A pH-responsive nanoplatform with dual-modality imaging for enhanced cancer phototherapy and diagnosis of lung metastasis.

To address the limitations of traditional photothermal therapy (PTT)/ photodynamic therapy (PDT) and real-time cancer metastasis detection, a pH-responsive nanoplatform (NP) with dual-modality imaging capability was rationally designed. Herein, 1 H,1 H-undecafluorohexylamine (PFC), served as both an oxygen carrier and a 19F magnetic resonance imaging (MRI) probe, and photosensitizer indocyanine green (ICG) were grafted onto the pH-responsive peptide hexahistidine (H6) to form H6-PFC-ICG (HPI). Subsequently, the heat shock protein 90 inhibitor, gambogic acid (GA), was incorporated into hyaluronic acid (HA) modified HPI (HHPI), yielding the ultimate HHPI@GA NPs. Upon self-assembly, HHPI@GA NPs passively accumulated in tumor tissues, facilitating oxygen release and HA-mediated cell uptake. Once phagocytosed by lysosomes, protonation of H6 was triggered due to the low pH, resulting in the release of GA. With near-infrared laser irradiation, GA-mediated decreased HSP90 expression and PFC-mediated increased ROS generation amplified the PTT/PDT effect of HHPI@GA, leading to excellent in vitro and in vivo anticancer efficacies. Additionally, the fluorescence and 19F MRI dual-imaging capabilities of HHPI@GA NPs enabled effective real-time primary cancer and lung metastasis monitoring. This work offers a novel approach for enhanced cancer phototherapy, as well as precise cancer diagnosis.

[Review of clinical effects of disk-up sinus reamer(DSR)-based internal sinus floor elevation with implantation in 10 years].

PURPOSE: To analyze the bone remodeling around the implant 10 years after disk-up sinus reamer(DSR)-based internal sinus floor elevation with implantation and to investigate the influence of different factors on implant retention. METHODS: The clinical and imaging data of patients undergoing DSR-based sinus floor elevation with simultaneous implantation were collected from the Department of Dental Implantology, Affiliated Hospital of Qingdao University from January 2008 to December 2011. Panoramic film and CBCT were used to measure the changes of bone mass around implant in different periods. Kaplan-Meier and Log-rank tests were used to analyze the effects of different factors on implant retention with SPSS 26.0 software package. RESULTS: The study included 98 patients with a total of 128 implants. During the follow-up of 0-168 months, 7 implants failed, and the remaining formed good osseointegration and functioned, with a 10-year cumulative retention rate of 94.53%. The height of bone formation was (0.29±0.15) mm at the top and (2.74±0.66) mm in the sinus of 75 implant sites with complete imaging data obtained ten years after surgery. Kaplan-Meier and Log-rank tests showed that 8 factors including initial bone height, elevated bone height, mucosal perforation, implant length, implant torsion, diabetes, smoking and periodontitis had significant effects on implant retention. CONCLUSIONS: The DSR-based internal sinus floor elevation with implantation is a reliable and stable bone augmentation operation for vertical bone defect in maxillary posterior region, with a 10-year cumulative retention rate of no less than 94%. Initial bone height, elevated bone height, mucosal perforation, implant length, implant torsion, diabetes, smoking and periodontitis are the important factors affecting the long-term retention rate of implants.

Diagnostic accuracy of interleukin-6 in multiple diseases: An umbrella review of meta-analyses.

Objective: This review aims to conduct a comprehensive study of the diagnostic accuracy of interleukin-6 (IL-6) for multiple diseases by utilizing existing systematic reviews and meta-analyses. Methods: We performed a thorough search of Embase, Web of Science, PubMed, and Cochrane Database of Systematic Reviews up to April 2023 to gather meta-analyses that investigate the diagnostic accuracy of IL-6. To assess the methodological quality of the studies, we employed the Assessing the Methodological Quality of Systematic Reviews-2 and Grading of Recommendations, Assessment, Development and Evaluation criteria. Results: We included 34 meta-analyses out of the 3024 articles retrieved from the search. These meta-analyses covered 9 categories of diseases of the International Classification of Diseases-11. Studies rated as "Critically Low" or "Very Low" in the quality assessment process were excluded, resulting in a total of 6 meta-analyses that encompassed sepsis, colorectal cancer, tuberculous pleural effusion (TPE), endometriosis, among others. Among these diseases, IL-6 demonstrated a relatively high diagnostic potential in accurately identifying TPE and endometriosis. Conclusions: IL-6 exhibited favorable diagnostic accuracy across multiple diseases, suggesting its potential as a reliable diagnostic biomarker in the near future. Substantial evidence supported its high diagnostic accuracy, particularly in the cases of TPE and endometriosis.

