Objective To evaluate the application value of artificial intelligence (AI) in the quality control of mammographic images and explore its feasibility for improving image quality. Methods A retrospective analysis was conducted on 500 mammographic images from 125 female patients. These images were acquired in December 2024 at the Department of Radiology, Obstetrics and Gynecology Hospital of Tongji University by two junior technologists with qualification certificate for junior radiologic technologists and ≤ 2 months of independent operation. The reference standard was the evaluation by a panel of senior experts comprising one associate chief radiologist and two intermediate radiologic technologists with over 10 years of experience in mammography. The evaluation was based on nine criteria within an AI quality control system. The correlation and consistency (Kappa test) of the assessment outcomes were compared among the junior technologist group, the AI quality control group, and the panel group. Additionally, the differences in the proportions of high-, medium-, and low-quality images rated by the three groups were analyzed. After one month of AI assistance, 200 images from 50 additional patients acquired by the same two junior technologists in January 2025 were collected. The area under the receiver operating characteristic curve and 95% confidence interval (95%CI) were calculated for the assessment performance of the junior technologist group before and after AI assistance. Differences were compared using the DeLong test. Changes in the disqualification rates for the nine criteria were analyzed. Results In consistency analysis, the AI quality control group showed high consistency with the panel group across multiple key criteria, with Kappa values ranging from 0.41 to 1.00. In contrast, the consistency between the junior technologist group and the panel group was generally low, with Kappa values ranging from 0.13 to 0.49. In comparison of image quality classification, no significant differences were observed in image quality classification between the AI group and the panel group (P>0.05). However, the proportion of images rated as high quality by the junior technologist group was significantly higher, while the proportions rated as medium and low quality were significantly lower, compared to those rated by the panel group (P<0.05). After AI assistance, the area under the receiver operating characteristic curve for the assessment performance of junior technologist group increased significantly from 0.56 (95%CI: 0.51-0.61) to 0.91 (95%CI: 0.87-0.94) (P<0.001). The disqualification rates for skin folds, incomplete inclusion of the pectoralis major muscle, and nipple not in profile decreased significantly. Conclusion AI demonstrates assessment capability comparable to the panel group in mammographic quality control. AI effectively enhances the quality of images acquired by junior technologists through real-time, objective feedback.

Objective To evaluate the application value of artificial intelligence (AI) in the quality control of mammographic images and explore its feasibility for improving image quality. Methods A retrospective analysis was conducted on 500 mammographic images from 125 female patients. These images were acquired in December 2024 at the Department of Radiology, Obstetrics and Gynecology Hospital of Tongji University by two junior technologists with qualification certificate for junior radiologic technologists and ≤ 2 months of independent operation. The reference standard was the evaluation by a panel of senior experts comprising one associate chief radiologist and two intermediate radiologic technologists with over 10 years of experience in mammography. The evaluation was based on nine criteria within an AI quality control system. The correlation and consistency (Kappa test) of the assessment outcomes were compared among the junior technologist group, the AI quality control group, and the panel group. Additionally, the differences in the proportions of high-, medium-, and low-quality images rated by the three groups were analyzed. After one month of AI assistance, 200 images from 50 additional patients acquired by the same two junior technologists in January 2025 were collected. The area under the receiver operating characteristic curve and 95% confidence interval (95%CI) were calculated for the assessment performance of the junior technologist group before and after AI assistance. Differences were compared using the DeLong test. Changes in the disqualification rates for the nine criteria were analyzed. Results In consistency analysis, the AI quality control group showed high consistency with the panel group across multiple key criteria, with Kappa values ranging from 0.41 to 1.00. In contrast, the consistency between the junior technologist group and the panel group was generally low, with Kappa values ranging from 0.13 to 0.49. In comparison of image quality classification, no significant differences were observed in image quality classification between the AI group and the panel group (P>0.05). However, the proportion of images rated as high quality by the junior technologist group was significantly higher, while the proportions rated as medium and low quality were significantly lower, compared to those rated by the panel group (P<0.05). After AI assistance, the area under the receiver operating characteristic curve for the assessment performance of junior technologist group increased significantly from 0.56 (95%CI: 0.51-0.61) to 0.91 (95%CI: 0.87-0.94) (P<0.001). The disqualification rates for skin folds, incomplete inclusion of the pectoralis major muscle, and nipple not in profile decreased significantly. Conclusion AI demonstrates assessment capability comparable to the panel group in mammographic quality control. AI effectively enhances the quality of images acquired by junior technologists through real-time, objective feedback.

