Purpose:This study aimed to validate the diagnostic value of kinetic heterogeneity (KH) across multiple cohorts and to develop a nomogram integrating KH, apparent diffusion coefficient (ADC), and key morphological and clinical features for preoperative discrimination between benign and malignant breast masses. Methods:This retrospective study included 501 female patients with 541 confirmed breast mass lesions. Training (n = 280) and internal validation (n = 119) cohorts were collected (2018-2019). Temporal validation (n = 101, 2024) and independent test cohorts (n = 41, 2025-2026), from a different scanner system) were also included. KH was computed using voxel-wise time-intensity curve analysis. Independent predictors were identified using logistic regression to construct a nomogram. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis, and compared with KH-, ADC-, and combined models. Results:Age, lesion margin, ADC, and KH were independent predictors. The nomogram achieved AUCs of 0.897 (95% confidence interval [CI]: 0.838-0.955) and 0.909 (95% CI: 0.811-1.000) in the temporal validation and independent test cohorts, outperforming the ADC + KH (0.855, 0.890), ADC (0.829, 0.848), and KH (0.803, 0.757) models. Sensitivities were 93.8% and 94.1%, with negative predictive values (NPVs) of 92.7% and 95.2%, respectively. Calibration and decision curve analysis demonstrated good agreement and positive net benefit across a wide range of threshold probabilities. Conclusion:The nomogram integrating KH, ADC, margin, and age demonstrated superior diagnostic performance in preoperatively discriminating between benign and malignant breast mass lesions.

γ-Al2O3 is one of the most widely used catalyst supports in heterogeneous catalysis, yet its catalytic role remains unclear due to intrinsic structural disorder and complex surface species. This paper describes the development of a high-accuracy Al–O–H machine learning interatomic potential combined with global optimization to explore the structural landscape of γ-Al2O3 and its influence on propane dehydrogenation over single-atom Pt catalysts. Two energetically favorable structures, γ-no aluminum vacancy (NAV) and γ-aluminum vacancy (AV), are identified with energies lower than the conventional model by up to 34.44 meV·(f.u.)−1. These optimized structures expose abundant penta-coordinated Al3+ sites on the (100) surface, serving as preferred anchoring sites for Pt atoms. Simulated X-ray diffraction patterns indicate that γ-AV shows better qualitative agreement with experimental data. Surface phase diagrams further reveal that defect-rich surfaces are thermodynamically stabilized under realistic reaction conditions. Catalytic calculations demonstrate that γ-NAV and γ-AV significantly reduce propane dehydrogenation activation barriers through enhanced metal-support interactions associated with penta-coordinated Al3+ sites and defect-modulated local environments. These results suggest that γ-Al2O3 is better described as an ensemble of defect-rich surface configurations rather than a single crystal structure. These findings establish a direct relationship between atomic structure, defect chemistry, and catalytic performance in γ-Al2O3.

Direct electrochemical ethylene (C2H4) epoxidation with water (H2O) represents a promising approach for the production of value-added ethylene oxide (EO) in a sustainable way. However, the activity remains limited due to the sluggish activation of C2H4 and the stiff formation of *OH intermediate. This paper describes the design of a Ag/SnO2 electrocatalyst to achieve efficient electrochemical C2H4 epoxidation with a high faradaic efficiency of 39.4% for EO and a high selectivity of 91.5% at 25 mA/cm2 in a membrane electrode assembly. Results of in situ attenuated total reflection infrared spectra characterizations and computational calculations reveal that the Ag/SnO2 interface promotes C2H4 adsorption and activation to obtain *C2H4. Moreover, electrophilic *OH is generated on the catalyst surface through H2O dissociation, which further reacts with *C2H4 to facilitate the formation of *C2H4OH, contributing to the enhanced electrochemical epoxidation activity. This work would provide general guidance for designing catalysts for electrochemical olefin epoxidation through interface engineering.

