Search

ObjectiveTo explore the correlation of RELL1 expression with clinical pathological features and prognosis of patients with colon cancer. MethodsImmunohistochemical experiments of the RELL1 protein were performed on tissue chips from 80 colon cancer tissues and 80 adjacent tissues. The relationship between different expression levels of RELL1 protein and clinical pathological parameters was compared. Univariate and multivariate Cox risk proportional regression analyses were conducted on factors affecting the survival of patients with colon cancer. Kaplan-Meier survival curve analysis was conducted on the survival rates of patients with colon cancer and different levels of RELL1 expression. Log rank test was performed to evaluate differences in survival rates. ResultsThe expression of RELL1 in colon cancer tissues was lower than that in adjacent tissues (P

Purpose: Radiation-induced lymphopenia (RIL) is common during chemoradiation therapy. Severe lymphopenia is associated with reduced survival. Proton beam therapy (PBT), with its substantially more compact dose distributions, spares circulating lymphocytes and immune organs at risk to a greater extent than photon therapy. Recent studies comparing PBT to photon radiation therapy, specifically intensity-modulated radiation therapy (IMRT) for esophageal cancer (EC), showed that the incidence of grade 4 RIL (G4RIL) is significantly reduced among patients receiving PBT for EC. However, whether the extent of this reduction has a direct causative link with improved survival is unknown. This study applies causal mediation analysis to answer this question. Methods and Materials: We retrospectively assessed 734 patients treated with concurrent chemoradiation therapy for biopsy-proven EC from 2004 to 2017. To address the potential for bias in the choice of radiation modality, propensity score analysis was used to evaluate and reduce imbalances between the PBT and IMRT cohorts. Causal mediation analysis was applied to decompose the total effect of radiation modality on overall survival (OS) into indirect (mediated through G4RIL) and direct effects. Results: We found that PBT was associated with a significantly lower incidence of G4RIL and prolonged OS compared with IMRT (odds ratio, 0.41; 95% CI, 0.28-0.60; P < .001). In the propensity-matched cohort of 506 patients (253 PBT, 253 IMRT), G4RIL risk reduction with PBT versus IMRT translated into a 5% reduction in the relative rate of death (P = .032). Mediation of G4RIL explained ∼14.5% of the difference in OS. Conclusions: G4RIL was found to mediate survival; however, a statistically significant direct effect of PBT on survival was not observed. In other words, the statistical significance of survival benefit from protons over photons in this EC cohort was lost in the absence of G4RIL risk reduction.

Summary: Orofacial clefts (OFCs) are among the most common human congenital birth defects. Previous multiethnic studies have identified dozens of associated loci for both cleft lip with or without cleft palate (CL/P) and cleft palate alone (CP). Although several nearby genes have been highlighted, the “casual” variants are largely unknown. Here, we developed DeepFace, a convolutional neural network model, to assess the functional impact of variants by SNP activity difference (SAD) scores. The DeepFace model is trained with 204 epigenomic assays from crucial human embryonic craniofacial developmental stages of post-conception week (pcw) 4 to pcw 10. The Pearson correlation coefficient between the predicted and actual values for 12 epigenetic features achieved a median range of 0.50–0.83. Specifically, our model revealed that SNPs significantly associated with OFCs tended to exhibit higher SAD scores across various variant categories compared to less related groups, indicating a context-specific impact of OFC-related SNPs. Notably, we identified six SNPs with a significant linear relationship to SAD scores throughout developmental progression, suggesting that these SNPs could play a temporal regulatory role. Furthermore, our cell-type specificity analysis pinpointed the trophoblast cell as having the highest enrichment of risk signals associated with OFCs. Overall, DeepFace can harness distal regulatory signals from extensive epigenomic assays, offering new perspectives for prioritizing OFC variants using contextualized functional genomic features. We expect DeepFace to be instrumental in accessing and predicting the regulatory roles of variants associated with OFCs, and the model can be extended to study other complex diseases or traits.

