PURPOSE:To evaluate the longitudinal evolution of transabdominal bowel ultrasound parameters during ustekinumab therapy in patients with postoperative recurrence of Crohn's disease (CD), and to investigate the association between ultrasound changes and clinical and biochemical outcomes. METHODS:This retrospective longitudinal study included 65 patients with endoscopically confirmed postoperative recurrence of CD who received ustekinumab therapy between April 2020 and December 2024. Patients were classified into improvement and non-improvement groups according to endoscopic outcomes at week 24. Serial transabdominal bowel ultrasound examinations were performed at baseline and at weeks 8, 16, and 24, including assessment of bowel wall thickness, Limberg vascularity grade, bowel wall stratification, perienteric fat edema, and luminal stenosis. Longitudinal changes in ultrasound parameters were analyzed using mixed-effects models and generalized estimating equations, as appropriate for each outcome type. Associations between bowel wall thickness changes and clinical/biochemical outcomes were evaluated using Spearman correlation analysis. RESULTS:Distinct temporal response patterns were observed among bowel ultrasound parameters during ustekinumab therapy. In the improvement group, bowel wall thickness and Limberg grade improved as early as week 8, bowel wall stratification improved at week 16, and perienteric fat edema improved at week 24, whereas no significant longitudinal changes were observed in the non-improvement group.Linear mixed-effects model analysis demonstrated a significant time × group interaction for bowel wall thickness (P < 0.001). Generalized estimating equation analyses demonstrated significant time × group interactions for Limberg grade, bowel wall stratification, and perienteric fat edema (all P < 0.001), while luminal stenosis remained relatively stable throughout follow-up (all P > 0.05).At week 24, patients in the improvement group showed significantly lower CRP levels (P < 0.001) and HBI scores (P = 0.006) than those in the non-improvement group. Furthermore, reductions in bowel wall thickness were significantly correlated with reductions in CRP levels (rs = 0.678, 95 % CI: 0.503-0.806, P < 0.001) and HBI scores (rs = 0.566, 95 % CI: 0.359-0.722, P < 0.001). CONCLUSION:Transabdominal bowel ultrasound may dynamically reflect transmural inflammatory changes during ustekinumab therapy in patients with postoperative recurrence of CD. Different ultrasound parameters demonstrated distinct temporal recovery patterns, with bowel wall thickness and vascularity showing the earliest improvement. The association between bowel wall thickness changes and CRP/HBI further supports the potential clinical utility of bowel ultrasound as a noninvasive longitudinal monitoring tool in this high-risk postoperative population.

The next-generation of artificial intelligence technology has contributed significantly to the development of medical intelligence. However, the widespread use of deep neural networks (DNNs) has also brought about serious security threats. In this paper, we present an adversarial attack approach for deep learning-based image segmentation models in the field of medical image analysis. In our solutions, we propose a novel adversarial attack method, which is designed to exploit the DNNs' generic down-sampling operation to ensure the effectiveness, stealthiness, and transferability of the attack. We perform the attack on two State-Of-The-Art (SOTA) models, DDANet and CaraNet in a general medical image dataset Kvasir-SEG, and a comprehensive evaluation shows that our attack is effective stealthy, and transferrable.

Masked Autoencoders (MAEs), the state-of-the-art self-supervised neural network architecture in miscellaneous vision tasks, show surprisingly effective potential in reconstructing images distorted by random masking. This paper first introduces an optical implementation of MAEs, employing digital micromirror devices in the optical path to capture partially blocked images. MAEs with multi-scale patches are deployed in the reconstruction procedure. By using an optical-specialized version of the reconstruction network, the system can reconstruct original scenes of high quality. Simulations and experimental measurements showed a significant performance, achieving 24.41 dB average peak-signal-to-noise on Davis2017 datasets and 29.92 dB (masked areas) on authentic captured images under 70% of pixels being blocked. This paves the way for the application of low-bandwidth sampling of high-throughput high-resolution images.