Although non-noble metal catalysts are appealing for propane dehydrogenation, achieving high propylene selectivity remains a persistent challenge, which necessitates the regulation of catalytic microenvironment. In this study, we comparatively investigate three commonly used active metals (Pt, Pd, and non-noble metal Ni) using both theoretical and experimental approaches. We find that the low selectivity of Ni-based catalysts is intrinsically attributed to a narrow interatomic distance (Delta d) between Ni atoms, which promotes side reactions. Thus, Ni-based intermetallic alloys are employed to modulate Delta d, whose surface microenvironment is quantified with a descriptor called degree-of-isolation. The established volcano-shaped isolation-selectivity plot provides a direct avenue for predicting propylene selectivity, which is determined by two competing variables: desorption and further dehydrogenation of propylene. The optimal catalyst, NiIn, manifests moderate Ni-C repulsion, obtaining >91% experimental propylene selectivity. This reveals the Sabatier principle over Ni-based catalysts for selective propane dehydrogenation and underscores the significance of microenvironment engineering.

Redox catalysts play a critical role in chemical looping oxidative dehydrogenation of propane (CL-ODH). However, challenges persist in modulating lattice oxygen in metal oxides and maintaining surface oxygen coverage to prolong the oxidative dehydrogenation stage. This paper describes the role of oxygen vacancies by evaluating numerous vacancy distribution patterns, including surface and bulk distributions, to identify VO x surfaces across a wide range of reduction degrees, guided by calculated oxygen vacancy formation energy. The surface reactions are classified into three distinct stages based on surface oxygen vacancy coverage (Ovc), with transitions between stages attributed to the excessive reactivity of lattice oxygen, variations in vanadium valence states, and the localized limitations of vacancy effects. Additionally, four high-valent metal dopants (W, Mo, Nb, and Os) identified through charge transfer energy (CTE)-based descriptors effectively reduce oxygen reactivity while optimizing the utilization of bulk lattice oxygen to maintain favorable surface Ovc. These findings provide essential theoretical insights and a strategic framework for the rational design of redox catalysts in CL-ODH applications.

Background:: Tertiary lymphoid structure (TLS) is an ectopic lymphoid structure that develops in non-lymphoid structures. Some studies have shown that the TLS formed in autoimmune diseases, such as lupus nephropathy (LN), can cause damage to normal tissues and continuous disease progression. Nevertheless, there is still a lack of efficient treatments for TLS in LN. Thus, the study aims to identify potential targets for therapy of TLS in LN. Methods:: Mice datasets relative to TLS were obtained from Gene Expression Omnibus (GEO). The differentially expressed genes (DEGs) were identified from mice datasets. Then, the Genetic Ontological (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were performed. The Protein-Protein Interaction (PPI) network was constructed. Additionally, the hub genes were selected by Cytoscape and verified by human databases from GEO. The relationships between the immune cells with hub genes were explored. Finally, the two genes PSMB9 and STAT1 were validated in the kidney tissues of LN patients and mice. Results:: 443 DEGs and 178 DEGs relative to TLS were filtered from GSE160488 and GSE155405, respectively. The enrichment results of these genes mostly focused on inflammatory response, cytokine-cytokine receptor interaction, and immune system process. Six genes were recognized by Cytoscape. According to the validation of six genes in human databases, the two hub genes (PSMB9 and STAT1) were also significantly expressed in LN patients. Immune infiltration analysis of hub genes shows immune cells are significantly crucial in LN patients with TLS. Conclusion:: PSMB9 and STAT1 may be identified as possible targets for the treatment of TLS in LN. According to the analysis of the interaction between these genes and immune cells, the immune process mediated by these signature targets takes part in the advancement and formation of TLS.

The study aims to develop a deep learning (DL) model based on multiparametric magnetic resonance imaging (MRI) for distinguishing between benign and malignant breast lesions. A total of 556 lesions (307 malignant, 249 benign) in 509 patients were pooled in the training/validation datasets between November 2018 and October 2019 in this retrospective study. A combined DL model based on the dynamic contrast enhanced-MRI (DCE-MRI) and apparent diffusion coefficient (ADC) map was developed to characterize breast lesions. Model performance was evaluated by using areas under the receiver operating characteristic curve (AUC) in the validation dataset and an independent testing dataset consisting of 243 lesions in 225 patients, and compared with other combined and single-parametric DL models. The predictive performance for malignancy was also compared between the DCE-ADC combined DL model and human readers. The DCE-ADC combined DL model achieved the highest diagnostic efficiency with the AUC, accuracy, sensitivity, and specificity of 0.889, 82.5