Transformaer-based model for lung adenocarcinoma subtypes.

BACKGROUND: Lung cancer has the highest morbidity and mortality rate among all types of cancer. Histological subtypes serve as crucial markers for the development of lung cancer and possess significant clinical values for cancer diagnosis, prognosis, and prediction of treatment responses. However, existing studies only dichotomize normal and cancerous tissues, failing to capture the unique characteristics of tissue sections and cancer types. PURPOSE: Therefore, we have pioneered the classification of lung adenocarcinoma (LAD) cancer tissues into five subtypes (acinar, lepidic, micropapillary, papillary, and solid) based on section data in whole-slide image sections. In addition, a novel model called HybridNet was designed to improve the classification performance. METHODS: HybridNet primarily consists of two interactive streams: a Transformer and a convolutional neural network (CNN). The Transformer stream captures rich global representations using a self-attention mechanism, while the CNN stream extracts local semantic features to optimize image details. Specifically, during the dual-stream parallelism, the feature maps of the Transformer stream as weights are weighted and summed with those of the CNN stream backbone; at the end of the parallelism, the respective final features are concatenated to obtain more discriminative semantic information. RESULTS: Experimental results on a private dataset of LAD showed that HybridNet achieved 95.12% classification accuracy, and the accuracy of five histological subtypes (acinar, lepidic, micropapillary, papillary, and solid) reached 94.5%, 97.1%, 94%, 91%, and 99% respectively; the experimental results on the public BreakHis dataset show that HybridNet achieves the best results in three evaluation metrics: accuracy, recall and F1-score, with 92.40%, 90.63%, and 91.43%, respectively. CONCLUSIONS: The process of classifying LAD into five subtypes assists pathologists in selecting appropriate treatments and enables them to predict tumor mutation burden (TMB) and analyze the spatial distribution of immune checkpoint proteins based on this and other clinical data. In addition, the proposed HybridNet fuses CNN and Transformer information several times and is able to improve the accuracy of subtype classification, and also shows satisfactory performance on public datasets with some generalization ability.

In Situ Remodeling of Efferocytosis via Lesion-Localized Microspheres to Reverse Cartilage Senescence.

Efferocytosis, an intrinsic regulatory mechanism to eliminate apoptotic cells, will be suppressed due to the delayed apoptosis process in aging-related diseases, such as osteoarthritis (OA). In this study, cartilage lesion-localized hydrogel microspheres are developed to remodel the in situ efferocytosis to reverse cartilage senescence and recruit endogenous stem cells to accelerate cartilage repair. Specifically, aldehyde- and methacrylic anhydride (MA)-modified hyaluronic acid hydrogel microspheres (AHM), loaded with pro-apoptotic liposomes (liposomes encapsulating ABT263, A-Lipo) and PDGF-BB, namely A-Lipo/PAHM, are prepared by microfluidic and photo-cross-linking techniques. By a degraded porcine cartilage explant OA model, the in situ cartilage lesion location experiment illustrated that aldehyde-functionalized microspheres promote affinity for degraded cartilage. In vitro data showed that A-Lipo induced apoptosis of senescent chondrocytes (Sn-chondrocytes), which can then be phagocytosed by the efferocytosis of macrophages, and remodeling efferocytosis facilitated the protection of normal chondrocytes and maintained the chondrogenic differentiation capacity of MSCs. In vivo experiments confirmed that hydrogel microspheres localized to cartilage lesion reversed cartilage senescence and promoted cartilage repair in OA. It is believed this in situ efferocytosis remodeling strategy can be of great significance for tissue regeneration in aging-related diseases.

Multi-omics analyses uncover metabolic signatures and gene expression profiles of interstitial cystitis/bladder pain syndrome.

