RESUMO
Exploring the landscape of protein phosphorylation, this investigation focuses on skin samples from LCG (Liaoning Cashmere Goats), characterized by different levels of cashmere fineness. Employing LC-MS/MS technology, we meticulously scrutinized FT-LCG (fine-type Liaoning Cashmere Goats) and CT-LCG (coarse-type Liaoning Cashmere Goats). Identifying 512 modified proteins, encompassing 1368 phosphorylated peptide segments and 1376 quantifiable phosphorylation sites, our exploration further revealed consistent phosphorylation sites in both groups. Analysis of phosphorylated peptides unveiled kinase substrates, prominently featuring Protein Kinase C, Protein Kinase B and MAPK3-MAPK1-MAPK7-NLK-group. Differential analysis spotlighted 28 disparate proteins, comprising six upregulated and twenty-two downregulated. Cluster analysis showcased the robust clustering efficacy of the two sample groups. GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) enrichment analyses underscored the significance of the purine metabolism pathway, suggesting its pivotal role in modulating cashmere fineness in LCG. Notably, through differential protein analysis, two crucial proteins were identified: HSL-X (hormone-sensitive lipase isoform X1) and KPRP (keratinocyte proline-rich protein). Further evidence supports LIPE and KPRP as key genes regulating cashmere fineness, paving the way for promising avenues in further research. These findings not only contribute to a nuanced understanding of protein-level dynamics in cashmere but also provide a theoretical foundation for the selective breeding of superior Liaoning Cashmere Goat strands.
RESUMO
The gold standard of milk is human milk, not cow milk. The present study expects to explored the comprehensive nutritional value of different kinds of milk and the differences between them through multi-omics analysis and found functional components that are more similar to human milk. This study employed untargeted LC-MS/MS metabolomics, untargeted LC-MS/MS lipidomics, and 4D label-free proteomics analysis techniques. The findings revealed substantial disparities in metabolites, lipids, and proteins among the five types of milk. Notably, pig milk exhibited a remarkable abundance of N-acetylneuraminic acid (Neu5Ac) and specific polar lipids. Yak milk stood out with significantly elevated levels of creatine and lipoprotein lipase (LPL) compared to other species. Buffalo milk boasted the highest concentrations of L-isoleucine, echinocystic acid, and alkaline phosphatase, tissue-nonspecific isozyme (ALPL). The concentrations of iminostilbene and osteopontin (OPN) were higher in cow milk.
Assuntos
Búfalos , Leite , Animais , Leite/química , Bovinos/metabolismo , Suínos , Espectrometria de Massas em Tandem , Metabolômica , Proteômica , Humanos , Leite Humano/química , Lipídeos/análise , Lipídeos/química , Cromatografia Líquida , MultiômicaRESUMO
In medical imaging, detecting tissue anomalies is vital for accurate diagnosis and effective treatment. Electrical impedance tomography (EIT) is a non-invasive technique that monitors the changes in electrical conductivity within tissues in real time. However, the current challenge lies in simply and accurately reconstructing multi-conductivity distributions. This paper introduces a layered fusion framework for EIT to enhance imaging in multi-conductivity scenarios. The method begins with pre-imaging and extracts the main object from the fuzzy image to form one layer. Then, the voltage difference in the other layer, where the local anomaly is located, is estimated. Finally, the corresponding conductivity distribution is established, and multiple layers are fused to reconstruct the multi-conductivity distribution. The simulation and experimental results demonstrate that compared to traditional methods, the proposed method significantly improves multi-conductivity separation, precise anomaly localization, and robustness without adding uncertain parameters. Notably, the proposed method has demonstrated exceptional accuracy in local anomaly detection, with positional errors as low as 1% and size errors as low as 33%, which significantly outperforms the traditional method with respective minimum errors of 9% and 228%. This method ensures a balance between the simplicity and accuracy of the algorithm. At the same time, it breaks the constraints of traditional linear methods, struggling to identify multi-conductivity distributions, thereby providing new perspectives for clinical EIT.
RESUMO
The periosteum plays a critical role in bone repair and is significantly influenced by the surrounding immune microenvironment. In this study, we employed 10× single-cell RNA sequencing to create a detailed cellular atlas of the swine cranial periosteum, highlighting the cellular dynamics and interactions essential for cranial bone injury repair. We noted that such injuries lead to an increase in M2 macrophages, which are key in modulating the periosteum's immune response and driving the bone regeneration process. These macrophages actively recruit periosteal stromal cells (PSCs) by secreting Neuregulin 1 (NRG1), a crucial factor in initiating bone regeneration. This recruitment process emphasizes the critical role of PSCs in effective bone repair, positioning them as primary targets for therapeutic interventions. Our results indicate that enhancing the interaction between M2 macrophages and PSCs could significantly improve the outcomes of treatments aimed at cranial bone repair and regeneration.