Objective To evaluate the application value of artificial intelligence (AI) in the quality control of mammographic images and explore its feasibility for improving image quality. Methods A retrospective analysis was conducted on 500 mammographic images from 125 female patients. These images were acquired in December 2024 at the Department of Radiology, Obstetrics and Gynecology Hospital of Tongji University by two junior technologists with qualification certificate for junior radiologic technologists and ≤ 2 months of independent operation. The reference standard was the evaluation by a panel of senior experts comprising one associate chief radiologist and two intermediate radiologic technologists with over 10 years of experience in mammography. The evaluation was based on nine criteria within an AI quality control system. The correlation and consistency (Kappa test) of the assessment outcomes were compared among the junior technologist group, the AI quality control group, and the panel group. Additionally, the differences in the proportions of high-, medium-, and low-quality images rated by the three groups were analyzed. After one month of AI assistance, 200 images from 50 additional patients acquired by the same two junior technologists in January 2025 were collected. The area under the receiver operating characteristic curve and 95% confidence interval (95%CI) were calculated for the assessment performance of the junior technologist group before and after AI assistance. Differences were compared using the DeLong test. Changes in the disqualification rates for the nine criteria were analyzed. Results In consistency analysis, the AI quality control group showed high consistency with the panel group across multiple key criteria, with Kappa values ranging from 0.41 to 1.00. In contrast, the consistency between the junior technologist group and the panel group was generally low, with Kappa values ranging from 0.13 to 0.49. In comparison of image quality classification, no significant differences were observed in image quality classification between the AI group and the panel group (P>0.05). However, the proportion of images rated as high quality by the junior technologist group was significantly higher, while the proportions rated as medium and low quality were significantly lower, compared to those rated by the panel group (P<0.05). After AI assistance, the area under the receiver operating characteristic curve for the assessment performance of junior technologist group increased significantly from 0.56 (95%CI: 0.51-0.61) to 0.91 (95%CI: 0.87-0.94) (P<0.001). The disqualification rates for skin folds, incomplete inclusion of the pectoralis major muscle, and nipple not in profile decreased significantly. Conclusion AI demonstrates assessment capability comparable to the panel group in mammographic quality control. AI effectively enhances the quality of images acquired by junior technologists through real-time, objective feedback.

Traditional Chinese medicine (TCM) demonstrates distinctive advantages in disease prevention and treatment. However, analyzing its biological mechanisms through the modern medical research paradigm of "single drug, single target" presents significant challenges due to its holistic approach. Network pharmacology and its core theory of network targets connect drugs and diseases from a holistic and systematic perspective based on biological networks, overcoming the limitations of reductionist research models and showing considerable value in TCM research. Recent integration of network target computational and experimental methods with artificial intelligence (AI) and multi-modal multi-omics technologies has substantially enhanced network pharmacology methodology. The advancement in computational and experimental techniques provides complementary support for network target theory in decoding TCM principles. This review, centered on network targets, examines the progress of network target methods combined with AI in predicting disease molecular mechanisms and drug-target relationships, alongside the application of multi-modal multi-omics technologies in analyzing TCM formulae, syndromes, and toxicity. Looking forward, network target theory is expected to incorporate emerging technologies while developing novel approaches aligned with its unique characteristics, potentially leading to significant breakthroughs in TCM research and advancing scientific understanding and innovation in TCM.
Artificial Intelligence ; Medicine, Chinese Traditional ; Humans ; Network Pharmacology/methods* ; Drugs, Chinese Herbal/pharmacology* ; Animals ; Multiomics

Stroke is a global disease that seriously threatens human life. The pathological mechanisms of ischemic stroke include neuroinflammation, oxidative stress, and the destruction of blood vessels at the lesion site. Here, a biocompatible in situ hydrogel platform was designed to target multiple pathogenic mechanisms post-stroke, including anti-inflammation, anti-oxidant, and promotion of angiogenesis. Double-crosslinked responsive multifunctional hydrogels could quickly respond to the pathological microenvironment of the ischemic damage site and mediate the delivery of nitric oxide (NO) and ISO-1 (inhibitor of macrophage migration inhibitory factor, MIF). The hydrogel demonstrated good biocompatibility and could scavenge reactive oxygen species (ROS) and inflammatory cytokines, such as interleukin-6 (IL-6), interleukin-10 (IL-10), and MIF. In a mouse stroke model, hydrogels, when situated within the microenvironment of cerebral infarction characterized by weak acidity and elevated ROS release, would release anti-inflammatory nanoparticles rapidly that exert an anti-inflammatory effect. Concurrently, NO was sustained release to facilitate angiogenesis and provide neuroprotective effects. Neurological function was significantly improved in treated mice as assessed by the modified neurological severity score, rotarod test, and open field test. These findings indicate that the designed hydrogel held promise for sustained delivery of NO and ISO-1 to alleviate cerebral ischemic injury by responding to the brain's pathological microenvironment.

