SPIRiT-Diffusion: Self-Consistency Driven Diffusion Model for Accelerated MRI.

Diffusion models have emerged as a leading methodology for image generation and have proven successful in the realm of magnetic resonance imaging (MRI) reconstruction. However, existing reconstruction methods based on diffusion models are primarily formulated in the image domain, making the reconstruction quality susceptible to inaccuracies in coil sensitivity maps (CSMs). k-space interpolation methods can effectively address this issue but conventional diffusion models are not readily applicable in k-space interpolation. To overcome this challenge, we introduce a novel approach called SPIRiT-Diffusion, which is a diffusion model for k-space interpolation inspired by the iterative self-consistent SPIRiT method. Specifically, we utilize the iterative solver of the self-consistent term (i.e., k-space physical prior) in SPIRiT to formulate a novel stochastic differential equation (SDE) governing the diffusion process. Subsequently, k-space data can be interpolated by executing the diffusion process. This innovative approach highlights the optimization model's role in designing the SDE in diffusion models, enabling the diffusion process to align closely with the physics inherent in the optimization model-a concept referred to as model-driven diffusion. We evaluated the proposed SPIRiT-Diffusion method using a 3D joint intracranial and carotid vessel wall imaging dataset. The results convincingly demonstrate its superiority over image-domain reconstruction methods, achieving high reconstruction quality even at a substantial acceleration rate of 10. Our code are available at https://github.com/zhyjSIAT/SPIRiT-Diffusion.

A dynamic approach for MR T2-weighted pelvic imaging.

OBJECTIVE: T2-weighted 2D fast spin echo sequence serves as the standard sequence in clinical pelvic MR imaging protocols. However, motion artifacts and blurring caused by peristalsis present significant challenges. Patient preparation such as administering antiperistaltic agents is often required before examination to reduce artifacts, which discomfort the patients. This work introduce a novel dynamic approach for T2 weighted pelvic imaging to address peristalsis-induced motion issue without any patient preparation. Approach: A rapid dynamic data acquisition strategy with complementary sampling trajectory is designed to enable highly undersampled motion-resistant data sampling, and an unrolling method based on deep equilibrium model is leveraged to reconstruct images from the dynamic sampled k-space data. Moreover, the fix-point convergence of the equilibrium model ensures the stability of the reconstruction. The high acceleration factor in each temporal phase, which is much higher than that in traditional static imaging, has the potential to effectively freeze pelvic motion, thereby transforming the imaging problem from conventional motion prevention or removal to motion reconstruction. Main results: Experiments on both retrospective and prospective data have demonstrated the superior performance of the proposed dynamic approach in reducing motion artifacts and accurately depicting structural details compared to standard static imaging. Significance: The proposed dynamic approach effectively captures motion states through dynamic data acquisition and deep learning-based reconstruction, addressing motion-related challenges in pelvic imaging.

Acquired factor XIII deficiency presenting with multiple intracranial hemorrhages and right hip hematoma: A case report.

BACKGROUND: Factor XIII (FXIII) deficiency is a rare yet profound coagulopathy. FXIII plays a pivotal role in hemostasis, and deficiencies in this factor can precipitate unchecked or spontaneous hemorrhaging. Immunological assays for detecting FXIII inhibitors are indispensable for diagnosing acquired FXIII deficiency; however, the availability of suitable testing facilities is limited, resulting in prolonged turnaround times for these assays. CASE SUMMARY: In this case study, a 53-year-old male devoid of significant medical history presented with recurrent intracranial hemorrhages and a hematoma in the right hip. Subsequent genetic analysis revealed a homozygous mutation in the ACE gene, confirming the diagnosis of acquired FXIII deficiency. CONCLUSION: This case underscores the significance of considering acquired deficiencies in clotting factors when evaluating patients with unexplained bleeding episodes.

Synthesis of Inherently Chiral Homo-heteracalixaromatics.

There is a resurgence of research interest in inherently chiral calixarenes and heteracalixaromatics. However, the examples of macrocyclic ring-expanded homocalixarenes and heterocalixaromatics of inherent chirality are not known. Here we report an efficient method to construct inherently chiral N2,O2-linked homo-heteracalixaromatics from reductive amination reactions between bis-aldehydes and aliphatic diamines. Examples of post-macrocyclization derivatization were also demonstrated. Enantiomers were obtained from resolution and did not undergo racemization at an elevated temperature. This study provides a new strategy to design novel and functional inherently chiral macrocycles that have potential applications in supramolecular chemistry.