Snapshot compressive imaging (SCI) can record the 3D information by a 2D measurement and from this 2D measurement to reconstruct the original 3D information by reconstruction algorithm. As we can see, the reconstruction algorithm plays a vital role in SCI. Recently, deep learning algorithm show its outstanding ability, outperforming the traditional algorithm. Therefore, to improve deep learning algorithm reconstruction accuracy is an inevitable topic for SCI. Besides, deep learning algorithms are usually limited by scalability, and a well trained model in general can not be applied to new systems if lacking the new training process. To address these problems, we develop the ensemble learning priors to further improve the reconstruction accuracy and propose the scalable learning to empower deep learning the scalability just like the traditional algorithm. What's more, our algorithm has achieved the state-of-the-art results, outperforming existing algorithms. Extensive results on both simulation and real datasets demonstrate the superiority of our proposed algorithm. The code and models will be released to the public.

Snapshot compressive imaging (SCI) can record a 3D datacube by a 2D measurement and algorithmically reconstruct the desired 3D information from that 2D measurement. The reconstruction algorithm thus plays a vital role in SCI. Recently, deep learning (DL) has demonstrated outstanding performance in reconstruction, leading to better results than conventional optimization-based methods. Therefore, it is desirable to improve DL reconstruction performance for SCI. Existing DL algorithms are limited by two bottlenecks: 1) a high-accuracy network is usually large and requires a long running time; 2) DL algorithms are limited by scalability, i.e., a well-trained network cannot generally be applied to new systems. To this end, this paper proposes to use ensemble learning priors in DL to achieve high reconstruction speed and accuracy in a single network. Furthermore, we develop the scalable learning approach during training to empower DL to handle data of different sizes without additional training. Extensive results on both simulation and real datasets demonstrate the superiority of our proposed algorithm. The code and model can be accessed at https://github.com/integritynoble/ELP-Unfolding/tree/master.

Ferroptosis, as a newly discovered cell death form, has become an attractive target for precision cancer therapy. Several ferroptosis therapy strategies based on nanotechnology have been reported by either increasing intracellular iron levels or by inhibition of glutathione (GSH)-dependent lipid hydroperoxidase glutathione peroxidase 4 (GPX4). However, the strategy by simultaneous iron delivery and GPX4 inhibition has rarely been reported. Herein, novel tumor microenvironments (TME)-activated metal-organic frameworks involving Fe & Cu ions bridged by disulfide bonds with PEGylation (FCSP MOFs) were developed, which would be degraded specifically under the redox TME, simultaneously achieving GSH-depletion induced GPX4 inactivation and releasing Fe ions to produce ROS via Fenton reaction, therefore causing ferroptosis. More ROS could be generated by the acceleration of Fenton reaction due to the released Cu ions and the intrinsic photothermal capability of FCSP MOFs. The overexpressed GSH and H2O2 in TME could ensure the specific TME self-activated therapy. Better tumor therapeutic efficiency could be achieved by doxorubicin (DOX) loading since it can not only cause apoptosis, but also indirectly produce H2O2 to amplify Fenton reaction. Remarkable anti-tumor effect of obtained FCSP@DOX MOFs was verified via both in vitro and in vivo assays.

Hydrogen therapy is an emerging and highly promising strategy for the treatment of inflammation-related diseases. However, nonpolarity and low solubility of hydrogen under the physiological conditions results in a limited therapeutic effect. Herein, we develop a biocompatible magnesium micromotor coated with hyaluronic acid as a hydrogen generator for precise rheumatoid arthritis management. The hydrogen bubbles generated locally not only function as a propellant for the motion but also function as the active ingredient for reactive oxygen species (ROS) and inflammation scavenging. Under ultrasound guidance, the micromotors are injected intra-articularly, and the dynamics of the micromotors can be visualized. By scavenging ROS and inflammation via active hydrogen, the oxidative stress is relieved and the levels of inflammation cytokines are reduced by our micromotors, showing prominent therapeutic efficacy in ameliorating joint damage and suppressing the overall arthritis severity toward a collagen-induced arthritis rat model. Therefore, our micromotors show great potential for the therapy of rheumatoid arthritis and further clinical transformation.

