The accurate segmentation of breast ultrasound images is an important precondition for the lesion determination. The existing segmentation approaches embrace massive parameters, sluggish inference speed, and huge memory consumption. To tackle this problem, we propose T ²KD Attention U-Net (dual-Teacher Knowledge Distillation Attention U-Net), a lightweight semantic segmentation method combined double-path joint distillation in breast ultrasound images. Primarily, we designed two teacher models to learn the fine-grained features from each class of images according to different feature representation and semantic information of benign and malignant breast lesions. Then we leveraged the joint distillation to train a lightweight student model. Finally, we constructed a novel weight balance loss to focus on the semantic feature of small objection, solving the unbalance problem of tumor and background. Specifically, the extensive experiments conducted on Dataset BUSI and Dataset B demonstrated that the T ²KD Attention U-Net outperformed various knowledge distillation counterparts. Concretely, the accuracy, recall, precision, Dice, and mIoU of proposed method were 95.26%, 86.23%, 85.09%, 83.59%and 77.78% on Dataset BUSI, respectively. And these performance indexes were 97.95%, 92.80%, 88.33%, 88.40% and 82.42% on Dataset B, respectively. Compared with other models, the performance of this model was significantly improved. Meanwhile, compared with the teacher model, the number, size, and complexity of student model were significantly reduced (2.2×10 ⁶ vs. 106.1×10 ⁶, 8.4 MB vs. 414 MB, 16.59 GFLOPs vs. 205.98 GFLOPs, respectively). Indeedy, the proposed model guarantees the performances while greatly decreasing the amount of computation, which provides a new method for the deployment of clinical medical scenarios.
Humans ; Breast Neoplasms/diagnostic imaging* ; Female ; Ultrasonography, Mammary/methods* ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Neural Networks, Computer ; Breast/diagnostic imaging*

Medical cross-modal retrieval aims to achieve semantic similarity search between different modalities of medical cases, such as quickly locating relevant ultrasound images through ultrasound reports, or using ultrasound images to retrieve matching reports. However, existing medical cross-modal hash retrieval methods face significant challenges, including semantic and visual differences between modalities and the scalability issues of hash algorithms in handling large-scale data. To address these challenges, this paper proposes a Medical image Semantic Alignment Cross-modal Hashing based on Transformer (MSACH). The algorithm employed a segmented training strategy, combining modality feature extraction and hash function learning, effectively extracting low-dimensional features containing important semantic information. A Transformer encoder was used for cross-modal semantic learning. By introducing manifold similarity constraints, balance constraints, and a linear classification network constraint, the algorithm enhanced the discriminability of the hash codes. Experimental results demonstrated that the MSACH algorithm improved the mean average precision (MAP) by 11.8% and 12.8% on two datasets compared to traditional methods. The algorithm exhibits outstanding performance in enhancing retrieval accuracy and handling large-scale medical data, showing promising potential for practical applications.
Algorithms ; Semantics ; Humans ; Ultrasonography ; Information Storage and Retrieval/methods* ; Image Processing, Computer-Assisted/methods*

To address the high computational complexity of the Transformer in the segmentation of ultrasound thyroid nodules and the loss of image details or omission of key spatial information caused by traditional image sampling techniques when dealing with high-resolution, complex texture or uneven density two-dimensional ultrasound images, this paper proposes a thyroid nodule segmentation method that integrates the receiving weighted key-value (RWKV) architecture and spherical geometry feature (SGF) sampling technology. This method effectively captures the details of adjacent regions through two-dimensional offset prediction and pixel-level sampling position adjustment, achieving precise segmentation. Additionally, this study introduces a patch attention module (PAM) to optimize the decoder feature map using a regional cross-attention mechanism, enabling it to focus more precisely on the high-resolution features of the encoder. Experiments on the thyroid nodule segmentation dataset (TN3K) and the digital database for thyroid images (DDTI) show that the proposed method achieves dice similarity coefficients (DSC) of 87.24% and 80.79% respectively, outperforming existing models while maintaining a lower computational complexity. This approach may provide an efficient solution for the precise segmentation of thyroid nodules.
Thyroid Nodule/diagnostic imaging* ; Humans ; Ultrasonography/methods* ; Algorithms ; Image Processing, Computer-Assisted/methods* ; Thyroid Gland/diagnostic imaging*