Soluble sugar, organic acid and phenolic composition and flavor evaluation of plum fruits.

Plums (Prunus salicina and Prunus domestica) are prevalent in southwestern China, and have attracted interest owing to their delectable taste and exceptional nutritional properties. Therefore, this study aimed to investigate the nutritional and flavor properties of plum to improve its nutritional utilization. Specifically, we determined the soluble sugars, organic acids, and phenolic components in 86 accessions using high-performance liquid chromatography. Notably, glucose, fructose, malic, and quinic acids were the predominant sweetness and acidity in plums, with sucrose contributing more to the sweetness of the flesh than the peel. Moreover, The peel contains 5.5 fold more phenolics than flesh, epicatechin, gallic acid, and proanthocyanidins C1 and B2 were the primary sources of astringency. Correlation and principal component analyses showed eight core factors for plum flavor rating, and a specific rating criterion was established. Conclusively, these findings provide information on the integrated flavor evaluation criteria and for enhancing optimal breeding of plums.

Major Bitter-Tasting Compounds from the Dichloromethane Fraction of Bitter Gourd (Fruit of Momordica charantia L.) Extract and Their Precursors.

The unripe fruit of the plant Momordica charantia L., commonly known as bitter gourd or bitter melon, is a popular vegetable and medical herb in many parts of the world and is characterized by its strong bitter taste. In our endeavor to decode its bitter taste, the dichloromethane fraction of 75% methanol extract of bitter gourd was found to be intensely bitter. Combining sensory analysis-guided fractionation and newly developed comparative high-pressure liquid chromatography (HPLC) analysis-guided purification led to the isolation of five known compounds including momordicoside L (1), (23E)-3ß-O-malonyl-7ß,25-dihydroxycucurbita-5,23-dien-19-al (2), 3-O-ß-d-allopyranosyl-7ß,25-dihydroxycucurbita-5,23(E)-dien-19-al (3), momordicine IV (5), and charantoside B (6) and three new compounds 3-O-ß-d-3-ketoglucopyranosly-7ß,25-dihydroxycucurbita-5,23(E)-dien-19-al (4), 6'-O-malonylmomordicoside L (7), and 6'-O-malonylmomordicine IV (8) from bitter gourd. Sensory analysis revealed compounds 3-8 had strong bitter taste with their bitter taste recognition thresholds in the range between 3.6 (4) and 13.6 ppm (6) in 3% ethanol. UPLC-MS/MS quantification showed that their concentration in bitter gourd ranged from 16.5 ± 1.3 mg/kg (4) to 214.8 ± 14.0 mg/kg (6) on a dry weight basis. Calculation of the dose-over-threshold (DoT) factor showed that momordicine IV (5) and charantoside B (6) should be considered as bitter principles of bitter gourd. In addition, the study also demonstrated the ubiquity of the isomerization reaction in the side chain of cucurbitane-type triterpenoids. Many isolated compounds were the isomerized products of their natural precursors, and these precursors should be the primary bitterness contributors of fresh fruits. In addition, comparative HPLC analysis-guided purification could be a practical approach for the fast isolation of acid-labile precursors.

Activated carbon fiber loaded nano zero-valent iron for Microcystis aeruginosa removal: Performance and mechanisms.

Cyanobacterial blooms caused by Microcystis aeruginosa threaten environmental safety and daily life. In this study, an activated carbon fiber-supported nano zero-valent iron composite (ACF-nZVI) was developed to remove Microcystis aeruginosa. The results showed that nZVI was evenly distributed on the activated carbon fibers, preventing aggregation and oxidation. ACF-nZVI achieved a removal efficiency of more than 90 % within a pH range of 3-7. During the reaction, H2O2, which was generated by Fe0, was activated to form ·OH and ·O-2, which dismantled antioxidant enzymes and induced lipid peroxidation. Additionally, ACF-nZVI destroyed the cell wall and membrane, resulting in protein and humus leakage and causing 92.34 % cell damage and death. In this study, an environmentally friendly and stable nanomaterial was developed, offering a novel approach for the safe, cost-effective, and efficient removal of cyanobacteria.

Disparities in the real-world prescription pattern of oral anticoagulants for US Medicare beneficiaries: a geospatial analysis.