Overproduction of hydrogen peroxide (H2O2) is a characteristic feature for inflammation and cancer. Ultrasonography is a safe and effective clinical diagnosis method to identify inflammation, yet a fully biodegradable ultrasound contrast agent (UCA) system that is able to reflect the pathophysiological level of H(2)O(2)at diseased tissue with high sensitivity is still scarce. Herein, a self-assembled biodegradable catalase-modified poly(ethylene glycol)(44)-b-poly(lactic acid)(100)(PEG-b-PLA) micelle (Cat-PEG-PLA) is presented as a sensitive ultrasound contrast agent. It is found that the imaging effect of Cat-PEG-PLA is remarkable with an obvious enhancement of ultrasound echogenic signal even in an ultralow (0.33x10(-3)m) inflammation-relevant H(2)O(2)concentration environment, close to or even lower than the reported lowest ultrasound detectable concentration limits. The ultrasonography of collagen-induced arthritis rat model's articular cavity by the Cat-PEG-PLA micelle platform is demonstrated, showing highly promising results. Therefore, with superior detection sensitivity and contrast, the system detects H(2)O(2)in arthritis rat model with full biodegradability, providing a novel, safe, and translatable platform for arthritis diagnosis.

A theranostic nanozyme (Au NCs-ICG) decomposes intratumoral H2O2 to O2, subsequently enhancing photodynamic therapy and radiotherapy with the guidance of multimodal imaging.

The efficiency of drug delivery and bioavailability to tumor cells are crucial for effective cancer chemotherapy. Herein, a doxorubicin (DOX) encapsulated lysolipid-based thermosensitive liposome decorated with cRGD peptide (RTSL) is conjugated on the surface of an IR780-loaded microbubble (IMB) to synthesize RTSL-IMBs. Sequentially taking advantage of acoustic-assisted early extravasation and thermo-triggered interstitium ultrafast drug release, RTSL-IMBs combine with ultrasound (US) and laser irradiation can advance drug delivery and bioavailability. In vitro experiments demonstrate that RTSL-IMBs associated with a two-step protocol (subsequently US irradiation for 1 min and laser irradiation for 5 min) can dramatically enhance the cellular uptake and bioavailability of DOX. In vivo fluorescence imaging studies reveal that the combination of RTSL-IMBs and US shows a 2.8-fold intratumoral drug accumulation increase at 0.5 h post-injection, while it will take 48 h to reach the same level of intratumoral drug accumulation for the RTSL-IMB group alone. Interestingly, the following localized application of a laser can further increase drug accumulation and slow tumor clearance. Histological analysis demonstrates that the combinational RTSL-IMBs, US and laser significantly improve the drug penetration distance and delivery efficiency in the tumor core. In this study, the acoustic/thermo-responsive hybrid system shows potential for advancing DOX chemotherapy in breast cancer cell MCF-7 xenograft nude mice.

Radiotherapy (RT) is one of the most widely used cancer treatments in the clinical setting, while hypoxia-associated resistance often occurs. Herein, a PEGylated TaOx-based oxygen-carrying nanoplatform was constructed for triple sensitizing tumor radiotherapy. The high-Z element based hollow mesoporous TaOx nanospheres were prepared following the in situ growth of ultrasmall CuS nanocrystals and then packaged with O2-saturated perfluoropentane (PFP). NIR laser-triggered mild hyperthermia would lead to the increase of intratumoral blood flow, together with the release of O2, the radiotherapeutic efficiency would be enhanced. Alternatively, radiant energy would be deposited inside the tumor by the Ta element, therefore triple sensitization of radiotherapy could be achieved. The in vivo studies showed that the as-prepared nanospheres could achieve almost total inhibition of tumor growth without obvious side effects, which provides new possibilities for multisensitizing tumor radiotherapy.