BACKGROUND AND OBJECTIVE: We explore molecular and metabolic pathways involved in interstitial cystitis (IC) with integrating multi-omics analysis for identifying potential diagnostic and therapeutic targets. METHODS: Mouse models of IC/bladder pain syndrome (BPS) were established by intraperitoneal injection of cyclophosphamide and bladder tissue samples were collected for metabolomics and transcriptome analysis. RESULTS: We found a total of 82 and 145 differential metabolites in positive ion modes and negative ion modes, respectively. Glycerophospholipid metabolism, choline metabolism in cancer, and nucleotide metabolism pathways were significantly enriched in the IC/BPS group. Transcriptome analysis demonstrated that 1069 upregulated genes and 1087 downregulated genes were detected. Importantly, the stronger enrichment for cell cycle pathway was observed in IC/BPS than that in normal bladder tissue, which may be involved in the process of bladder remodeling. Moreover, the inflammatory response and inflammatory factors related pathways were enriched in the IC/BPS group. CONCLUSIONS: Our findings provide critical directions for further exploration of the molecular pathology underlying IC/BPS.

Efficacy and Safety of Adeno-Associated Virus-Based Clinical Gene Therapy for Hemophilia: A Systematic Review and Meta-Analysis.

Clinical trials of adeno-associated virus (AAV)-based gene therapy have made remarkable progress in recent years. We aimed to perform a systematic review and meta-analysis of the literature to assess the efficacy and safety of AAV-based gene therapy for hemophilia. We systematically searched the Web of Science, Embase, PubMed, and the Cochrane Database of Systematic Reviews databases, for clinical trials involving patients diagnosed with hemophilia and treated with AAV-mediated gene therapy. Data on the annualized bleeding rate (ABR), annualized infusion rate (AIR), the incidence of treatment-related adverse events (TRAEs), severe adverse events (SAEs), and alanine aminotransferase (ALT) elevation were extracted as our outcomes. A total of 12 articles from 11 clinical trials were selected from 868 articles for meta-analysis. Pooled analyses showed that AAV-based gene therapy in hemophilia patients reduced the number of bleeding events and the number of factor infusion events by an approximate average of 7 per year and 103 per year, respectively. Eighty percent, 18%, and 63% of hemophilia patients had elevated TRAE, SAE, and ALT levels, respectively. Moreover, subgroup analysis found a significant reduction in ABR and AIR 2-3 years after the therapy. Additional findings that were not pooled including coagulation factor activity are presented in the accompanying tables. Our analysis supported the efficacy and safety of AAV-mediated gene therapy for hemophilia, providing evidence for its application as a therapeutic option for widespread clinical use in hemophilia patients in the future.

Stem Cells Expansion Vector via Bioadhesive Porous Microspheres for Accelerating Articular Cartilage Regeneration.

Stem cell tissue engineering is a potential treatment for osteoarthritis. However, the number of stem cells that can be delivered, loss of stem cells during injection, and migration ability of stem cells limit applications of traditional stem cell tissue engineering. Herein, kartogenin (KGN)-loaded poly(lactic-co-glycolic acid) (PLGA) porous microspheres is first engineered via emulsification, and then anchored with chitosan through the amidation reaction to develop a new porous microsphere (PLGA-CS@KGN) as a stem cell expansion vector. Following 3D co-culture of the PLGA-CS@KGN carrier with mesenchymal stem cells (MSCs), the delivery system is injected into the capsule cavity in situ. In vivo and in vitro experiments show that PLGA-CS microspheres have a high cell-carrying capacity up to 1 × 104 mm-3 and provide effective protection of MSCs to promote their controlled release in the osteoarthritis microenvironment. Simultaneously, KGN loaded inside the microspheres effectively cooperated with PLGA-CS to induce MSCs to differentiate into chondrocytes. Overall, these findings indicate that PLGA-CS@KGN microspheres held high cell-loading ability, adapt to the migration and expansion of cells, and promote MSCs to express markers associated with cartilage repair. Thus, PLGA-CS@KGN can be used as a potential stem cell carrier for enhancing stem cell therapy in osteoarthritis treatment.

Semantic guidance network for video captioning.

video captioning is a more challenging task that aims to generate abundant natural language descriptions, and it has become a promising direction for artificial intelligence. However, most existing methods are prone to ignore the problems of visual information redundancy and scene information omission due to the limitation of the sampling strategies. To address this problem, a semantic guidance network for video captioning is proposed. More specifically, a novel scene frame sampling strategy is first proposed to select key scene frames. Then, the vision transformer encoder is applied to learn visual and semantic information with a global view, alleviating information loss of modeling long-range dependencies caused in the encoder's hidden layer. Finally, a non-parametric metric learning module is introduced to calculate the similarity value between the ground truth and the prediction result, and the model is optimized in an end-to-end manner. Experiments on the benchmark MSR-VTT and MSVD datasets show that the proposed method can effectively improve the description accuracy and generalization ability.

