the aid of ChatGPT, for thyroid nodules, utilizing ultrasound images. The cytopathology of the fine needle aspiration biopsy (FNAB) serves as the baseline.

Material and methods: After securing IRB approval, a retrospective study was conducted, analyzing thyroid ultrasound images and FNAB results from 1,061 patients between January 2017 and January 2022. Detailed examinations of their demographic profiles, imaging characteristics, and cytological features were conducted. The images were used for training a deep learning model to identify various thyroid pathologies. ChatGPT assisted in developing this model by aiding in code writing, preprocessing, model optimization, and troubleshooting.

Results: The model demonstrated an accuracy of 0.81 on the testing set, within a 95% confidence interval of 0.76 to 0.87. It presented remarkable results across thyroid subgroups, particularly in the benign category, with high precision (0.78) and recall (0.96), yielding a balanced F1-score of 0.86. The malignant category also displayed high precision (0.82) and recall (0.92), with an F1-score of 0.87.

Conclusions: The study demonstrates the potential of artificial intelligence, particularly ChatGPT, in aiding the creation of robust deep learning models for medical image analysis.

.

.

.

Aim: To assess the effectiveness of shear wave elastography (SWE) in diagnosing delaminated partial-thickness rotator cuff tears (DPT-RCT).

Material and methods: A retrospective study was carried out on 137 patients with DPT-RCT. The study included complete clinical data, including the images of conventional ultrasound (US), SWE, Magnetic Resonance Imaging (MRI) and shoulder arthroscopic surgery. The features of US, SWE, and MRI were evaluated. The study analysed the Shear-Wave Velocity (SWV) among three types of DPT-RCT, and between the regions of tears, normal contralateral, and affected unilateral supraspinatus tendon. Furthermore, receiver operating characteristic (ROC) curves were evaluated.

Results:  The SWE detection rate was significantly higher (91.2%) compared to US (73.7%) and MRI (87.6%) for the overall diagnosis of DPT-RCT. Similarly, SWE yielded higher rates of detection for types 1 (89.5%) and 2 (92.3%) of DPT-RCT as compared to US (71.7%, 69.2%) and MRI (81.6%, 94.9%), respectively. However, there was no significant difference in the accuracy of diagnosing type 3 among the three methods. The SWV of the 137 supraspinatus tendon tears was 3.64±0.60 m/s, which was higher than that of the normal supraspinatus tendon (2.43±0.47 m/s, p<0.01) as well as the region of tears (1.61±0.54 m/s, p<0.01). Nevertheless, there was no significant difference in SWV among the three types of DPT-RCT. The cutoff thresholds of SWV for identifying normal tendon from DPT-RCT and for identifying DPT-RCT from the region of tears were 2.96m/s and 2.39m/s, respectively.

Conclusions: SWE with SWV can provide both quantitative and qualitative diagnostic information for DPT-RCT, which can be used as a crucial supplement imaging method.

Aim: To evaluate the feasibility of ultrasound (US) in identification of nerve lesions after breast cancer surgery in patients with neuropathic pain and assess the effect of a targeted US-guided therapy.

Material and methods: Patients with neuropathic pain after breast cancer surgery underwent US examination. Nerve lesions identified by US were treated by a US-guided application of a mixture of local anesthetics and corticoids. The patients reported pain relief on a 100-point scale (0% = no effect, 100% = complete relief) and its duration in the next 18 months.

Results: We performed 17 interventions in 11 women. A neuroma was observed in 2 patients, edema of the nerve in 5 patients, and scarring across the nerve in 4 patients. The affected nerves were the intercostobrachial nerve (5 patients), the long thoracic nerve (4), cutaneous branch of the pectoral nerve (1), and both the intercostobrachial and the long thoracic nerve (1). After 15 (88%) interventions, the patients reported relief (55±32%) with a median duration of 3 months (0.5-18 months).

Conclusion: In patients after breast cancer surgery, ultrasound can reliably identify small painful neural lesions which can be efficiently treated by ultrasound-guided intervention.