BACKGROUND:Ultrasound-targeted microbubble destruction (UTMD) has been shown to be a promising noninvasive technique to change the tumor circulation, thus providing a potential method to increase reactive oxygen species (ROS) levels in tumors by inducing tumor tissue ischemia-reperfusion (IR). In this study, we investigated the feasibility of local tumor IR through UTMD to enhance the anti-tumor efficacy of doxorubicin (DOX) chemotherapy.METHODS:UTMD was used to induce local tumor IR. After the major blood supply of the tumor was restored, DOX was intravenously injected into the tumor-bearing mice. The superoxide dismutase (SOD) and catalase (CAT) activity and ROS levels were examined, and the anti-tumor efficacy was evaluated.RESULTS:UTMD blocked the circulation to the tumor for 30 mins. Slow reperfusion began to occur after 30 mins, and major blood supply was restored after 1 hr. The blood perfusion of the tumor completely recovered at 2 hrs. The activity of SOD in the tumors was significantly decreased at 2 hrs and 1 day after IR treatment with or without DOX treatment. The CAT activity showed no obvious changes at 2 hrs after IR treatment, whereas a significant decrease was found after 1 day in both the IR and DOX/IR groups. Moreover, higher levels of ROS were produced in the IR group and IR/DOX group. In vivo anti-tumor study indicated that the local tumor IR strategy may significantly enhance the anti-tumor efficacy of DOX chemotherapy.CONCLUSION:UTMD provides a novel, simple and non-invasive technique for tumor IR. In combination with chemotherapy, UTMD may have high great potential to improve the anti-tumor efficacy of chemotherapeutic drugs.

Purpose: To examine diagnostic performance of qualitative shear wave elastography (SWE) for evaluation of status of axillary lymph nodes (ALN) in comparison with conventional ultrasonograghy (US) and quantitative SWE parameters. Methods: A total of 118 patients were enrolled, who were all scheduled for breast cancer surgery and core needle biopsy. Conventional US and SWE were performed before biopsy. Based on qualitative evaluation of each ALN, the SWE images were classified into four color patterns: Color Pattern 1: homogeneous; Color Pattern 2: filling defect within lymph node (LN); Color Pattern 3: homogeneous within LN with a localized colored area at the margin; and Color Pattern 4: filling defect within LN with a localized colored area at the margin. The diagnostic performances of the three methods were compared. Results: There were 60 metastatic nodes and 61 benign nodes in the 121 ALNs. Benign ALNs were presented as Color Pattern 1 while metastatic ALNs usually were presented as Color Pattern 2 to 4 (p < 0.05). The AUC of qualitative SWE classification was 0.983, higher than that of quantitative SWE parameters and conventional US (p< 0.05). The highest diagnostic performance, with AUC of 0.998, could be achieved if both conventional US and qualitative SWE were applied. Conclusion: The qualitative SWE classification of ALNs proposed in our study exhibited better diagnostic performance than quantitative SWE parameters and conventional US, especially for differentiating metastatic ALNs from benign reactive ALNs. More accurate diagnosis could be reached with this new method and unnecessary biopsy might be avoided in the meantime.

The application of microbubble (MB)-assisted ultrasound (US) can combine the advantages of real-time imaging and targeted drug delivery. However, the drug loading capacity of MB is limited restricting its application in antitumor procedure. In contrast, nanoparticles (NPs) can carry drugs more efficiently, but adverse side effect induced by unspecific accumulation can not be ignored. Herein, we developed a dual-functionalized NP loaded MB to investigate its potential feasibility for tumor-targeted drug delivery. Firstly, we prepared NPs using heparin as backbone. Targeting ligand folate and cell-penetrating ligand Tat peptide were conjugated to the backbone to deliver paclitaxel (H-F-Tat-P NPs). Subsequently, the dual-functionalized NPs were incorporated with MBs via avidin-biotin linkage to fabricate H-F-Tat-P NPs loaded MBs (NPs-loaded MBs). The combined strategy can take profit of dual functionalities from NPs and sonoporation effect from MBs triggered by US. The prepared NPs have been characterized. The excellent cellular uptake of NPs were qualitative and quantitative analysis by flow cytometry and confocal microscope, the results indicated that it was attributed to not only dual functionalities but also US effect. Foremost, the NPs-loaded MBs combined with US exhibited significant cytotoxicity on both folate receptor (FR) overexpressing and deficiency cells. The combination of dual-functionalized NPs and MBs with US is expected to be a promising strategy for targeted anticancer drug delivery and ultrasound imaging simultaneously. (C) 2018 Elsevier B.V. All rights reserved.