Abstract An edge-cloud computing collaborative dust concentration detection architecture is proposed for real-time operation of intelligent algorithms to reduce the warning delay. And, an end-to-end three-channel convolutional neural network (E2E-SCNN) method is proposed in the paper to facilitate intelligent monitoring and management of dust concentration in tobacco production workshops. This model, which includes three sub-networks-a local feature branch, a global feature branch, and a spatial feature branch, learns the detail texture, overall layout, and spatial distribution information of the input image respectively. A fusion of the three complementary features is performed at the end of the network for the final dust concentration regression prediction. The design, when compared with the single network structure that directly regresses the entire image, is shown to more fully represent the overall information of the image and enhance monitoring performance. A richly annotated image dataset of tobacco production workshops is constructed to verify the effectiveness of the proposed method. The prediction error of E2E-SCNN is compared with existing image estimation algorithms, dual-channel networks, and other methods on this dataset using indicators such as Mean Absolute Error (MAE) and $${R}^{2}$$ R 2 . It is shown by the results that excellent performance is achieved by the E2E-SCNN algorithm, significantly surpassing other comparison methods. The paper demonstrates that the accuracy and robustness of dust concentration prediction can be greatly improved by using a three-channel convolutional neural network spatial information monitoring framework. This achievement provides an effective means for dust supervision and governance during the tobacco production process and offers a technical route that can be referred to for image analysis tasks in other similar fields.

Abstract To achieve real‐time and effective prediction of industrial site risks, this paper proposes an industrial risk prediction framework for multimodal data based on edge computing. First, the authors gather and annotate industrial risk multimodal data that consist of text descriptions, images, and videos. Then, the authors transfer the data to the edge server, and apply deep learning models such as Bidirectional Encoder Representations from Transformers (BERT), ResNet etc., to extract features and learn representations for text, image, and video data respectively. The authors input the fused feature data into an enhanced long short term memory (LSTM) model and train it on the dataset. Finally, the authors perform the risk prediction based on the collected multimodal data. The experimental results demonstrate that the method proposed in this paper exhibits superior performance, achieving a 1.4% enhancement in predictive accuracy.

Abstract Background Monitoring blood pressure and peripheral capillary oxygen saturation plays a crucial role in healthcare management for patients with chronic diseases, especially hypertension and vascular disease. However, current blood pressure measurement methods have intrinsic limitations; for instance, arterial blood pressure is measured by inserting a catheter in the artery causing discomfort and infection. Method Photoplethysmogram (PPG) signals can be collected via non-invasive devices, and therefore have stimulated researchers’ interest in exploring blood pressure estimation using machine learning and PPG signals as a non-invasive alternative. In this paper, we propose a Transformer-based deep learning architecture that utilizes PPG signals to conduct a personalized estimation of arterial systolic blood pressure, arterial diastolic blood pressure, and oxygen saturation. Results The proposed method was evaluated with a subset of 1,732 subjects from the publicly available ICU dataset MIMIC III. The mean absolute error is 2.52 ± 2.43 mmHg for systolic blood pressure, 1.37 ± 1.89 mmHg for diastolic blood pressure, and 0.58 ± 0.79% for oxygen saturation, which satisfies the requirements of the Association of Advancement of Medical Instrumentation standard and achieve grades A for the British Hypertension Society standard. Conclusions The results indicate that our model meets clinical standards and could potentially boost the accuracy of blood pressure and oxygen saturation measurement to deliver high-quality healthcare.

The increasing load difference between peak and off-peak seasons has led to the unstable operation of China’s power systems in recent years. In addition, the current electricity package structure system of China, which is gradually liberalizing its electricity market, remains imperfect. This paper establishes a model of residential electricity packages that considers load difference penalty during peak and off-peak seasons. First, consumers’ utility of different packages is established and thus the optimal is determined by maximizing the utility function. Then, a mathematical model focusing on the minimization of load difference of retail electricity packages during peak and off-peak seasons is established. We thus design a particle-swarm based algorithm to solve the problem. Through simulation results of several numerical examples, we demonstrate that the profit of retailer is increased by 37.35% and Load difference during peak and off-peak seasons is reduced by 78.01%, improving the power demand curve of users and the operation efficiency of power systems.