Thyroid nodules are a common endocrine disorder, and their early detection and accurate diagnosis are crucial for the prevention of thyroid cancer. However, the highly heterogeneous morphology and boundaries of thyroid nodules pose significant challenges to their precise identification and classification. Traditional diagnostic approaches rely heavily on physicians' experience, which increases the risk of misdiagnosis and missed diagnoses. With the rapid advancement of computer-aided diagnosis (CAD) technologies, applying deep learning algorithms to the analysis of thyroid nodule ultrasound images has shown great potential. This paper reviews the latest research progress on deep learning-based CAD methods for thyroid nodules, with a focus on their applications in image preprocessing, segmentation and classification. The advantages and limitations of current techniques are analyzed, and potential future directions are discussed. This review aims to highlight the potential of deep learning in thyroid nodule diagnosis and to provide a foundation for selecting feasible pathways for future clinical applications.
Humans ; Thyroid Nodule/diagnostic imaging* ; Deep Learning ; Ultrasonography/methods* ; Diagnosis, Computer-Assisted/methods* ; Algorithms ; Thyroid Neoplasms/diagnostic imaging* ; Image Processing, Computer-Assisted/methods*

OBJECTIVE: To approach the evaluation of relative α variant score monitored by bedside continuous electroencephalography and optic nerve sheath diameter (ONSD) evaluated by ultrasound in patients with ischemic hypoxic encephalopathy, and to observe the effect of neurofeedback training on brain function. METHODS: A prospective observational study was conducted. The patients admitted to the emergency and intensive care department of Shanghai Pudong New Area People's Hospital from January 2021 to December 2023, who meet the diagnostic criteria of ischemic hypoxic encephalopathy with the Glasgow coma score (GCS) ≤ 8 at admission receiving neurofeedback training were enrolled as the study object (observation group), and the patients without neurofeedback training and GCS score ≤ 8 at admission were enrolled as the controls (control group). Both groups received intravenous neurotrophic therapy combining ganglioside and cerebrolysin for 10 days as one course of treatment. On this basis, the observation group additionally received continuous neurofeedback training including visual feedback, auditory feedback, meditation and relaxation for 14 days. Bedside continuous electroencephalography was used for monitoring relative α variation score, and ultrasound was used to determine ONSD. The average power and slow wave power [expressed as delta-theta ratio (DTR)] of five channels in electroencephalography before and 14 days after neurofeedback training were examined. The differences in peripheral blood neutrophil/lymphocyte ratio (NLR), Hamilton depression scale (HAMD) score, National Institutes of Health stroke scale (NIHSS) score, plasma levels of 5-hydroxytryptamine (5-HT) and brain-derived neurotrophic factor (BDNF). RESULTS: A total of 60 patients were enrolled in the observation group and 50 patients in the control group finally. There was no significant difference in gender, age or course of disease between the two groups. The ONSD and relative α variant score in the observation group were significantly higher than those in the control group [ONDS (mm): 5.59±0.42 vs. 3.23±0.34, relative α variant score: 2.28±0.39 vs. 0.83±0.28, both P < 0.01]. After neurofeedback training for 14 days, the mean power and DTR in five channels of electroencephalography in the observation group were significantly lower than those before treatment [mean power (μV²/Hz): 95.35±3.61 vs. 102.58±4.23 in frontal pole 1 (Fp1), 38.56±4.73 vs. 46.13±2.36 in frontal 3 (F3), 34.33±5.87 vs. 51.71±4.65 in central 3 (C3), 58.37±4.45 vs. 62.95±3.22 in F7, 45.23±2.41 vs. 54.14±2.45 in temporal 3 (T3); DTR (μV²/Hz): 75.21±11.34 vs. 84.12±11.35 in ground electrode (GND), 72.31±21.67 vs. 88.23±10.25 in reference electrode (REF), 81.34±8.57 vs. 92.41±8.56 in F4, 71.25±5.42 vs. 87.23±5.64 in parietal 3 (P3), 70.12±5.88 vs. 85.67±6.12 in P4; all P < 0.05]. However, there was no significant difference in the mean power of five channels before and after treatment in the control group. There was no significant difference in the HAMD score or NIHSS score before treatment between the two groups. The above scores at 14 days after treatment were significantly lower than before, and the decrease was more significant in the observation group (HAMD score: 4.59±1.06 vs. 10.69±0.97, NIHSS score: 6.81±0.66 vs. 8.45±0.87, both P < 0.01). There was no significant difference in the plasma 5-HT, BDNF or peripheral blood NLR before treatment between the two groups. The above parameters at 14 days after treatment were improved as compared with before, and the levels in the observation group were superior to control group [5-HT (mg/L): 150.25±17.37 vs. 123.34±16.18, BDNF (mg/L): 19.37±2.35 vs. 12.48±2.18, NLR: 4.78±0.83 vs. 5.81±1.17, all P < 0.01]. CONCLUSIONS Both ONDS determined by ultrasound and relative α variation score monitored by electroencephalography changed significantly in the patients with ischemic hypoxic encephalopathy. Neurofeedback training can effectively improve brain function in patients with ischemic hypoxic encephalopathy.
Humans ; Electroencephalography ; Prospective Studies ; Neurofeedback ; Optic Nerve/diagnostic imaging* ; Ultrasonography ; Hypoxia-Ischemia, Brain/physiopathology* ; Male ; Female ; Middle Aged