Objective To prepare dual-targeted pH-sensitive DOX prodrug-microbubble complex and explore the characterization of complex with ultrasound as well as drug release in vitro . Methods Dual-targeted ligands ,cRGD and folate were conjugated with heparin using carbodiimide method ,and then the dual-targeted pH-sensitive DOX prodrug was synthesized by coupling DOX via a pH-sensitive hydrazone bond . The prodrug was combined with microbubbles to prepare complex by biotin-avidin system . The characterization of complex with/without ultrasound was investigated for size ,morphology and drug loaded capacity .In vitro drug release manner of complex with/without at different pH was analyzed . Results DOX content of the prodrug determined by UV Spectrophotometry was about 18 .9％ . Dynamic laser light scattering analysis( DLS) ,corresponding to transmission electron microscope( TEM ) findings ,revealed its inhomogeneous size distribution [ mean size ( 159 .7 ± 24 .5) nm and ( 1089 .0 ± 174 .9) nm ] . However ,the complex was dispersed into uniform fragment after ultrasound irradiation [ mean size ( 155 .9 ± 29 .8) nm , polymer dispersity index( PDI) 0 .22 ,Zeta potential - ( 20 .6 ± 3 .4) mV ] . The cumulative release rate of DOX from both complex and complex with ultrasound at pH 5 .0 were much faster than that at pH 7 .4 , displaying a pH-triggered release manner . Conclusions Dual-targeted pH-sensitive DOX prodrug-microbubble complex displays excellent drug release activity in acid environment . Uniform fragment and smaller particle size of complex could be achieved via ultrasound irradiation ,promoting DOX accumulation within tumor tissue and facilitating in vivo antitumor ability .

To investigate the vaporization threshold of perfluoropentane encapsulated lipids nanodroplets and their feasibilities in assessing radio-frequency ablation margins in vitro. The nanodroplets with lipids shell and perfluoropentane (PFP) kemel was prepared by a Homogenization/emulsion method of which the size and zeta potential were measured by dynamic light scattering instrument (DLS). The phase change of perfluoropentane (PFP) nanodroplets (PFPNDs) stimulated by temperature rise was observed with the inverted optical microscope and ultrasonography. Echo intensity of bubbles generation from nanodroplets was calculated at different temperature points to plot the temperature-intensity curve. The temperature at the margins of hyperechogenicity zones generated from radiofrequency ablation (RFA) under ultrasonic guidance in tissue-mimicking phantom were measured with a thermocouple probe. The mean diameter of PFPNDs was 479.9±18.8nm. A few bubbles were observed at 50°C for PFPNDs with the inverted optical microscope, while a mass of bubbles generated rapidly from PFPNDs at 60°C, partly coalesced and ruptured ,and most of them ruptured at a higher temperature. No bubbles were observed in the Control group. With the ultrasonography, weak echo intensity of bubbles generated from PFPND at 50°C was detected. And when temperature rose to 60°C for PFPNDs or higher, the echo intensity reached peak value. Echo intensity in the Control group was not enhanced. The hyperechogenicity zones in the tissue-mimicking phantom dispersed with PFPNDs presented as elliptic shape along with the RFA needle, and the mean temperature at margin was 60.4°C. Little hyperechogenicity zones irregularly occurred in the Control group. The prepared PFPNDs can vaporize at given temperature. PFPNDs, with the vaporization threshold closed to the temperature threshold of tissues coagulative necrosis caused by RFA, is potential to assess the radiofrequency ablation margins in real time.