In bearing fault diagnosis, machine learning methods have been proven effective on the basis of the heterogeneous features extracted from multiple domains, including deep representation features. However, comparatively little research has been performed on fusing these multi-domain heterogeneous features while dealing with the interrelation and redundant problems to precisely discover the bearing faults. Thus, in the current study, a novel diagnostic method, namely the method of incorporating heterogeneous representative features into the random subspace, or IHF-RS, is proposed for accurate bearing fault diagnosis. Primarily, via signal processing methods, statistical features are extracted, and via the deep stack autoencoder (DSAE), deep representation features are acquired. Next, considering the different levels of predictive power of features, a modified lasso method incorporating the random subspace method is introduced to measure the features and produce better base classifiers. Finally, the majority voting strategy is applied to aggregate the outputs of these various base classifiers to enhance the diagnostic performance of the bearing fault. For the proposed method’s validity, two bearing datasets provided by the Case Western Reserve University Bearing Data Center and Paderborn University were utilized for the experiments. The results of the experiment revealed that in bearing fault diagnosis, the proposed method of IHF-RS can be successfully utilized.

Abstract Background Abnormal calcium homeostasis related to the development of hypertension. As the key regulator of intracellular calcium concentration, voltage-dependent calcium channels (VDCCs), the variations in these genes may have important effects on the development of hypertension. Here we evaluate VDCCs variability with respect to hypertension in the Dai ethnic group of China. Methods A total of 1034 samples from Dai individuals were collected, of which 495 were used as cases, and 539 were used as controls. Blood pressure was measured using a standard mercury measurement method, three times with a rest for 5 min, and the average was used for analyses. Seventeen single nucleotide polymorphisms (SNPs) in the four protein-coding genes (CACNA1A, CACNA1C, CACNA1S, CACNB2) of VDCCs were identified by multiplex PCR-SNP typing technique. Chi-square tests and regression models were used to analyse the associations of SNPs with hypertension. Results The results of chi-square tests showed that the allele frequencies of 5 SNPs were significantly different between the case and the control groups (P

ABSTRACT A novel Klebsiella pneumoniae carbapenemase (KPC) variant, KPC-93, was identified in two Klebsiella pneumoniae clinical isolates from a patient from China treated with ceftazidime-avibactam. KPC-93 possessed a five-amino-acids insertion (Pro-Asn-Asn-Arg-Ala) between Ambler positions 267 and 268 in KPC-2. Cloning and expression of the blaKPC-93 gene in Escherichia coli, followed by determination of minimum inhibitory concentration (MIC) values and kinetic parameters, showed that KPC-93 exhibited increased resistance to ceftazidime-avibactam, but a drastic decrease in carbapenemase activity. Our data highlight that a KPC variant conferring resistance to ceftazidime-avibactam could be easily induced by ceftazidime-avibactam treatment and that actions are required to control dissemination of these determinants. IMPORTANCE Ceftazidime-avibactam (CZA) is a novel β-lactam/β-lactamase inhibitor combination with activity against serine β-lactamases, including the Ambler class A enzyme KPC. However, during recent years, there have been increasing reports of emergence of new KPC variants that could confer resistance to CZA. This has limited its clinical application. Here, we reported a new KPC variant, KPC-93, that could confer CZA resistance. KPC-93 possessed a five-amino-acids insertion (Pro-Asn-Asn-Arg-Ala) between Ambler positions 267 and 268 in KPC-2. Our findings have revealed the potential risk of blaKPC gene mutations associated with CZA exposure over a short period of time.