Abstract Background Recent data has shown a considerable advancement in understanding the role of lymphotoxin-β receptor (LTβR) in inflammation. However, the functions and underlying mechanisms of LTβR in acute kidney injury (AKI) remain largely unknown. Methods AKI was induced in mice by renal ischemia-reperfusion (I/R). HK-2 cells and primary renal tubular epithelial cells (RTECs) were subjected to hypoxia/reoxygenation (H/R) injury. The effects of LTβR depletion were examined in mice, as well as primary RTECs. Bone marrow chimeric mice was generated to determine whether the involvement of LTβR expression by parenchymal cells or bone marrow derived cells contributes to renal injury during AKI. RNA sequencing techniques were employed to investigate the mechanism via which LTβR signaling provides protection against I/R-induced AKI Results LTβR expression was downregulated both in vivo and in vitro models of AKI. Moreover, depletion of LTβR decreased renal damage and inflammation in I/R-induced AKI. We also found that LTβR deficient mice engrafted with wild type bone marrow had significantly less tubular damage, implying that LTβR in renal parenchymal cells may play dominant role in I/R-induced AKI. RNA sequencing indicated that the protective effect of LTβR deletion was associated with activation of PPARα signaling. Furthermore, upregulation of PPARα was observed upon depletion of LTβR. PPARα inhibitor, GW6471, aggravated the tubular damage and inflammation in LTβR−/− mice following I/R injury. Then we further demonstrated that LTβR depletion down-regulated non-canonical NF-κB and Bax/Bcl-2 apoptosis pathway through PPARα. Conclusions Our results suggested that the LTβR/PPARα axis may be a potential therapeutic target for the treatment of AKI.

The inherent chemical tunability of perovskite materials has spurred extensive research into composition engineering within the perovskite community. However, identifying the optimal composition across a broad range of variations still remains a significant challenge. Conventional trial-and-error methods are prohibitively expensive and environmentally taxing for comprehensive screening. Here, we employed machine learning-accelerated atomic simulation to guide the design of stable perovskite solar cells absorbers. Our approach entailed training of a neural network (NN) potential using data generated from first-principles calculations, yielding a perovskite NN potential exhibiting high accuracy. Utilizing this NN potential, we constructed a phase diagram for FA x Cs1-x Pb(I y Br1-y )3 (where 0 ≤ x ≤ 1 and 0 ≤ y ≤ 1, FA denotes formamidinium cation). Integrating this with a band gap diagram, we successfully identified global optimal perovskite compositions for tandem applications with 1.7 and 1.8 eV band gaps. We have identified that all FA x Cs1-x Pb(I y Br1-y )3 with >1.8 eV band gaps are thermodynamically vulnerable to phase segregation and developed a strategy to stabilize thermodynamically unstable phases by suppressing phase segregation kinetics. Finally, theoretical predictions were confirmed by the corresponding experiments. Our results suggest that creating perovskites/Si tandem solar cells with 1.7 eV FA x Cs1-x Pb(I y Br1-y )3 encounters less severe challenges in addressing phase segregation issues than perovskites/perovskites tandem solar cells with 1.8 eV FA x Cs1-x Pb(I y Br1-y )3.

Maximizing atomic utilization of noble metals is crucial for efficient industrial catalysis. We demonstrate that minimal platinum (Pt) loading for propane dehydrogenation (PDH) can be achieved through atom abstraction. At low loadings of Pt with copper (Cu), reduction over silica or other oxide supports formed nanoparticles (NPs) with Pt mainly dispersed in the bulk. Addition of tin (Sn) to the alloy led to formation of surface Pt1Sn1 dimers. The larger atomic radius of Sn compared with Cu drove it to the surface, and its stronger interactions with Pt abstracted Pt from the bulk. Single metallic Pt atoms were stabilized on fully open surfaces, resulting in nearly 100% surface exposure. This configuration reduced Pt usage by one order of magnitude for propane dehydrogenation and improved catalytic stability.