Objective:To describe a low-cost, reproducible phantom model for training medical practitioners in ultrasound-guided ethanol ablation (EA) of a cystic thyroid nodule.

Methods:The model of cyst contents was created using a mixture of coffee powder, cornstarch and water to mimic colloidal contents. This was injected into the finger of a cut surgical glove secured with transparent tape to serve as an inflatable capsule and placed inside a chicken breast. This setup allows practitioners to perform key ultrasound-guided ethanol ablation techniques including spinal needle insertion, cyst aspiration, saline flushing, re-aspiration and ethanol injection.

Results:The phantom model was assembled for PhP 150.30 per unit and took approximately 5 minutes to construct. This model was demonstrated to be a cost-effective and simple method for enabling specialists to practice and enhance their skills in ultrasound-guided ethanol ablation of a cystic thyroid nodule.

Conclusion:The described phantom model provides an accessible and practical training tool for healthcare providers to gain proficiency in ultrasound-guided ethanol ablation of a cystic thyroid nodule in a safe and controlled setting prior to actual patient handling.

Endoscopic ultrasonography (EUS) has progressed beyond diagnostic imaging to include EUS-guided tissue acquisition and EUS-directed therapies. This review provides an update on EUS-guided drainage and anastomotic procedures, and other therapeutic procedures. Today, EUS-guided drainage of symptomatic walled-off pancreatic fluid collections is the norm, with endoscopic necrosectomy as an adjunct. For high-risk surgical patients unsuitable for cholecystectomy, EUS-guided gallbladder drainage of acute cholecystitis is an option. Additionally, EUS-guided drainage of obstructed biliary and pancreatic ductal system can be performed as salvage procedures after unsuccessful endoscopic retrograde cholangiopancreatography (ERCP). Bariatric procedures such as Roux-en-Y gastric bypass alter the gastric anatomy, hindering access to the major papilla. This can be overcome by creating a conduit through the excluded stomach using EUS-directed transgastric ERCP. Gastric outlet obstruction and afferent loop syndrome can be treated using EUS-guided gastrojejunostomy. These therapeutic interventions are a major advancement in the field of interventional EUS, achieving significant clinical impact.
Humans ; Endosonography/methods* ; Drainage/methods* ; Cholangiopancreatography, Endoscopic Retrograde/methods* ; Ultrasonography, Interventional/methods* ; Anastomosis, Surgical/methods*