Deep neural network (DNN) has been recently applied to many safety-critical environments. Unfortunately, recent research has proven that DNN can be vulnerable to well-designed examples, called adversarial examples. Adversarial examples can easily fool a well-performed deep learning model with little perturbations imperceptible to humans. In this paper, to tackle the DNN security issue, we propose a Model Adversarial Score (MAS) index to evaluate the vulnerability of a deep neural network, and introduce a deep learning vulnerability assessment system (SecureAS) using adversarial samples to assess the vulnerability and risk of a trained DNN in a blackbox way. We also present two adversary algorithms (FGNM and PINM) that provide better adversary images with the similar attack effect compared to existing approaches like FGSM and BIM. Our experimental results confirm the effectiveness of MAS algorithm, SecureAS, FGNM and PINM.

In recent research, supervised image clustering based on Graph Neural Networks (GNN) connectivity prediction has demonstrated considerable improvements over traditional clustering algorithms. However, existing supervised image clustering algorithms are usually time-consuming and limit their applications. In order to infer the connectivity between image instances, they usually created a subgraph for each image instance. Due to the creation and process of a large number of subgraphs as the input of GNN, the computation overheads are enormous. To address the high computation overhead problem in the GNN connectivity prediction, we present a time-efficient and effective GNN-based supervised clustering framework based on density division namely DDC-GNN. DDC-GNN divides all image instances into high-density parts and low-density parts, and only performs GNN subgraph connectivity prediction on the low-density parts, resulting in a significant reduction in redundant calculations. We test two typical models in the GNN connectivity prediction module in the DDC-GNN framework, which are the graph convolutional networks (GCN)-based model and the graph auto-encoder (GAE)-based model. Meanwhile, adaptive subgraphs are generated to ensure sufficient contextual information extraction for low-density parts instead of the fixed-size subgraphs. According to the experiments on different datasets, DDC-GNN achieves higher accuracy and is almost five times quicker than those without the density division strategy.

BackgroundThere is an unmet need for noninvasive imaging markers that can help identify the aggressive subtype(s) of pancreatic ductal adenocarcinoma (PDAC) at diagnosis and at an earlier time point, and evaluate the efficacy of therapy prior to tumor reduction. In the past few years, there have been two major developments with potential for a significant impact in establishing imaging biomarkers for PDAC and pancreatic cancer premalignancy: (1) hyperpolarized metabolic (HP)-magnetic resonance (MR), which increases the sensitivity of conventional MR by over 10,000-fold, enabling real-time metabolic measurements; and (2) applications of artificial intelligence (AI). ObjectiveOur objective of this review was to discuss these two exciting but independent developments (HP-MR and AI) in the realm of PDAC imaging and detection from the available literature to date. MethodsA systematic review following the PRISMA extension for Scoping Reviews (PRISMA-ScR) guidelines was performed. Studies addressing the utilization of HP-MR and/or AI for early detection, assessment of aggressiveness, and interrogating the early efficacy of therapy in patients with PDAC cited in recent clinical guidelines were extracted from the PubMed and Google Scholar databases. The studies were reviewed following predefined exclusion and inclusion criteria, and grouped based on the utilization of HP-MR and/or AI in PDAC diagnosis. ResultsPart of the goal of this review was to highlight the knowledge gap of early detection in pancreatic cancer by any imaging modality, and to emphasize how AI and HP-MR can address this critical gap. We reviewed every paper published on HP-MR applications in PDAC, including six preclinical studies and one clinical trial. We also reviewed several HP-MR–related articles describing new probes with many functional applications in PDAC. On the AI side, we reviewed all existing papers that met our inclusion criteria on AI applications for evaluating computed tomography (CT) and MR images in PDAC. With the emergence of AI and its unique capability to learn across multimodal data, along with sensitive metabolic imaging using HP-MR, this knowledge gap in PDAC can be adequately addressed. CT is an accessible and widespread imaging modality worldwide as it is affordable; because of this reason alone, most of the data discussed are based on CT imaging datasets. Although there were relatively few MR-related papers included in this review, we believe that with rapid adoption of MR imaging and HP-MR, more clinical data on pancreatic cancer imaging will be available in the near future. ConclusionsIntegration of AI, HP-MR, and multimodal imaging information in pancreatic cancer may lead to the development of real-time biomarkers of early detection, assessing aggressiveness, and interrogating early efficacy of therapy in PDAC.