Chemical looping partial oxidation of methane (CL-POM) offers a promising approach to produce syngas with high selectivity and reduced explosion risk. However, the design of metal oxide oxygen carriers with excellent performance and continuous oxygen release capacity remains a challenge. In this study, we developed a composite oxygen carrier (LaFeO 3− δ /Ca 1− η Sr η MnO 3 ) with the aim of modulating the oxygen transport capacity of Ca 1− η Sr η MnO 3 to maintain active structures on the LaFeO 3− δ (121) defected surface, thereby enhancing the activity and selectivity of CL-POM. The Fe-O 4 (O V ) and Fe-O 3 (O V ) 2 local structures were found to serve as active sites for methane oxidation on the LaFeO 3− δ (121) defected surface, with comparable free energy barriers of reaction (Δ G a = 1.44 and 1.40 eV, respectively). Based on the oxygen migration energy barriers and reaction energy barriers calculated by DFT, we determined oxygen transport coefficients and surface reaction rate constants to further assess the degree of rate matching between bulk oxygen transport and surface oxygen consumption. Finally, LaFeO 3− δ /Ca 0.75 Sr 0.25 -MnO 3 was proposed as a potential candidate for CL-POM. This composite material achieves commendable rate matching between surface reactivity and bulk oxygen transport, and notably exhibits the highest phase transition energy barrier, effectively inhibiting adverse phase transitions.

We addressed the challenges of designing catalysts for selective CO2 hydrogenation by incorporating Fe oxide species onto Rh nanoparticles. Nanoscopic FeOx domains created a "reverse catalyst" structure (i.e., a metal oxide supported on a metal) that increased the density of interfacial sites compared to traditional supported catalysts. The contact between the metal nanoparticle and the oxide overlayer induced the formation of a surface Rh-Fe alloy that stabilized methoxy groups while suppressing hydrogenolysis to methane. Sites at FeOx-metal interfaces interact with CO2 much stronger than sites on metal surfaces, show larger energy barriers to cleave the C-O bonds, and offer a barrierless pathway for the hydrogenation of methoxy species to methanol. Consequently, the multifunctional sites over FeOx/Rh-Fe catalysts highlight and meet the requirements of a selective methanol catalyst: strong interaction with CO2 to ensure a high density of transition states, metal sites to activate and make hydrogen available to surface intermediates, and high energy barriers for C-O bond cleavage to form carbides. These synthetic and catalytic chemistries, demonstrated for Rh-Fe-FeOx interfaces, enable us to overcome the limitations to the design of methanol production catalysts.

CRISPR/Cas12a systems have been repurposed as powerful tools for developing next-generation molecular diagnostics due to their trans-cleavage ability. However, it was long considered that the CRISPR/Cas12a system could only recognize DNA targets. Herein, we systematically investigated the intrinsic trans-cleavage activity of the CRISPR/Cas12a system (LbCas12a) and found that it could be activated through fragmented ssDNA activators. Remarkably, we discovered that the single-stranded DNA (ssDNA) activators in the complementary crRNA-distal domain could be replaced by target miRNA sequences without the need for pre-amplification or specialized recognition mechanisms. Based on these findings, we proposed the “Fragment Complementary Activation Strategy” (FCAS) and designed reverse fluorescence-enhanced lateral flow test strips (rFLTS) for the direct detection of miRNA-10b, achieving a limit of detection (LOD) of 5.53 fM and quantifying the miRNA-10b biomarker in clinical serum samples from glioma patients. Moreover, for the first time, we have developed the FCAS-based CRISPR/Cas12a system for miRNA in situ imaging, effectively recognizing tumor cells. The FCAS not only broadens the scope of CRISPR/Cas12a system target identification but also unlocks the potential for in-depth studies of CRISPR technology in many diagnostic settings.

Background: Renal tertiary lymphoid structures (TLSs) are involved in renal pathology and prognosis of IgA nephropathy (IgAN). CD30 and its ligands participate in the formation of renal TLSs. However, the relationship between circulating CD30 and renal prognosis is unclear. The objective of this study was to evaluate the relationship between circulating CD30 and prognosis in patients with IgAN. Methods: We conducted a retrospective study including 351 patients with biopsy proved IgAN. We collected clinical and pathologic features at the time of biopsy and recorded renal follow-up outcomes. Circulating CD30 levels in IgAN patients at the time of biopsy were measured via enzyme-linked immunosorbent assay (ELISA). The association between elevated CD30 levels and the composite endpoint (defined as a >= 50 % decline in eGFR from baseline, end-stage renal disease, or death) was investigated using Cox regression analysis. Results: During a median follow-up period of 5.12 years, 44 (12.5 %) patients in the cohort reached the composite endpoint. Kaplan-Meier survival curve analysis revealed a significant association between higher circulating CD30 levels and a poorer renal prognosis (log-rank P < 0.001). Cox regression analysis showed that high CD30 was an independent factor for the composite endpoints in multivariable-adjusted models (HR 3.397, 95 % CI: 1.230-9.384, P = 0.018). These associations were also observed in a subgroup of patients with concomitant renal TLSs formation (10.443, 95 % CI: 1.680-65.545, P = 0.012), proteinuria > 1 g/d (HR 12.287, 95 % CI: 1.499-100.711, P = 0.019), and female patients (HR 22.372, 95 % CI: 1.797-278.520, P = 0.016). Conclusion: Elevated level of circulating CD30 is an independent risk factor for renal disease progression in patients with IgAN.