OBJECTIVES: This study was conducted to examine urban-rural differences in the real-world prescribing pattern of oral anticoagulants and geographic variations in the prescribing pattern among clinicians serving Medicare beneficiaries in the USA. DESIGN: A cross-sectional study. SETTING: A real-world setting. PARTICIPANTS: 232 665 clinicians who prescribed oral anticoagulants for Medicare beneficiaries from the 2020 Medicare Provider Utilisation and Payment Data were classified as warfarin only, direct oral anticoagulants (DOACs) only or warfarin+DOAC prescribers. MAIN OUTCOME MEASURES: Urban-rural differences in the prescribing patterns were examined using multivariate multinominal logistic regression analysis. A geospatial analysis was conducted to estimate standardised prescriber ratios (SPR) for DOAC only or warfarin only prescribers versus warfarin+DOAC prescribers to illustrate hot and cold spots for DOAC adoption in the USA. RESULTS: 92% of clinicians who prescribed oral anticoagulants prescribed DOAC in 2020. Clinicians from rural regions were more likely to prescribe warfarin only (adjusted OR=1.335, 95% CI=(1.281 to 1.391)) and DOAC only (adjusted OR=2.052, 95% CI=(1.999 to 2.106)), compared with clinicians from urban regions. Hot spots for SPR of DOAC only versus warfarin+DOAC prescribers were mostly found in California, southern and southeastern states; cold spots were mostly found in Minnesota and Iowa. Hot spots for SPR of warfarin only versus warfarin+DOAC prescribers were mostly found in several metropolitan areas on the west coast and in Midwest; cold spots were mostly found on the east coast. CONCLUSIONS: Urban-rural status of clinicians was associated with their prescribing patterns of oral anticoagulants. The study identifies geographical heterogeneity in DOAC adoption and highlights gaps that may need to be addressed for increased accessibility of DOAC for patients in need.

HA-TiNet: Learning a Distinctive and General 3D Local Descriptor for Point Cloud Registration.

Extracting geometric features from 3D point clouds is widely applied in many tasks, including registration and recognition. We propose a simple yet effective method, termed height-azimuth image based transformation-invariant net (HA-TiNet), to learn a distinctive, general and rotation-invariant 3D local descriptor. HA-TiNet is composed of a height-azimuth image generator and a feature extraction net. Based on a local reference axis (LRA), the height-azimuth image generator first partitions local region along the plane-radial direction, and then implements a statistic of height and azimuth information in each divided space to generate a set of height-azimuth images. The generated height-azimuth images are invariant in the rotation around x- and y-axes and have high accuracy due to the high repeatability of an LRA. Besides, they can be easily embedded in 2D convolutional neural networks (CNNs). Our feature extraction net learns the information on the height-azimuth images using a ResNet-based backbone and a rotation-invariant layer. The ResNet-based backbone is lightweight while very effective. The rotation-invariant layer removes the rotation-variance around z-axis, making our descriptor have full rotation-invariance. Extensive experiments on indoor and outdoor datasets show that our method presents superior overall performance, and exhibits strong descriptiveness and generalization ability compared to the state-of-the-art descriptors. The source code will be made publicly available at https://github.com/ahulq/HA-TiNet.

Physics-Informed DeepMRI: k-Space Interpolation Meets Heat Diffusion.

Recently, diffusion models have shown considerable promise for MRI reconstruction. However, extensive experimentation has revealed that these models are prone to generating artifacts due to the inherent randomness involved in generating images from pure noise. To achieve more controlled image reconstruction, we reexamine the concept of interpolatable physical priors in k-space data, focusing specifically on the interpolation of high-frequency (HF) k-space data from low-frequency (LF) k-space data. Broadly, this insight drives a shift in the generation paradigm from random noise to a more deterministic approach grounded in the existing LF k-space data. Building on this, we first establish a relationship between the interpolation of HF k-space data from LF k-space data and the reverse heat diffusion process, providing a fundamental framework for designing diffusion models that generate missing HF data. To further improve reconstruction accuracy, we integrate a traditional physics-informed k-space interpolation model into our diffusion framework as a data fidelity term. Experimental validation using publicly available datasets demonstrates that our approach significantly surpasses traditional k-space interpolation methods, deep learning-based k-space interpolation techniques, and conventional diffusion models, particularly in HF regions. Finally, we assess the generalization performance of our model across various out-of-distribution datasets. Our code are available at https://github.com/ZhuoxuCui/Heat-Diffusion.