RESUMO
The inducing activation event of secondary hair follicle (SHF)-stem cells is considered a key biological process in the SHF regeneration, and the morphogenesis of cashmere fiber in cashmere goats. The miR-361-5p was essentially implicated in the induced activation of SHF-stem cells of cashmere goats, but its functional mechanisms are unclear. Here, we confirmed miR-361-5p was significantly downregulated in anagen SHF bugle of cashmere goats compared with that at telogen, and miR-361-5p expression was significantly lower in SHF-stem cells after activation than its counterpart before activation. Further, we found that miR-361-5p could negatively regulate the induced activation event of SHF-stem cells in cashmere goats. Mechanistically, through dual-luciferase reporter assays, miR-361-5p specifically bound with FOXM1 mRNA in SHF-stem cells of cashmere goats and negatively regulated the expression of FOXM1 gene. Also, through overexpression/knockdown analysis of FOXM1 gene, our results indicated that FOXM1 upregulated the expression of Wnt/ß-catenin pathway related genes in SHF-stem cells. Moreover, based on TOP/FOP-flash Wnt report assays, the knockdown of miR-361-5p promotes the Wnt/ß-catenin pathway activation through upregulating the FOXM1 expression in SHF-stem cells. Finally, we demonstrated that miR-361-5p negatively regulated the induced activation of SHF-stem cells through FOXM1 mediated Wnt/ß-catenin pathway in cashmere goats.
Assuntos
Proteína Forkhead Box M1 , Cabras , Folículo Piloso , MicroRNAs , Células-Tronco , Via de Sinalização Wnt , Animais , Cabras/genética , MicroRNAs/genética , MicroRNAs/metabolismo , Via de Sinalização Wnt/genética , Folículo Piloso/metabolismo , Proteína Forkhead Box M1/genética , Proteína Forkhead Box M1/metabolismo , Células-Tronco/fisiologia , Células-Tronco/metabolismo , Técnicas de Silenciamento de GenesRESUMO
This study systematically reviewed the application of large language models (LLMs) in medicine, analyzing 550 selected studies from a vast literature search. LLMs like ChatGPT transformed healthcare by enhancing diagnostics, medical writing, education, and project management. They assisted in drafting medical documents, creating training simulations, and streamlining research processes. Despite their growing utility in assisted diagnosis and improving doctor-patient communication, challenges persisted, including limitations in contextual understanding and the risk of over-reliance. The surge in LLM-related research indicated a focus on medical writing, diagnostics, and patient communication, but highlighted the need for careful integration, considering validation, ethical concerns, and the balance with traditional medical practice. Future research directions suggested a focus on multimodal LLMs, deeper algorithmic understanding, and ensuring responsible, effective use in healthcare.
RESUMO
The myostatin (MSTN) gene exhibits significant nucleotide sequence variations in sheep, impacting growth characteristics and muscular traits of the body. However, its influence on specific growth traits in some sheep remains to be further elucidated. This study utilized single nucleotide polymorphism sequence analysis to investigate the role of the MSTN gene in meat production performance across four sheep breeds: Charolais sheep, Australian White sheep, crossbreeds of Australian White and Small-tailed Han, and crossbreeds of Charolais and Small-tailed Han. At a SNP locus of the MSTN gene, the C2361T site was identified, with three genotypes detected: CC, CT, and TT, among which CC predominated. Gene substitution effect analysis revealed that replacing C with T could elevate the phenotypic value. Comparative analysis of data from different genotypes within the same breed highlighted the superiority of CC and TT genotypes in phenotypic values, underscoring the significance of specific genotypes in influencing key traits. Contrasting the performance of different genotypes across breeds, Charolais sheep and Charolais Han hybrids demonstrated superiority across multiple indicators, offering valuable insights for breeding new sheep varieties. Analysis of gender effects on growth characteristics indicated that ewes exhibited significantly wider chest, waist, and hip widths compared to rams, while rams displayed better skeletal growth and muscle development. Additionally, the MSTN gene also exerted certain effects on lamb growth characteristics, with the CC genotype closely associated with weight. These findings not only contribute crucial insights for sheep breeding but also pave the way for future research exploring the interaction of this gene with others.