In this study, we investigated the potential of a dual-targeted pH-sensitive doxorubicin prodrug-microbubble complex (DPMC) in ultrasound (US)-assisted antitumor therapy. The doxorubicin prodrug (DP) consists of a succinylated-heparin carrier conjugated with doxorubicin (DOX) via hydrazone linkage and decorated with dual targeting ligands, folate and cRGD peptide. Combination of microbubble (MB) and DP, generated via avidin-biotin binding, promoted intracellular accumulation and improved therapeutic efficiency assisted by US cavitation and sonoporation. Aggregates of prepared DP were observed with an inhomogeneous size distribution (average diameters: 149.6±29.8 nm and 1036.2±38.8 nm, PDI: 1.0) while DPMC exhibited a uniform distribution (average diameter: 5.804±2.1 μm), facilitating its usage for drug delivery. Notably, upon US exposure, DPMC was disrupted and aggregated DP dispersed into homogeneous small-sized nanoparticles (average diameter: 128.6±42.3 nm, PDI: 0.21). DPMC could target to angiogenic endothelial cells in tumor region via αvβ3-mediated recognition and subsequently facilitate its specific binding to tumor cells mediated via recognition of folate receptor (FR) after US exposure. In vitro experiments showed higher tumor specificity and killing ability of DPMC with US than free DOX and DP for breast cancer MCF-7 cells. Furthermore, significant accumulation and specificity for tumor tissues of DPMC with US were detected using in vivo fluorescence and ultrasound molecular imaging, indicating its potential to integrate tumor imaging and therapy. In particular, through inducing apoptosis, inhibiting cell proliferation and antagonizing angiogenesis, DPMC with US produced higher tumor inhibition rates than DOX or DPMC without US in MCF-7 xenograft tumor-bearing mice while inducing no obvious body weight loss. Our strategy provides an effective platform for the delivery of large-sized or aggregated particles to tumor sites, thereby extending their therapeutic applications in vivo.

We have fabricated a negative-charged nanoparticle (Heparin-Folate-Tat-Taxol NP, H-F-Tat-T NP) with dual ligands, tumor targeting ligand folate and cell-penetrating peptide Tat, to deliver taxol presenting great anticancer activity for sensitive cancer cells, while it fails to overcome multidrug resistance (MDR) in MCF-7/T cells (taxol-resistant breast cancer cells). Ultrasound (US) can increase the sensitivity of positive-charged NPs thereby making it possible to reverse MDR through inducing NPs' drug release. However, compared with the negative-charged NPs, positive-charged NPs may cause higher toxic effect. Hence, the combination of negative-charged NPs and US may be an efficient strategy for overcoming MDR. The conventional procedure to treat with NPs followed by US exposure possibly destruct multifunctional NPs resulting in its bioactivity inhibition. Herein, we have further improved the operating approach to eliminate US mechanical damage and keep the integrity of negative-charged NPs: cells are exposed to US with microbubbles (MBs) prior to the treatment of H-F-Tat-T NPs. Superior to the conventional method, US sonoporation affects the physiological property of cancer cells while preventing direct promotion of drug release from NPs. The results of the present study displayed that US in condition (1MHz, 10% duty cycle, duration of 80s, US intensity of 0.6W/cm2 and volume ratio of medium to MBs 20:1) combined with H-F-T-Tat-T NPs can achieve optimal reversal MDR effect in MCF-7/T cells. Mechanism study further disclosed that the individual effect of US was responsible for the enhancement of cell membrane permeability, inhibition of cell proliferation rate and down-regulation of MDR-related genes and proteins. Simultaneously, US sonoporation on resistant cancer cells indirectly increased the accumulation of NPs by inducing endosomal escape of negative-charged NPs. Taken together, the overcoming MDR ability for the combined strategy was achieved by the synergistic effect from individual function of NPs, physiological changes of resistant cancer cells and behavior changes of NPs caused by US.

目的 探讨青年与中老年乳腺癌患者的临床病理及超声表现.方法 对经手术病理证实的288例乳腺癌患者进行回顾性研究,分析青年组(≤35岁)和中老年组(＞35岁)乳腺癌临床病理及超声图像特征的差异.结果 青年组较中老年组癌肿组织学分级高(P=0.03)、肿瘤偏大(P＜0.01)、腋窝淋巴结易转移(P=0.03)、Ki-67表达增高(P=0.01)、三阴型多(P=0.04)、Luminal A型少(P=0.04)、肿块形状规则(P＜0.05)、后方回声无衰减或回声增强(P=0.01)、边缘无毛刺征(P＜0.01)及BI-RADS分类低(P=0.03).结论 与中老年比较,青年乳腺癌侵袭性强,预后差,但其超声表现偏向良性.