Accuracy of the Yakebot dental implant robotic system versus fully guided static computer-assisted implant surgery template in edentulous jaw implantation: A preliminary clinical study.

AIMS: To compare the accuracy of the Yakebot dental implant robotic system with that of fully guided static computer-assisted implant surgery (CAIS) template in edentulous implantation. MATERIALS AND METHODS: Thirteen patients with edentulous were recruited and divided into two groups: the Yake robotic system group (experimental) (n = 5) and the CAIS group (control) (n = 8). Postoperative cone-beam computed tomography (CBCT) was performed immediately, and the 3-dimensional positions of implants were obtained and compared with that in the preoperative design. The comparison showed platform, apical, depth, and angular deviations. A value of p < 0.05 was considered statistically significant. RESULTS: A total of 84 implants (36 in the robotic group and 48 in the CAIS group) were placed. The mean deviation at the implant platform, apex, depth, and angle in the CAIS group was 1.37 ± 0.72 mm, 1.28 ± 0.68 mm, 0.88 ± 0.47 mm, and 3.47 ± 2.02°, respectively. However, the mean deviation at the implant platform, apex, depth, and angle in the robotic group was 0.65 ± 0.25 mm, 0.65 ± 0.22 mm, 0.49 ± 0.24 mm, and 1.43 ± 1.18°, respectively. Significant differences in the four types of deviation (p < 0.05) between the two groups were observed. CONCLUSION: The accuracy of robotic system in edentulous implant placement was superior to that of the CAIS template, suggesting that robotic system is more accurate, safe, and flexible, can be considered a promising treatment in clinical practice.

mRNA nanodelivery systems: targeting strategies and administration routes.

With the great success of coronavirus disease (COVID-19) messenger ribonucleic acid (mRNA) vaccines, mRNA therapeutics have gained significant momentum for the prevention and treatment of various refractory diseases. To function efficiently in vivo and overcome clinical limitations, mRNA demands safe and stable vectors and a reasonable administration route, bypassing multiple biological barriers and achieving organ-specific targeted delivery of mRNA. Nanoparticle (NP)-based delivery systems representing leading vector approaches ensure the successful intracellular delivery of mRNA to the target organ. In this review, chemical modifications of mRNA and various types of advanced mRNA NPs, including lipid NPs and polymers are summarized. The importance of passive targeting, especially endogenous targeting, and active targeting in mRNA nano-delivery is emphasized, and different cellular endocytic mechanisms are discussed. Most importantly, based on the above content and the physiological structure characteristics of various organs in vivo, the design strategies of mRNA NPs targeting different organs and cells are classified and discussed. Furthermore, the influence of administration routes on targeting design is highlighted. Finally, an outlook on the remaining challenges and future development toward mRNA targeted therapies and precision medicine is provided.

Revolutionizing cancer treatment: The power of cell-based drug delivery systems.

Intravenous administration of drugs is a widely used cancer therapy approach. However, the efficacy of these drugs is often hindered by various biological barriers, including circulation, accumulation, and penetration, resulting in poor delivery to solid tumors. Recently, cell-based drug delivery platforms have emerged as promising solutions to overcome these limitations. These platforms offer several advantages, including prolonged circulation time, active targeting, controlled release, and excellent biocompatibility. Cell-based delivery systems encompass cell membrane coating, intracellular loading, and extracellular backpacking. These innovative platforms hold the potential to revolutionize cancer diagnosis, monitoring, and treatment, presenting a plethora of opportunities for the advancement and integration of pharmaceuticals, medicine, and materials science. Nevertheless, several technological, ethical, and financial barriers must be addressed to facilitate the translation of these platforms into clinical practice. In this review, we explore the emerging strategies to overcome these challenges, focusing specifically on the functions and advantages of cell-mediated drug delivery in cancer treatment.

Using optimal transport theory to optimize a deep convolutional neural network microscopic cell counting method.