RESUMO
Intimal hyperplasia is a complicated pathophysiological phenomenon attributable to in-stent restenosis, and the underlying mechanism remains unclear. Interleukin enhancer-binding factor 3 (ILF3), a double-stranded RNA-binding protein involved in regulating mRNA stability, has been recently demonstrated to assume a crucial role in cardiovascular disease; nevertheless, its impact on intimal hyperplasia remains unknown. In current study, we used samples of human restenotic arteries and rodent models of intimal hyperplasia, we found that vascular smooth muscle cell (VSMC) ILF3 expression was markedly elevated in human restenotic arteries and murine ligated carotid arteries. SMC-specific ILF3 knockout mice significantly suppressed injury induced neointimal formation. In vitro, platelet-derived growth factor type BB (PDGF-BB) treatment elevated the level of VSMC ILF3 in a dose- and time-dependent manner. ILF3 silencing markedly inhibited PDGF-BB-induced phenotype switching, proliferation, and migration in VSMCs. Transcriptome sequencing and RNA immunoprecipitation sequencing depicted that ILF3 maintained its stability upon binding to the mRNA of the high-mobility group box 1 protein (HMGB1), thereby exerting an inhibitory effect on the transcription of dual specificity phosphatase 16 (DUSP16) through enhanced phosphorylation of signal transducer and activator of transcription 3 (STAT3). Therefore, the results both in vitro and in vivo indicated that the loss of ILF3 in VSMC ameliorated neointimal hyperplasia by regulating the STAT3/DUSP16 axis through the degradation of HMGB1 mRNA. Our findings revealed that vascular injury activates VSMC ILF3, which in turn promotes intima formation. Consequently, targeting specific VSMC ILF3 may present a potential therapeutic strategy for ameliorating cardiovascular restenosis.
Assuntos
Proteína HMGB1 , Hiperplasia , Camundongos Knockout , Músculo Liso Vascular , Miócitos de Músculo Liso , Proteínas do Fator Nuclear 90 , Estabilidade de RNA , Fator de Transcrição STAT3 , Túnica Íntima , Animais , Humanos , Masculino , Camundongos , Movimento Celular , Proliferação de Células , Modelos Animais de Doenças , Regulação da Expressão Gênica , Proteína HMGB1/metabolismo , Proteína HMGB1/genética , Músculo Liso Vascular/metabolismo , Músculo Liso Vascular/patologia , Miócitos de Músculo Liso/metabolismo , Neointima/metabolismo , Neointima/patologia , Proteínas do Fator Nuclear 90/metabolismo , Proteínas do Fator Nuclear 90/genética , RNA Mensageiro/genética , RNA Mensageiro/metabolismo , Transdução de Sinais , Fator de Transcrição STAT3/metabolismo , Túnica Íntima/metabolismo , Túnica Íntima/patologiaRESUMO
The milk proteins from samples of 13 different animals were identified utilizing 4D-Label-Free proteomics technology, leading to the identification of a substantial number of proteins. Among the various samples, Chinese people (CHP) milk proteins exhibited the highest count, with 1149 distinct proteins. Simultaneously, we identified common proteins present in these animal milk. It's notable presence in goat milk contributes to enhancing infant infection resistance, showcasing the beneficial role of lactoperoxidase. Galectin-3 binding protein (Gal-3BP) and tetraspanin in human milk are significantly higher than those in other animals, which determine the prominent antiviral effect of human milk and the important processes related to cell transduction. Furthermore, human milk, camel milk, goat milk and sheep milk proved to be rich sources of milk fat globule membrane (MFGM) proteins. The insights obtained from this study can serve as a foundational framework for exploring the role of different animal milk proteins in disease treatment and the composition of infant formula.
RESUMO
This paper introduces a sensitivity matrix decomposition regularization (SMDR) method for electric impedance tomography (EIT). Using k-means clustering, the EIT-reconstructed image can be divided into four clusters, derived based on image features, representing posterior information. The sensitivity matrix is then decomposed into distinct work areas based on these clusters. The elimination of smooth edge effects is achieved through differentiation of the images from the decomposed sensitivity matrix and further post-processing reliant on image features. The algorithm ensures low computational complexity and avoids introducing extra parameters. Numerical simulations and experimental data verification highlight the effectiveness of SMDR. The proposed SMDR algorithm demonstrates higher accuracy and robustness compared to the typical Tikhonov regularization and the iterative penalty term-based regularization method (with an improvement of up to 0.1156 in correlation coefficient). Moreover, SMDR achieves a harmonious balance between image fidelity and sparsity, effectively addressing practical application requirements.