Inflammatory bowel disease (IBD) is an autoimmune disorder involving complex immune regulation, where balancing localized and systemic immunosuppression is a key challenge. This study aimed to enhance the therapeutic efficacy by engineering the probiotic Escherichia coli Nissle 1917 (EcN). We removed endogenous plasmids pMUT1 and pMUT2 from wild-type EcN and expressed the mPD-L1 (19‒238 aa)-mFc fusion protein on the bacterial surface using a cytolysin A (ClyA) fragment. This modification stabilized mPD-L1 (19‒238 aa) protein expression and promoted its recruitment to outer membrane vesicles (OMVs). The engineered strain, EcNΔ_pMUT1/2-ClyA-mPD-L1-mFc (EcN-ePD-L1-mFc), features conditional ePD-L1-mFc expression under the araBAD promoter, enhancing gut-targeted release and reducing systemic side effects. This strain improved treatment targeting and efficiency by enabling direct ePD-L1-mFc interaction with immune cells at inflammation sites. OMVs from this strain induced Treg proliferation, inhibited effector T cell proliferation in vitro, and significantly improved intestinal inflammation and colonic epithelial barrier repair in vivo. Additionally, the bacterium restored intestinal microbiota balance, increasing Lactobacillaceae and reducing Bacteroides. This study highlights the engineered bacterium's potential for targeted intestinal immune modulation and offers a novel local IBD treatment approach with promising clinical prospects.

General anesthesia, pivotal for surgical procedures, requires precise depth monitoring to mitigate risks ranging from intraoperative awareness to postoperative cognitive impairments. Traditional assessment methods, relying on physiological indicators or behavioral responses, fall short of accurately capturing the nuanced states of unconsciousness. This study introduces a machine learning-based approach to decode anesthesia depth, leveraging EEG data across different anesthesia states induced by propofol and esketamine in rats. Our findings demonstrate the model's robust predictive accuracy, underscored by a novel intra-subject dataset partitioning and a 5-fold cross-validation method. The research diverges from conventional monitoring by utilizing anesthetic infusion rates as objective indicators of anesthesia states, highlighting distinct EEG patterns and enhancing prediction accuracy. Moreover, the model's ability to generalize across individuals suggests its potential for broad clinical application, distinguishing between anesthetic agents and their depths. Despite relying on rat EEG data, which poses questions about real-world applicability, our approach marks a significant advance in anesthesia monitoring.
Animals ; Machine Learning ; Electroencephalography/methods* ; Ketamine/administration & dosage* ; Rats ; Male ; Propofol/administration & dosage* ; Rats, Sprague-Dawley ; Anesthesia, General/methods* ; Brain/physiology* ; Intraoperative Neurophysiological Monitoring/methods*

Learning-associated functional plasticity at hippocampal synapses remains largely unexplored. Here, in a single session of reward-based trace conditioning, we examine learning-induced synaptic plasticity in the dorsal CA1 hippocampus (dCA1). Local field-potential recording combined with selective optogenetic inhibition first revealed an increase of dCA1 synaptic responses to the conditioned stimulus (CS) induced during conditioning at both Schaffer collaterals to the stratum radiatum (Rad) and temporoammonic input to the lacunosum moleculare (LMol). At these dCA1 inputs, synaptic potentiation of CS-responding excitatory synapses was further demonstrated by locally blocking NMDA receptors during conditioning and whole-cell recording sensory-evoked synaptic responses in dCA1 neurons from naive animals. An overall similar time course of the induction of synaptic potentiation was found in the Rad and LMol by multiple-site recording; this emerged later and saturated earlier than conditioned behavioral responses. Our experiments demonstrate a cued memory-associated dCA1 synaptic plasticity induced at both Schaffer collaterals and temporoammonic pathways.
Animals ; CA1 Region, Hippocampal/physiology* ; Male ; Association Learning/physiology* ; Neuronal Plasticity/physiology* ; Cues ; Memory/physiology* ; Synapses/physiology* ; Conditioning, Classical/physiology* ; Excitatory Postsynaptic Potentials/physiology* ; Receptors, N-Methyl-D-Aspartate/antagonists & inhibitors* ; Rats ; Optogenetics

Drug development remains a critical issue in the field of biomedicine. With the rapid advancement of information technologies such as artificial intelligence (AI) and the advent of the big data era, AI-assisted drug development has become a new trend, particularly in predicting drug-target associations. To address the challenge of drug-target prediction, AI-driven models have emerged as powerful tools, offering innovative solutions by effectively extracting features from complex biological data, accurately modeling molecular interactions, and precisely predicting potential drug-target outcomes. Traditional machine learning (ML), network-based, and advanced deep learning architectures such as convolutional neural networks (CNNs), graph convolutional networks (GCNs), and transformers play a pivotal role. This review systematically compiles and evaluates AI algorithms for drug- and drug combination-target predictions, highlighting their theoretical frameworks, strengths, and limitations. CNNs effectively identify spatial patterns and molecular features critical for drug-target interactions. GCNs provide deep insights into molecular interactions via relational data, whereas transformers increase prediction accuracy by capturing complex dependencies within biological sequences. Network-based models offer a systematic perspective by integrating diverse data sources, and traditional ML efficiently handles large datasets to improve overall predictive accuracy. Collectively, these AI-driven methods are transforming drug-target predictions and advancing the development of personalized therapy. This review summarizes the application of AI in drug development, particularly in drug-target prediction, and offers recommendations on models and algorithms for researchers engaged in biomedical research. It also provides typical cases to better illustrate how AI can further accelerate development in the fields of biomedicine and drug discovery.