Medical image processing has become increasingly important in recent years, particularly in the field of microscopic cell imaging. However, accurately counting the number of cells in an image can be a challenging task due to the significant variations in cell size and shape. To tackle this problem, many existing methods rely on deep learning techniques, such as convolutional neural networks (CNNs), to count cells in an image or use regression counting methods to learn the similarities between an input image and a predicted cell image density map. In this paper, we propose a novel approach to monitor the cell counting process by optimizing the loss function using the optimal transport method, a rigorous measure to calculate the difference between the predicted count map and the dot annotation map generated by the CNN. We evaluated our algorithm on three publicly available cell count benchmarks: the synthetic fluorescence microscopy (VGG) dataset, the modified bone marrow (MBM) dataset, and the human subcutaneous adipose tissue (ADI) dataset. Our method outperforms other state-of-the-art methods, achieving a mean absolute error (MAE) of 2.3, 4.8, and 13.1 on the VGG, MBM, and ADI datasets, respectively, with smaller standard deviations. By using the optimal transport method, our approach provides a more accurate and reliable cell counting method for medical image processing.

RNA m6A modification in prostate cancer: A new weapon for its diagnosis and therapy.

Prostate cancer (PCa) is the most common malignant tumor and the second leading cause of cancer-related mortality in men worldwide. Despite significant advances in PCa therapy, the underlying molecular mechanisms have yet to be fully elucidated. Recently, epigenetic modification has emerged as a key player in tumor progression, and RNA-based N6-methyladenosine (m6A) epigenetic modification was found to be crucial. This review summarizes comprehensive state-of-art mechanisms underlying m6A modification, its implication in the pathogenesis, and advancement of PCa in protein-coding and non-coding RNA contexts, its relevance to PCa immunotherapy, and the ongoing clinical trials for PCa treatment. This review presents potential m6A-based targets and paves a new avenue for diagnosing and treating PCa, providing new guidelines for future related research through a systematic review of previous results.

Cationic Phosphinidene as a Versatile P1 Building Block: [LC-P]+ Transfer from Phosphonio-Phosphanides [LC-P-PR3]+ and Subsequent LC Replacement Reactions (LC = N-Heterocyclic Carbene).

Cationic imidazoliumyl(phosphonio)-phosphanides [LC-P-PR3]+ (1a-e+, LC = 4,5-dimethyl-1,3-diisopropylimidazolium-2-yl; R = alkyl, aryl) are obtained via the nucleophilic fragmentation of tetracationic tetraphosphetane [(LC-P)4][OTf]4 (2[OTf]4) with tertiary phosphanes. They act as [LC-P]+ transfer reagents in phospha-Wittig-type reactions, when converted with various thiocarbonyls, giving unprecedented cationic phosphaalkenes [LC-P═CR2]+ (5a-f[OTf]) or phosphanides [LC-P-CR(NR2')]+ (6a-d[OTf]). Theoretical calculations suggest that three-membered cyclic thiophosphiranes are crucial intermediates of this reaction. To test this hypothesis, treatment of [LC-P-PPh3]+ with phosphaalkenes, that are isolobal to thioketones, permits the isolation of diphosphirane salts 11a,b[OTf]. Furthermore, preliminary studies suggest that the cationic phosphaalkene [LC-P═CPh2]+ may be employed to access rare examples of η2-P═C π-complexes with Pd0 and Pt0 when treated with [Pd(PPh3)4] and [Pt(PPh3)3] for which analogous complexes of neutral phosphaalkenes are scarce. The versatility of [LC-P]+ as a valuable P1 building block was showcased in substitution reactions of the transferred LC-substituent using nucleophiles. This is demonstrated through the reactions of 5a[OTf] and 6c[OTf] with Grignard reagents and KNPh2, providing a convenient, high-yielding access to MesP═CPh2 (16) and otherwise difficult-to-synthesize 1,3-diphosphetane 17 and P-aminophosphaalkenes.

Safety and efficacy of intra-articular injection of platelet-rich plasma for the treatment of ankle osteoarthritis: a systematic review and meta-analysis.