RESUMO
D-ribose, an ubiquitous pentose compound found in all living cells, serves as a vital constituent of numerous essential biomolecules, including RNA, nucleotides, and riboflavin. It plays a crucial role in various fundamental life processes. Within the cellular milieu, exogenously supplied D-ribose can undergo phosphorylation to yield ribose-5-phosphate (R-5-P). This R-5-P compound serves a dual purpose: it not only contributes to adenosine triphosphate (ATP) production through the nonoxidative phase of the pentose phosphate pathway (PPP) but also participates in nucleotide synthesis. Consequently, D-ribose is employed both as a therapeutic agent for enhancing cardiac function in heart failure patients and as a remedy for post-exercise fatigue. Nevertheless, recent clinical studies have suggested a potential link between D-ribose metabolic disturbances and type 2 diabetes mellitus (T2DM) along with its associated complications. Additionally, certain in vitro experiments have indicated that exogenous D-ribose exposure could trigger apoptosis in specific cell lines. This article comprehensively reviews the current advancements in D-ribose's digestion, absorption, transmembrane transport, intracellular metabolic pathways, impact on cellular behaviour, and elevated levels in diabetes mellitus. It also identifies areas requiring further investigation.
Assuntos
Diabetes Mellitus Tipo 2 , Insuficiência Cardíaca , Doenças Metabólicas , Humanos , Diabetes Mellitus Tipo 2/tratamento farmacológico , Ribose/metabolismo , Trifosfato de AdenosinaRESUMO
In this investigation, a composite material comprising Ti-MOF and chitosan, denoted as BD-MOF(Ti)@CS/Fe3O4, was successfully designed for the efficient adsorption of Pb(II) from aqueous solutions. A comprehensive array of characterization techniques, including SEM, XRD, BET, FT-IR, and XPS, were meticulously employed to scrutinize the structural attributes and morphological features of the Pb(II) adsorbent. Notably, the material exhibits adaptability to a broad pH range, with adsorption efficiency reaching 99 % between pH 3 and 6. Kinetic studies reveal that the adsorption process of Pb(II) by BD-MOF(Ti)@CS/Fe3O4 adheres closely to a pseudo-second-order kinetic model. Impressively, within a short duration of 40 min, the adsorption efficiency can reach 85 %. Furthermore, the adsorption isotherm aligns with the Hill isotherm model, signifying cooperative adsorption. This observation underscores the synergistic interplay among the functional groups on the surface of BD-MOF(Ti)@CS/Fe3O4 in capturing Pb(II). As per the Hill model, the theoretical maximum capacity was an impressive 944.9 mg/g. Thermodynamic assessments suggested that the adsorption process was spontaneous, entropy increasing and exothermic. Even in the presence of various interfering ions, BD-MOF(Ti)@CS/Fe3O4 exhibited robust adsorption performance, thereby affirming its utility in complex environments. Moreover, the material demonstrates noteworthy reusability, sustaining effective Pb(II) removal across five consecutive cycles in aqueous solutions.
Assuntos
Quitosana , Poluentes Químicos da Água , Adsorção , Quitosana/química , Chumbo , Espectroscopia de Infravermelho com Transformada de Fourier , Cinética , Titânio , Fenômenos Magnéticos , Poluentes Químicos da Água/química , Concentração de Íons de HidrogênioRESUMO
The cashmere, a kind of nature protein fiber, is one of the main use of cashmere goats. The induced activation of secondary hair follicle (SHF) stem cells by the dermal papilla cell-derived signals is a key biological process for the morphogenesis and growth of cashmere fiber in cashmere goats. Previously, the circRNA-ERCC6 (circERCC6) was identified from cashmere goat SHFs; however, its biological significance is unclear in the SHF physiology process of cashmere goats. In this study, we found that circERCC6 exhibited significantly higher expression at anagen SHF bulge compared with the counterpart of telogen and harbored three m6A modified sites (named m6A-685, m6A-862, and m6A-995) through methylation immunoprecipitation using a real-time quantitative polymerase chain reaction (Me-RIP-qPCR) technique. The knockdown experiments of circERCC6 in SHF stem cells showed that circERCC6 positively regulates the induced activation of SHF stem cells in cashmere goats. Through a dual-luciferase reporter assay, we demonstrated that m6A-modified circERCC6 (m6A-circERCC6) sponged miR-412-3p to upregulate the expression of BNC2 mRNA in SHFstem cells. Through m6A-deficient mutant assay in circERCC6 knockdown SHF stem cells, we further showed that m6A modification within circERCC6 is required to mediate the miR-412-3p/BNC2 axis to finally promote the proper induced activation of SHF stem cells in cashmere goats.