Objective Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) is a primary cause of hospitalization and death in COPD. Serum CA-125 and red blood cell distribution width (RDW) are related to AECOPD. We investigated correlations between serum markers and AECOPD. Methods In total, 132 patients with AECOPD were included from January 2017 to December 2019. Participants were followed for 1 year. Patients were assigned to the poor prognosis (n = 40) or good prognosis (n = 92) group. We collected serum samples and general clinical information and conducted routine blood tests. We used logistic regression, receiver operating characteristic (ROC), and area under the ROC curve (AUC) analyses to assess differences between groups. Results We found significant differences between groups (odds ratio, 95% confidence interval) for age (1.046, 1.005–1.09), RDW (2.012, 1.339–3.023), and cancer antigen 125 (CA-125; 1.022, 1.006–1.039); these remained risk factors for AECOPD prognosis in multivariate analyses. RDW and CA-125 in combination was significant in ROC curve analysis. The AUC of RDW, CA-125, and these combined were 0.691, 0.779, and 0.772, respectively. Patients with RDW >12.75% and CA-125 >15.65 U/mL were predicted to have poor prognosis. Conclusions We found that RDW and CA-125 are potential prognostic indicators for AECOPD.

Down syndrome (DS) is a common genetic condition in which a person is born with an extra copy of chromosome 21. Intellectual disability is the most common characteristic of DS. N6-methyladenosine (m6A) is a common RNA modification that is implicated in many biological processes. It is highly enriched within the brain and plays an essential role in human brain development. However, the mRNA m6A modification in the fetal brain of DS has not been explored. Here, we report m6A mRNA profiles and mRNA expression profiles of fetal brain cortex tissue from DSs and controls. We observed that the m6A modification in DS brain tissues was reduced genome-wide, which may be due to decreased the m6A methyltransferase like 3 (METTL3) protein expression. The nuclear receptor-interacting protein 1 (NRIP1/RIP140) is coded by a highly conserved chromosome 21 (Hsa21) gene. Overexpression of NRIP1 is associated with mitochondrial dysfunction in DS. The NRIP1 mRNA increased in fetal brain tissues of DS, whereas the m6A modification of the NRIP1 mRNA significantly decreased. METTL3 knockdown reduced the m6A modification of NRIP1 mRNA and increased its expression, and an increase in NRIP1 m6A modification and a decrease in its expression were observed in METTL3-overexpressed cells. The Luciferase reporter assay confirmed that METTL3 regulates NRIP1 expression in an m6A-dependent manner. The decay rate of NRIP1 mRNA was significantly reduced in METTL3-knockdown cells but increased in METTL3-overexpressed cells. We proposed that the m6A modification of NRIP1 mRNA in DS fetal brain tissue is reduced, reducing its transcript degradation rate, resulting in abnormally increased expression of NRIP1, at least partially, in the DS brain. It provides a new mechanism for the molecular pathology of DS and leads to a new insight that may become therapeutically relevant.

Background: Variations in TOR1A were thought to be associated with early-onset isolated dystonia. The variant S287Y (NM_000113.2: c.860C > A, p. Ser287Tyr, rs766483672) was found in our late-onset isolated dystonia patient. This missense variant is adjacent to R288Q (c.863G > A, p. Arg288Gln), which was reported to be associated with isolated dystonia. The potentially pathogenic role of S287Y is not conclusively known. Methods: Cytological and molecular biological analyses were performed in vitro to determine whether this variant damages the structure and function of the cell. Results: Compared with the SH-SY5Y cells overexpressing wild-type TOR1A, the cells overexpressing the protein with S287Y have an enlarged peri-nuclear space. The same changes in nuclear morphology were also found in the cells overexpressing the pathogenic variants ΔE (NM_000113.2:c.904_906delGAG, p. Glu302del), F205I (NM_000113.2:c.613 T > A, p. Phe205Ile), and R288Q (NM_000113.2:c.863G > A, p. Arg288Gln). Mutated proteins with S287Y presented a higher tendency to form dimers under reducing conditions. The same tendencies were observed in other mutated proteins but not in wild-type torsinA. Conclusions: TorsinA with S287Y damages the structure of the cell nucleus and may be a novel pathogenic mutation that causes isolated dystonia.