Intermetallic nanoparticles (NPs) possess significant potentials for catalytic applications, yet their production presents challenges as achieving the disorder-to-order transition during the atom ordering process involves overcoming a kinetic energy barrier. Here, we demonstrate a robust approach utilizing atomic gas-migration for the in-situ synthesis of stable and homogeneous intermetallic alloys for propane dehydrogenation (PDH). This approach relies on the physical mixture of two separately supported metal species in one reactor. The synthesized platinum-zinc intermetallic catalysts demonstrate exceptional stability for 1300 h in continuous propane dehydrogenation under industrially relevant industrial conditions, with extending 95% propylene selectivity and propane conversions approaching thermodynamic equilibrium values at 550-600 oC. In situ characterizations and density functional theory/molecular dynamics simulation reveal Zn atoms adsorb on the particle surface and then diffuse inward, aiding in the formation of ultrasmall and highly ordered intermetallic alloys. This in-situ gas-migration strategy is applicable to a wide range of intermetallic systems.

Strain engineering plays an important role in tuning electronic structure and improving catalytic capability of biocatalyst, but it is still challenging to modify the atomic-scale strain for specific enzyme-like reactions. Here, we systematically design Pt single atom (Pt1), several Pt atoms (Ptn) and atomically-resolved Pt clusters (Ptc) on PdAu biocatalysts to investigate the correlation between atomic strain and enzyme-like catalytic activity by experimental technology and in-depth Density Functional Theory calculations. It is found that Ptc on PdAu (Ptc-PA) with reasonable atomic strain upshifts the d-band center and exposes high potential surface, indicating the sufficient active sites to achieve superior biocatalytic performances. Besides, the Pd shell and Au core serve as storage layers providing abundant energetic charge carriers. The Ptc-PA exhibits a prominent peroxidase (POD)-like activity with the catalytic efficiency (Kcat/Km) of 1.50 × 109 mM-1 min-1, about four orders of magnitude higher than natural horseradish peroxidase (HRP), while catalase (CAT)-like and superoxide dismutase (SOD)-like activities of Ptc-PA are also comparable to those of natural enzymes. Biological experiments demonstrate that the detection limit of the Ptc-PA-based catalytic detection system exceeds that of visual inspection by 132-fold in clinical cancer diagnosis. Besides, Ptc-PA can reduce multi-organ acute inflammatory damage and mitigate oxidative stress disorder.

目的:分析乳腺实性乳头状癌(solid papillary carcinoma,SPC)的磁共振成像(magnetic resonance imaging,MRI)特征,探究MRI对于SPC的诊断价值.方法:回顾并收集2017年1月—2021年12月上海交通大学医学院附属新华医院经手术后病理学检查证实为SPC且行术前MRI检查的患者57例(共61个SPC病灶).57例患者中,行术前乳腺X线摄影及超声检查者分别为45例(48个SPC)和52例(55个SPC).根据术前乳腺影像报告和数据系统(Breast Imaging Reporting and Data System,BI-RADS)分类结果,以BI-RADS≥4A类为可疑恶性,计算乳腺X线摄影、超声及MRI对SPC的检出率及诊断准确度.病灶形态分为非肿块强化(non-mass enhancement,NME)与肿块两组,两组大小比较采用Mann-Whitney U检验,伴随导管扩张的差异采用χ2检验.结果:乳腺X线摄影、超声及MRI对SPC的检出率为分别为64.6%(31/48)、83.6%(46/55)和100.0%(61/61),诊断准确度分别为52.1%(25/48)、65.5%(36/55)和98.4%(60/61).在MRI上,SPC表现为NME较肿块更多见(67.2%vs 32.8%).NME较肿块病灶更大[2.5(1.6,4.0)cm vs 1.4(1.0,1.8)cm,P＜0.001],伴随导管扩张的阳性率更高[82.9%(34/41)vs 25.0%(5/20),P＜0.001].结论:乳腺MRI对于SPC的检出率及诊断准确度均高于乳腺X线摄影和超声检查.在MRI上,SPC表现为NME较肿块更多见,前者病灶更大,更常伴随导管扩张.