OBJECTIVE: To evaluate the clinical efficacy of ultrasound-guided needle knife release with different pathways for carpal tunnel syndrome (CTS). METHODS: Sixty CTS patients were randomly divided into a transverse group and a longitudinal group, with 30 patients in each group. The transverse group received the needle knife release under ultrasound above and below the median nerve along the short axis, while the longitudinal group received the needle knife release under ultrasound above the median nerve along the long axis. The patients' visual analogue scale (VAS) scores, Levine scores, and median nerve cross-section area (CSA) were assessed before treatment, on the 7th, 14th, and 30th days after treatment. The median nerve conduction indexes (motor conduction: latency [Lat], amplitude[Amp], motor conduction velocity [MCV]; sensory conduction: Lat, Amp, sensory conduction velocity [SCV]) were evaluated before treatment and on the 30th day after treatment. RESULTS: Compared before treatment, the VAS scores, Levine scores, and median nerve CSA were reduced in both groups on the 7th, 14th, and 30th days after treatment (P<0.05); the transverse group showed lower VAS and Levine scores and smaller CSA than those in the longitudinal group (P<0.05). On the 30th day after treatment, motor and sensory Lat was reduced (P<0.05), motor and sensory Amp was increased (P<0.05), and MCV and SCV were faster (P<0.05) in both groups. The transverse group had shorter Lat, higher Amp, and faster MCV and SCV than those in the longitudinal group (P<0.05). CONCLUSION The ultrasound-guided needle knife release with transverse and longitudinal pathways could both effectively alleviate pain and improve median nerve conduction in CTS patients, with the transverse pathway showing superior efficacy to longitudinal pathway.
Humans ; Carpal Tunnel Syndrome/therapy* ; Female ; Middle Aged ; Male ; Adult ; Median Nerve/physiopathology* ; Acupuncture Therapy/instrumentation* ; Treatment Outcome ; Neural Conduction ; Aged ; Ultrasonography

OBJECTIVE: To evaluate the clinical efficacy of ultrasound-guided acupuncture at myofascial trigger points (MTrPs) on treating post-stroke foot drop. METHODS: Sixty patients with post-stroke foot drop were randomly assigned to an observation group 1 (20 cases, 1 case dropped out), an observation group 2 (20 cases, 2 casses dropped out), and a control group (20 cases). The control group received conventional acupuncture at Yanglingquan (GB34), Jiexi (ST41), Taichong (LR3), Zusanli (ST36), Xuanzhong (GB39), and Qiuxu (GB40) on the affected side, once daily. In addition to the treatment of the control group , the observation group 1 received acupuncture at the tibialis anterior and gastrocnemius MTrPs, once every other day, while the observation group 2 received ultrasound-guided acupuncture at the tibialis anterior and gastrocnemius MTrPs, once every other day. All groups were treated for two weeks. Three-dimensional gait analysis was performed using an infrared motion capture system, and the Holden walking scale was used to evaluate walking ability before and after treatment in the three groups. RESULTS: Compared before treatment, the patients in the observation groups 1 and 2 showed increased walking speed (P<0.05, P<0.01), and improved Holden walking scale grades (P<0.05, P<0.01) after treatment; the patients in the observation group 2 also showed increased ankle dorsiflexion angles (P<0.05). The walking speeds of the observation groups 1 and 2 were faster than those of the control group after treatment (P<0.05), the Holden walking scale grade in the observation group 2 was superior to that in the control group (P<0.05). CONCLUSION The ultrasound-guided acupuncture at MTrPs could effectively improve gait function in post-stroke foot drop patients.
Humans ; Acupuncture Therapy ; Male ; Female ; Middle Aged ; Stroke/physiopathology* ; Aged ; Trigger Points/physiopathology* ; Gait ; Acupuncture Points ; Adult ; Ultrasonography ; Treatment Outcome ; Gait Disorders, Neurologic/etiology*

This paper analyzed the causes of peripheral nerve injury induced by acupoint injection, and proposed methods for prevention. These methods included emphasizing the physicochemical properties of medications and strengthening research on medication compatibility, classifying high-risk acupoints and establishing international standards for safe acupoint needling, standardizing clinical procedures for acupoint injection, and incorporating ultrasound technology when necessary to improve the accuracy and safety of the procedure. These strategies aimed to reduce the risk associated with the clinical application of acupoint injection.
Humans ; Peripheral Nerve Injuries/prevention & control* ; Ultrasonography ; Acupuncture Points ; Injections/adverse effects*