PURPOSE: To evaluate the safety and efficacy of platelet-rich plasma (PRP) intra-articular injective treatments for ankle osteoarthritis (OA). METHODS: A systematic literature search was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines in PubMed, Scopus, Embase, Google Scholar, and the Cochrane library until May 2022. Both randomized and non-randomized studies were included with the assessment of the risk of bias. We recorded the participant's age, gender, type of PRP, injection volume, the kit used, and activating agent. We subsequently assessed the short-term and long-term efficacy of PRP using the functional scores and visual analog scale (VAS). RESULTS: We included four studies with a total of 127 patients, with a mean age of 56.1 years. 47.2% were male (60/127), according to eligibility criteria. There were three cohort studies and one randomized controlled trial (RCT) study, and no study reported severe adverse events. All included studies used the Leukocyte-poor PRP. Short-term follow-up results suggested significant improvement of the American Orthopaedic Foot and Ankle Society (AOFAS) score in the PRP injection group compared to the control group (n = 87 patients; MD: 6.94 [95% CI: 3.59, 10.29]; P < 0.01). Consistently, there was a statistical difference in AOFAS score between PRP injection and control groups in the final follow-up (≥ 6 months) (n = 87 patients; MD: 9.63 [95% CI: 6.31, 12.94]; P < 0.01). Furthermore, we found a significant reduction in VAS scores in the PRP groups at both the short-term follow-up (n = 59 patients; MD, - 1.90 [95% CI, - 2.54, - 1.26]; P < 0.01) and the ≥ six months follow-up (n = 79 patients; MD, - 3.07 [95% CI, - 5.08, - 1.05]; P < 0.01). The improvement of AOFAS and VAS scores at ≥ six months follow-up reached the minimal clinically important difference (MCID). Nevertheless, the treatment effect of AOFAS and VAS scores offered by PRP at short-term follow-up did not exceed the MCID. Substantial heterogeneity was reported at the ≥ six months follow-up in VAS scores (I2: 93%, P < 0.01). CONCLUSION: This meta-analysis supports the safety of PRP intra-articular injection for ankle OA. The improvements of AOFAS and VAS scores in the PRP group at short-term follow-up do not exceed the MCID to be clinically significant. PRP injection provides significant improvement of AOFAS score and reduced pain at ≥ six months follow-up. The efficacy of PRP should be interpreted with caution regarding the high heterogeneity and the scarcity of available literature, which urges large-scale RCTs with longer follow-up to confirm the potential efficacy of PRP injection for ankle OA.

An interpretable transformer network for the retinal disease classification using optical coherence tomography.

Retinal illnesses such as age-related macular degeneration and diabetic macular edema will lead to irreversible blindness. With optical coherence tomography (OCT), doctors are able to see cross-sections of the retinal layers and provide patients with a diagnosis. Manual reading of OCT images is time-consuming, labor-intensive and even error-prone. Computer-aided diagnosis algorithms improve efficiency by automatically analyzing and diagnosing retinal OCT images. However, the accuracy and interpretability of these algorithms can be further improved through effective feature extraction, loss optimization and visualization analysis. In this paper, we propose an interpretable Swin-Poly Transformer network for performing automatically retinal OCT image classification. By shifting the window partition, the Swin-Poly Transformer constructs connections between neighboring non-overlapping windows in the previous layer and thus has the flexibility to model multi-scale features. Besides, the Swin-Poly Transformer modifies the importance of polynomial bases to refine cross entropy for better retinal OCT image classification. In addition, the proposed method also provides confidence score maps, assisting medical practitioners to understand the models' decision-making process. Experiments in OCT2017 and OCT-C8 reveal that the proposed method outperforms both the convolutional neural network approach and ViT, with an accuracy of 99.80% and an AUC of 99.99%.

Cancer detection for small-size and ambiguous tumors based on semantic FPN and transformer.

Early detection of tumors has great significance for formative detection and determination of treatment plans. However, cancer detection remains a challenging task due to the interference of diseased tissue, the diversity of mass scales, and the ambiguity of tumor boundaries. It is difficult to extract the features of small-sized tumors and tumor boundaries, so semantic information of high-level feature maps is needed to enrich the regional features and local attention features of tumors. To solve the problems of small tumor objects and lack of contextual features, this paper proposes a novel Semantic Pyramid Network with a Transformer Self-attention, named SPN-TS, for tumor detection. Specifically, the paper first designs a new Feature Pyramid Network in the feature extraction stage. It changes the traditional cross-layer connection scheme and focuses on enriching the features of small-sized tumor regions. Then, we introduce the transformer attention mechanism into the framework to learn the local feature of tumor boundaries. Extensive experimental evaluations were performed on the publicly available CBIS-DDSM dataset, which is a Curated Breast Imaging Subset of the Digital Database for Screening Mammography. The proposed method achieved better performance in these models, achieving 93.26% sensitivity, 95.26% specificity, 96.78% accuracy, and 87.27% Matthews Correlation Coefficient (MCC) value, respectively. The method can achieve the best detection performance by effectively solving the difficulties of small objects and boundaries ambiguity. The algorithm can further promote the detection of other diseases in the future, and also provide algorithmic references for the general object detection field.