Tertiary lymphoid structure (TLS) represents lymphocyte clusters in non-lymphoid organs. The formation and maintenance of TLS are dependent on follicular helper T (TFH) cells. However, the role of TFH cells during renal TLS formation and the renal fibrotic process has not been comprehensively elucidated in chronic kidney disease. Here, we detected the circulating TFH cells from 57 IgAN patients and found that the frequency of TFH cells was increased in IgA nephropathy patients with renal TLS and also increased in renal tissues from the ischemic-reperfusion-injury (IRI)-induced TLS model. The inducible T-cell co-stimulator (ICOS) is one of the surface marker molecules of TFH. Remarkably, the application of an ICOS-neutralizing antibody effectively prevented the upregulation of TFH cells and expression of its canonical functional mediator IL-21, and also reduced renal TLS formation and renal fibrosis in IRI mice in vivo. In the study of this mechanism, we found that recombinant IL-21 could directly promote renal fibrosis and the expression of p65. Furthermore, BAY 11-7085, a p65 selective inhibitor, could effectively alleviate the profibrotic effect induced by IL-21 stimulation. Our results together suggested that TFH cells contribute to TLS formation and renal fibrosis by IL-21. Targeting the ICOS-signaling pathway network could reduce TFH cell infiltration and alleviate renal fibrosis.

Redox chemistry plays an important role in the precise production of targeted valuable chemicals. However, achieving accurate control of the redox process still lacks the guidance of mechanistic insight. This paper describes how doping can enhance redox performance of vanadia for chemical looping oxidative propane dehydrogenation. By elucidating the dynamic redox process of vanadia, we discovered that molybdenum (Mo) dopant lowers the phase trans-formation barrier from defective V2O5 toward VO2. In the VO2 phase, Mo doping modulates chemical properties of every oxygen atom, while doping in V2O5 only affects the oxygen atoms that directly bond with the dopants. Charge analysis indicated that the conductive nature of VO2 delocalizes charge within the lattice and alters the overall chemical properties. Experiments confirmed that Mo-doped VO2 is the dominant phase during the redox process, exhibiting the best catalytic performance. Hence, delocalization could be key to future chemical applications of efficiently doped materials.

The objective of this study is to develop a radiomic signature constructed from deep learning features and a nomogram for prediction of axillary lymph node metastasis (ALNM) in breast cancer patients. Preoperative magnetic resonance imaging data from 479 breast cancer patients with 488 lesions were studied. The included patients were divided into two cohorts by time (training/testing cohort, n = 366/122). Deep learning features were extracted from diffusion-weighted imaging–quantitatively measured apparent diffusion coefficient (DWI-ADC) imaging and dynamic contrast-enhanced MRI (DCE-MRI) by a pretrained neural network of DenseNet121. After the selection of both radiomic and clinicopathological features, deep learning signature and a nomogram were built for independent validation. Twenty-three deep learning features were automatically selected in the training cohort to establish the deep learning signature of ALNM. Three clinicopathological factors, including LN palpability (odds ratio (OR) = 6.04; 95% confidence interval (CI) = 3.06–12.54, P = 0.004), tumor size in MRI (OR = 1.45, 95% CI = 1.18–1.80, P = 0.104), and Ki-67 (OR = 1.01; 95% CI = 1.00–1.02, P = 0.099), were selected and combined with radiomic signature to build a combined nomogram. The nomogram showed excellent predictive ability for ALNM (AUC 0.80 and 0.71 in training and testing cohorts, respectively). The sensitivity, specificity, and accuracy were 65%, 80%, and 75%, respectively, in the testing cohort. MRI-based deep learning radiomics in patients with breast cancer could be used to predict ALNM, providing a noninvasive approach to structuring the treatment strategy.