Bortezomib, lenalidomide, and dexamethasone versus bortezomib, doxorubicin, and dexamethasone in newly diagnosed multiple myeloma.

Bortezomib, lenalidomide, and dexamethasone (VRD), and bortezomib, doxorubicin, and dexamethasone (PAD), are commonly used in induction regimens for patients with newly diagnosed multiple myeloma (NDMM) in China. This real-world study enrolled 390 patients, 195 receiving VRD and 195 receiving PAD induction. The primary endpoint was progression-free survival (PFS) and stringent complete remission/complete remission. Across the entire cohort, VRD demonstrated significantly improved five-year overall survival (OS) (74% vs. 59%, p = 0.0024) and five-year PFS (67% vs. 37%, p = 0.0018) compared to PAD. Notably, the median OS and PFS were not reached for VRD-treated patients, while they were 77 months (60-not reached [NR]) and 46 months (36-NR), respectively, for PAD. In patients with standard-risk cytogenetics, VRD showed superior five-year OS (83% vs. 58%, p = 0.0038) and PFS (78% vs. 48%, p = 0.0091) compared to PAD. However, these differences were not statistically significant in high-risk patients. For transplanted patients, VRD was associated with superior five-year OS (91% vs. 67%, p = 0.014) and PFS (79% vs. 47%, p = 0.015) compared to PAD. In non-transplanted patients, VRD showed a trend towards improved five-year OS (p = 0.085) and PFS (p = 0.073) compared to the PAD group. In conclusion, VRD displayed superior OS and PFS outcomes in standard-risk patients and those who underwent transplantation. These findings suggest potential advantages of VRD over PAD in real-world clinical settings for NDMM treatment. However, due to the imbalance in transplantation rates between the VRD and PAD groups, limitations in testing for high-risk cytogenetic abnormalities (HRA), and the difference between the received cycles and salvage therapies, the conclusions of this study should be interpreted with caution.

Transcriptome analysis reveals the promoting effects of exogenous melatonin on the selenium uptake in grape under selenium stress.

Introduction: Exogenous melatonin (MT) can promote horticultural crops growth under stress conditions. Methods: In this study, the effects of exogenous MT on the accumulation of selenium (Se) in grape were studied under Se stress. Results and discussion: Under Se stress, exogenous MT increased the biomass, content of photosynthetic pigments and antioxidant enzyme activity of grapevines. Compared with Se treatment, MT increased the root biomass, shoot biomass, chlorophyll a content, chlorophyll b content, carotenoids, superoxide dismutase activity, and peroxidase activity by 18.11%, 7.71%, 25.70%, 25.00%, 25.93%, 5.73%, and 9.41%, respectively. Additionally, MT increased the contents of gibberellin, auxin, and MT in grapevines under Se stress, while it decreased the content of abscisic acid. MT increased the contents of total Se, organic Se and inorganic Se in grapevines. Compared with Se treatment, MT increased the contents of total Se in the roots and shoots by 48.82% and 135.66%, respectively. A transcriptome sequencing analysis revealed that MT primarily regulated the cellular, metabolic, and bioregulatory processes of grapevine under Se stress, and the differentially expressed genes (DEGs) were primarily enriched in pathways, such as aminoacyl-tRNA biosynthesis, spliceosome, and flavonoid biosynthesis. These involved nine DEGs and nine metabolic pathways in total. Moreover, a field experiment showed that MT increased the content of Se in grapes and improved their quality. Therefore, MT can alleviate the stress of Se in grapevines and promote their growth and the accumulation of Se.

Stability and passivity analysis of delayed neural networks via an improved matrix-valued polynomial inequality.

The stability and passivity of delayed neural networks are addressed in this paper. A novel Lyapunov-Krasovskii functional (LKF) without multiple integrals is constructed. By using an improved matrix-valued polynomial inequality (MVPI), the previous constraint involving skew-symmetric matrices within the MVPI is removed. Then, the stability and passivity criteria for delayed neural networks that are less conservative than the existing ones are proposed. Finally, three examples are employed to demonstrate the meliority and feasibility of the obtained results.

An additional proofreader contributes to DNA replication fidelity in mycobacteria.