RESUMO
OBJECTIVE: The incidence of pulmonary nodules has been increasing over the past 30 years. Different types of nodules are associated with varying degrees of malignancy, and they engender inconsistent treatment approaches. Therefore, correct distinction is essential for the optimal treatment and recovery of the patients. The commonly-used medical imaging methods have limitations in distinguishing lung nodules to date. A new approach to this problem may be provided by electrical properties of lung nodules. Nevertheless, difference identification is the basis of correct distinction. So, this paper aims to investigate the differences in electrical properties between various lung nodules. METHODS: At variance with existing studies, benign samples were included for analysis. A total of 252 specimens were collected, including 126 normal tissues, 15 benign nodules, 76 adenocarcinomas, and 35 squamous cell carcinomas. The dispersion properties of each tissue were measured over a frequency range of 100 Hz to 100 MHz. And the relaxation mechanism was analyzed by fitting the Cole-Cole plot. The corresponding equivalent circuit was estimated accordingly. RESULTS: Results validated the significant differences between malignant and normal tissue. Significant differences between benign and malignant lesions were observed in conductivity and relative permittivity. Adenocarcinomas and squamous cell carcinomas are significantly different in conductivity, first-order, second-order differences of conductivity, α-band Cole-Cole plot parameters and capacitance of equivalent circuit. The combination of the different features increased the tissue groups' differences measured by Euclidean distance up to 94.7%. CONCLUSION AND SIGNIFICANCE: In conclusion, the four tissue groups reveal dissimilarity in electrical properties. This characteristic potentially lends itself to future diagnosis of non-invasive lung cancer.
Assuntos
Adenocarcinoma , Carcinoma de Células Escamosas , Neoplasias Pulmonares , Lesões Pré-Cancerosas , Humanos , Neoplasias Pulmonares/diagnóstico por imagem , Pulmão , Condutividade Elétrica , Carcinoma de Células Escamosas/diagnóstico por imagemRESUMO
Lipids are essential organic components in milk and have been associated with various health benefits for newborns. However, a comprehensive analysis of lipid profiles across multiple species and levels has been lacking. In this study, we employed liquid chromatography-tandem mass spectrometry (LC-MS/MS) to accurately determine the absolute content of lipid molecules. It revealed that ruminants exhibit a higher concentration of short-chain fatty acids compared to non-ruminants. Additionally, we identified ALC (camel), MGH (horse), and DZD (donkey) as species that display similarities to components found in human milk fat. Remarkably, it reveals that porcine milk fat is characterized by long chain lengths, low saturation, and a high proportion of essential fatty acids. PS (22:5_18:2) could potentially serve as a biomarker in porcine milk. These unique characteristics present potential opportunities for the utilization of porcine milk. Overall, our findings provide valuable insights into the lipidomics profiles of milk from different species.
RESUMO
BACKGROUND: Numerous factors influence the growth and development of cashmere. Existing research on cashmere has predominantly emphasized a single omics level. Integrating multi-omics analyses can offer a more comprehensive understanding by encompassing the entire spectrum. This study more accurately and comprehensively identified the key factors influencing cashmere fineness using multi-omics analysis. METHODS: This study used skin tissues of coarse cashmere type (CT_LCG) and fine cashmere type Liaoning cashmere goats (FT_LCG) for the analysis. This study employed an integrated approach involving transcriptomics, translatomics, proteomics, and metabolomics to identify substances associated with cashmere fineness. The findings were validated using parallel reaction monitoring (PRM) and multiple reaction monitoring (MRM) techniques. RESULTS: The GO functional enrichment analysis identified three common terms: multicellular organismal process, immune system process, and extracellular region. Furthermore, the KEGG enrichment analysis uncovered the involvement of the arachidonic acid metabolic pathway. Protein expression trends were verified using PRM technology. The expression trends of KRT79, as confirmed by PRM, were consistent with those observed in TMT proteomics and exhibited a positive regulatory effect on cashmere fineness. Metabolite expression trends were confirmed using MRM technology. The expression trends of 9 out of 15 validated metabolites were in agreement with those identified in the non-targeted metabolomics analysis. CONCLUSIONS: This study employed multi-omics analysis to identify key regulators of cashmere fineness, including PLA2G12A, KRT79, and prostaglandin B2. The findings of this study offer valuable data and establish a theoretical foundation for conducting comprehensive investigations into the molecular regulatory mechanisms and functional aspects of cashmere fineness.