Effective object tracking is a great challenging task in computer vision due to the appearance changes of an object. Extreme learning machine (ELM) can analytically determine the weights of feed-forward neural networks, while compressive features can be extracted via a random sparse matrix. Thus, their common advantages are running fast. Moreover, the strong learning ability of ELM is helpful to make a tracker robust target appearance change. In this paper, we propose an effective tracking method using ELM and compressive features. First, a bank of compressive Haar-like features is extracted with the sparse random matrix. To make the feature invariant to illumination, the mean illumination of each image patch is removed from each Haar-like feature. Second, a bank of ELM-based weak classifiers are combined together to form a strong classifier to complete tracking task. Considering the discrimination performance of features is uneven, the training accuracy of each ELM is employed as the confidence (weight) of the weak classifier. Thus, less discriminative features will be weakened. The proposed tracker has been tested on several representative sequences, and the results show that it achieves the highest tracking success rate and the tracking accuracy among the compared trackers including the FCT, CNT, MIL, KCF, and ESPT.

The effect of polyvinylpyrrolidone (PVP) as glycyrrhetic acid (GA) solid dispersions carrier at different molecular weights on the dissolution behavior and physicochemical properties was investigated. PVP-GA-SDs prepared with all four molecular weight PVPs displayed good enhancement of dissolution rate and equilibrium solubility compared with pure drug and corresponding physical mixtures. The results showed that the enhancement effect of molecular weight on dissolution rate and equilibrium solubility follows PVP K30>PVP K60>PVP K17>PVP K15. In addition, the dissolution rate and solubility of the SDs with a carrier-drug ratio of 8 : 1 were better than the samples of 4 : 1. The DSC and XRD patterns showed that the crystallization of GA in SDs prepared by PVP K30 and PVP K60 was significantly inhibited, and both were transformed to amorphous. Based on FTIR and Raman detection, a hydrogen-bond between PVP and drug molecules is formed. SEM results showed that there were no significant differences in the appearance of SDs prepared with four PVPs, and no crystalline morphology of GA was seen. In conclusion, the findings of this study demonstrated that the dissolution performance of the PVP-GA-SDs prepared by the solvent method is related to the molecular weight of PVP, and the change in the molecular weight of PVP does not cause a monotonic change in dissolution of GA. The samples with PVP K30 as the carrier have the best dissolution performance.

Objective: The aim of this report is to describe the phenotype-genotype correlation of chromosome 9p deletion syndrome cases, particularly the prenatal cases. Case report: We present a female fetus at 17+4 weeks of gestation age detected with an 8 × 6 mm2 omphalocele. Karyotype analysis of the fetus cell showed a karyotype of 46,XX,del(9)(p22). Whole genome array comparative genomic hybridization analysis detected a deletion as arr 9p24.2p22.2 (226,7812–1,7466,907 bp) × 1. The prenatal abnormality and cytogenetic findings in our and other similar patients are compared. Moreover, we describe two male cases demonstrating de novo and pure 9p deletion at 9p24.3p22.1 (1–18,743,296 bp) and 9p24.3p23 (271,257–12,048,612 bp). These two children shared development delay, learning difficulties, hypotonia, and trigonocephaly. However, both the boys apparently did not present external genitalia malformation, although the deletion region included the sex reversal region 9p24.3. Conclusion: Phenotypes of 9p deletion cases are broadly in line. The prenatal diagnosis of the omphalocele provides evidence for a correlation with distal 9q deletion.