The removal of mis-incorporated nucleotides by proofreading activity ensures DNA replication fidelity. Whereas the ε-exonuclease DnaQ is a well-established proofreader in the model organism Escherichia coli, it has been shown that proofreading in a majority of bacteria relies on the polymerase and histidinol phosphatase (PHP) domain of replicative polymerase, despite the presence of a DnaQ homolog that is structurally and functionally distinct from E. coli DnaQ. However, the biological functions of this type of noncanonical DnaQ remain unclear. Here, we provide independent evidence that noncanonical DnaQ functions as an additional proofreader for mycobacteria. Using the mutation accumulation assay in combination with whole-genome sequencing, we showed that depletion of DnaQ in Mycolicibacterium smegmatis leads to an increased mutation rate, resulting in AT-biased mutagenesis and increased insertions/deletions in the homopolymer tract. Our results showed that mycobacterial DnaQ binds to the ß clamp and functions synergistically with the PHP domain proofreader to correct replication errors. Furthermore, the loss of dnaQ results in replication fork dysfunction, leading to attenuated growth and increased mutagenesis on subinhibitory fluoroquinolones potentially due to increased vulnerability to fork collapse. By analyzing the sequence polymorphism of dnaQ in clinical isolates of Mycobacterium tuberculosis (Mtb), we demonstrated that a naturally evolved DnaQ variant prevalent in Mtb lineage 4.3 may enable hypermutability and is associated with drug resistance. These results establish a coproofreading model and suggest a division of labor between DnaQ and PHP domain proofreader. This study also provides real-world evidence that a mutator-driven evolutionary pathway may exist during the adaptation of Mtb.

The unfolded protein response sensor PERK mediates mechanical stress-induced maturation of focal adhesion complexes in glioblastoma cells.

Stiffening of the brain extracellular matrix (ECM) in glioblastoma promotes tumor progression. Previously, we discovered that protein kinase R (PKR)-like endoplasmic reticulum kinase (PERK) plays a role in glioblastoma stem cell (GSC) adaptation to matrix stiffness through PERK/FLNA-dependent F-actin remodeling. Here, we examined the involvement of PERK in detecting stiffness changes via focal adhesion complex (FAC) formation. Compared to control GSCs, PERK-deficient GSCs show decreased vinculin and tensin expression, while talin and integrin-ß1 remain constant. Furthermore, vimentin was also reduced while tubulin increased, and a stiffness-dependent increase of the differentiation marker GFAP expression was absent in PERK-deficient GSCs. In conclusion, our study reveals a novel role for PERK in FAC formation during matrix stiffening, which is likely linked to its regulation of F-actin remodeling.

Analysis of structural composition and antioxidant activity of traditional fermented sour meat peptides.

Chinese traditional fermented sour meat has a unique flavor and nutritional value. The antioxidant activity of sour meat peptides is related to their molecular weights, amino acid compositions, and structural characteristics. Therefore, this study explores the relationships between them. The results indicate that sour meat peptides with molecular weights <1 kDa exhibit significant antioxidant properties both in vitro and in vivo. The smaller the molecular weights, the higher the content of typical amino acids with antioxidant activity (p <0.05), and the characteristic peaks of ultraviolet absorption decrease. The absorption peak at 284.5 nm blue-shifted, and the polarity of the microenvironment increased. The peak intensity and peak area of the Raman characteristic peaks of tyrosine residues and aliphatic amino acids were enhanced. In the secondary structure, there is a high content of ß-turns and a low content of α-helix, which are closely related to the enhancement of antioxidant activity.

Magnetic resonance imaging combined with serum endolipin and galactagoglobin-3 to diagnose cerebral infarction in the elderly with diabetes mellitus.