Assuntos
Multiômica , Pele , Animais , Pele/metabolismo , Cabras/genéticaRESUMO
Breast milk plays a crucial role in the taste development of infants, which cannot be replicated by other mammalian milk or formulas. This study aimed to identify and characterize the flavor substances in 15 different types of milk and analyze the differences among them. The results showed that human milk contained high levels of esters, particularly fatty acid ethyl esters, which contribute to its unique flavor. The four substances that had the highest flavor contribution in all species were identified as 2,3-butanedione, trimethylamine, isophorone, and acetaldehyde. Furthermore, the analysis of differences revealed that thermal-oxidation of lipids could explain the variation between human milk and other species in terms of flavor compounds. The key differential flavor compounds identified in milk from all species were trimethylamine, propanal, 1-pentanol, pyridine 2-methyl, and 2-butanone. These findings can potentially aid in developing formulas that better meet the taste needs of infants.
RESUMO
Liaoning cashmere goat (LCG) is one of the excellent cashmere goat breeds in China. Because of its larger size, better cashmere, and better cashmere production performance, people pay special attention to it. This article mainly studied the relationship between SNP loci of LIPE gene and ITGB4 gene and milk production, cashmere production and body measurement traits of LCGs. We further identified potential SNP loci by PCR-Seq polymorphism detection and gene sequence comparison of LIPE and ITGB4 genes. Further, we use SPSS and SHEsis software to analyze their relationship to production performance. The consequence indicated that CC genotype of LIPE gene T16409C locus was dominant genotype in milk production and cashmere production, while CT genotype of LIPE gene T16409C locus was dominant in body size. The CT genotype of C168T locus of ITGB4 gene is the dominant genotype of body type and cashmere production, while the dominant genotype of milk production is TT genotype. Through joint analysis, in haploid combinations, H1H2:CCCT is the dominant haplotype combination in cashmere fineness. H3H4:TTCT is a dominant haplotype combination of milk production traits and body measurement traits. These dominant genotypes can provide a reliable basis for the study of production performance of LCG.
Assuntos
Cabras , Polimorfismo de Nucleotídeo Único , Animais , Polimorfismo de Nucleotídeo Único/genética , Cabras/genética , Leite , Fenótipo , GenótipoRESUMO
Mature milk, as a nutrient-rich endogenous metabolite, has various beneficial effects on the human body. In order to investigate the specific nutrients provided by different dairy products to humans, we used UHPLC-Q-TOF MS to analyze the highly significantly differentially expressed metabolites in 13 species of mammalian mature milk, which were grouped into 17 major metabolite classes with 1992 metabolites based on chemical classification. KEGG shows that 5 pathways in which differentially significant metabolites are actively involved are ABC transporters, Purine metabolism, Pyrimidine metabolism, Phosphotransferase system, Galactose metabolism. The study found that pig milk and goat milk are closer to human milk and contain more nutrients that are beneficial to human health, followed by camel milk and cow milk. In the context of dairy production, the development of goat milk is more likely to meet human needs and health.
RESUMO
Myocardial ischemia diagnosis with CT perfusion imaging (CTP) is important in coronary artery disease management. Traditional analysis procedure is time-consuming and error-prone due to the semi-manual and operator-dependent nature. To improve the diagnostic performance, a deep learning-based, fully automatic, and clinical-ready framework was developed. Two collaborating deep learning networks including a 3D U-Net for left ventricle segmentation and a CNN for anatomical landmarks detection were trained on 276 subjects. With our processing framework, the 17-segment left ventricular model was automatically generated conformed to the clinical standard. Myocardial blood flow computed by commercial software was extracted within each segment and visualized against the bull's eye plot. The performance was validated on another 45 subjects. Coronary angiography and invasive fractional flow reserve measurements were also performed in these patients to serve as the gold standard for myocardial ischemia diagnosis. As a result, the diagnostic accuracy for our method was 81.08%, much higher than that for commercially available CTP analysis software (56.75%), and our method demonstrated a higher consistency (Kappa coefficient 0.759 vs. 0.585). Besides, the average processing time of our method was much lower (30 ± 10.5 s/subject vs. over 30 min/subject). In conclusion, the proposed deep learning-based framework could be a promising tool for assisting CTP analysis.