In this paper, an adaptive dual-regularization super-resolution reconstruction algorithm based on sub-pixel convolution (MPSR) is proposed. There are two novel features of the algorithm: First, the traditional regularization algorithm and sub-pixel convolution algorithm are combined to enrich the details; then, a regularization function with two adaptive parameters and two regularization terms is proposed to enhance the edge. MPSR firstly enhances the multi-scale detail of low-resolution images; then, regular processing and feature extraction are carried out; finally, sub-pixel convolution is used to fuse the extracted features to generate high-resolution images. The experimental results show that the subjective and objective evaluation indexes (PSNR/SSIM) of the algorithm have achieved satisfactory results.

Ultra-high performance liquid chromatography/ quadrupole time of flight mass spectrometry-based metabonomics platform was employed to profile the plasma metabolites of patients with metabolic syndrome and the healthy controls. Data analysis revealed lots of differential metabolites between the two groups, and most of them were identified as lipids. Several fatty acids and lysophosphatidylcholines were of higher plasma levels in the patient group, indicating the occurrence of insulin resistance and inflammation. The identified ether phospholipids were decreased in the patient group, reflecting the oxidative stress and some metabolic disorders. These identified metabolites can also be used to aid diagnosis of patients with metabolic syndrome. These results showed that metabonomics was a promising and powerful method to study metabolic syndrome.

Background: As women with low ovarian reserve embark on the challenging journey of in-vitro fertilisation (IVF) treatment, the choice between natural and mildly stimulated cycles becomes a pivotal consideration. It is unclear which of these two regimens is superior for women with low ovarian reserve. Our study aims to assess the impact of natural cycles on embryo quality and pregnancy outcomes in women with low ovarian reserve undergoing IVF treatment compared to mildly stimulated cycles. Methods: This retrospective study enrolled consecutive patients with low ovarian reserve who underwent IVF/intracytoplasmic sperm injection (ICSI) at Guangdong Second Provincial General Hospital between January 2017 and April 2021. The primary outcome for pregnancy rate of 478 natural cycles and 448 mild stimulated cycles was compared. Secondary outcomes included embryo quality and oocyte retrieval time of natural cycles. Results: The pregnancy rate in the natural cycle group was significantly higher than that in the mildly stimulated cycle group (51.8% vs. 40.1%, p = 0.046). Moreover, natural cycles exhibited higher rates of available embryos (84.1% vs. 78.6%, p = 0.040), high-quality embryos (61.8% vs. 53.2%, p = 0.008), and utilisation of oocytes (73% vs. 65%, p = 0.001) compared to mildly stimulated cycles. Oocyte retrievals in natural cycles were predominantly performed between 7:00 and 19:00, with 94.9% occurring during this time frame. In natural cycles with high-quality embryos, 96.4% of oocyte retrievals were also conducted between 7:00 and 19:00. Conclusion: Natural cycles with appropriately timed oocyte retrieval may present a valuable option for patients with low ovarian reserve. PLAIN LANGUAGE SUMMARY: In the realm of in-vitro fertilisation (IVF) treatment, women with low ovarian reserve often face the crucial decision of opting for natural or mildly stimulated cycles. This retrospective study, conducted between January 2017 and April 2021 at Guangdong Second Provincial General Hospital, delves into the impact of these cycles on pregnancy outcomes. Examining 478 natural cycles and 448 mildly stimulated cycles, the study reveals a notably higher pregnancy rate in the natural cycle group (51.8% vs. 40.1%). Additionally, natural cycles demonstrated higher rates of available embryos, high-quality embryos, and oocyte utilisation compared to their mildly stimulated counterparts. The findings suggest that natural cycles, with proper oocyte retrieval timing, could be a favourable choice for those with low ovarian reserve seeking IVF treatment. [ABSTRACT FROM AUTHOR]