BACKGROUND: Magnetic resonance imaging (MRI) combined with serum endothelin and galactagoglobin-3 (Gal-3) can improve the clinical diagnosis of diabetes mellitus complicated with cerebral infarction. AIM: To analyze the clinical value of MRI combined with serum endolipin and Gal-3 for the diagnosis of cerebral infarction in the elderly with diabetes mellitus. METHODS: One hundred and fifty patients with acute cerebral infarction hospitalized between January 2021 and December 2023 were divided into two groups according to comorbid diabetes mellitus, including 62 and 88 cases in the diabetic and nondiabetic cerebral infarction groups. Serum samples were collected to detect the expression of serum endolipoxins, and Gal-3, and cranial MRI was performed at admission. Differences between the two groups were compared to analyze the diagnostic value of these parameters. RESULTS: Serum endolipin and Gal-3 expression were higher in the diabetic cerebral infarction group (P < 0.05). The arterial wall area, vessel area, normalized wall index, and lumen stenosis rate were higher in the diabetic cerebral infarction group, while the rate of arterial lumen moderate and severe stenosis was 48.39% higher (36.36%, P < 0.05). The percentage of large (29.03%) and multiple infarcts (33.87%) in the diabetic cerebral infarction group was higher (13.64% and 20.45%), and the incidence rate of lacunar infarcts was lower (37.10% vs 65.91%) (P < 0.05). The total incidence of arterial plaque in patients in the diabetic cerebral infarction group was 96.77% higher (69.32%), while the incidence of necrotic lipid core plaque was 58.06% higher (26.14%) (P < 0.05). Receiver operating characteristic curve analysis was performed to assess the diagnosis utility of these techniques. MRI in combination with serum endoglin and Gal-3 had the highest area under the curve, the Yoden index, sensitivity and specificity (P < 0.05). CONCLUSION: The expression of serum endolipin and Gal-3 in elderly patients with diabetes mellitus with cerebral infarction showed an elevated trend, and the degree of luminal stenosis was severe. MRI predominantly revealed large and multiple infarct foci. This combined index examination can improve the clinical diagnosis of diabetes mellitus combined with cerebral infarction.

ß-resorcylic acid released by Limosilactobacillus reuteri protects against cisplatin-induced ovarian toxicity and infertility.

Chemotherapy-induced premature ovarian insufficiency (CIPOI) triggers gonadotoxicity in women undergoing cancer treatment, leading to loss of ovarian reserves and subfertility, with no effective therapies available. In our study, fecal microbiota transplantation in a cisplatin-induced POI mouse model reveals that a dysbiotic gut microbiome negatively impacts ovarian health in CIPOI. Multi-omics analyses show a significant decrease in Limosilactobacillus reuteri and its catabolite, ß-resorcylic acid , in the CIPOI group in comparison to healthy controls. Supplementation with L. reuteri or ß-RA mitigates cisplatin-induced hormonal disruptions, morphological damages, and reductions in follicular reserve. Most importantly, ß-RA pre-treatment effectively preserves oocyte function, embryonic development, and fetus health, thereby protecting against chemotherapy-induced subfertility. Our results provide evidence that ß-RA suppresses the nuclear accumulation of sex-determining region Y-box 7, which in turn reduces Bcl-2-associated X activation and inhibits granulosa cell apoptosis. These findings highlight the therapeutic potential of targeting the gut-ovary axis for fertility preservation in CIPOI.

Multimodal radiomics-based methods using deep learning for prediction of brain metastasis in non-small cell lung cancer with18F-FDG PET/CT images.

Objective. Approximately 57% of non-small cell lung cancer (NSCLC) patients face a 20% risk of brain metastases (BMs). The delivery of drugs to the central nervous system is challenging because of the blood-brain barrier, leading to a relatively poor prognosis for patients with BMs. Therefore, early detection and treatment of BMs are highly important for improving patient prognosis. This study aimed to investigate the feasibility of a multimodal radiomics-based method using 3D neural networks trained on18F-FDG PET/CT images to predict BMs in NSCLC patients.Approach. We included 226 NSCLC patients who underwent18F-FDG PET/CT scans of areas, including the lung and brain, prior to EGFR-TKI therapy. Moreover, clinical data (age, sex, stage, etc) were collected and analyzed. Shallow lung features and deep lung-brain features were extracted using PyRadiomics and 3D neural networks, respectively. A support vector machine (SVM) was used to predict BMs. The receiver operating characteristic (ROC) curve and F1 score were used to assess BM prediction performance.Main result. The combination of shallow lung and shallow-deep lung-brain features demonstrated superior predictive performance (AUC = 0.96 ± 0.01). Shallow-deep lung-brain features exhibited strong significance (P < 0.001) and potential predictive performance (coefficient > 0.8). Moreover, BM prediction by age was significant (P < 0.05).Significance. Our approach enables the quantitative assessment of medical images and a deeper understanding of both superficial and deep tumor characteristics. This noninvasive method has the potential to identify BM-related features with statistical significance, thereby aiding in the development of targeted treatment plans for NSCLC patients.