Nuclear mTOR Signaling Orchestrates Transcriptional Programs Underlying Cellular Growth and Metabolism.

mTOR is a central regulator of cell growth and metabolism in response to mitogenic and nutrient signals. Notably, mTOR is not only found in the cytoplasm but also in the nucleus. This review highlights direct involvement of nuclear mTOR in regulating transcription factors, orchestrating epigenetic modifications, and facilitating chromatin remodeling. These effects intricately modulate gene expression programs associated with growth and metabolic processes. Furthermore, the review underscores the importance of nuclear mTOR in mediating the interplay between metabolism and epigenetic modifications. By integrating its functions in nutrient signaling and gene expression related to growth and metabolism, nuclear mTOR emerges as a central hub governing cellular homeostasis, malignant transformation, and cancer progression. Better understanding of nuclear mTOR signaling has the potential to lead to novel therapies against cancer and other growth-related diseases.

[Changes of hemodynamic parameters in normal pregnant women and their relationship with maternal basic characteristics].

Objective: To investigate the variation of reference ranges of hemodynamic parameters in normal pregnancy and their relation to maternal basic characteristics. Methods: A total of 598 healthy pregnant women who underwent regular prenatal examination at the Third Affiliated Hospital of Guangzhou Medical University from January to December 2023 were prospectively enrolled, and noninvasive hemodynamic monitors were used to detect changes in hemodynamic parameters of the pregnant women with the week of gestation, including cardiac output (CO), stroke volume (SV), thoracic fluid content (TFC), systemic vascular resistance (SVR), mean arterial pressure (MAP), and heart rate (HR). Relationships between hemodynamic parameters and maternal basic characteristics, including age, height, and weight, were analyzed using restricted cubic spline. Results: (1) CO (r=0.155, P<0.001), TFC (r=0.338, P<0.001), MAP (r=0.204, P<0.001), and HR (r=0.352, P<0.001) were positively correlated with the week of gestation, and SV was negatively correlated with the week of gestation (r=-0.158, P<0.001). There was no significant correlation between SVR and gestational age (r=-0.051, P=0.258). (2) CO exhibited a positive correlation with maternal height and weight (all P<0.001). The taller and heavier of pregnant women, the higher their CO. A linear relationship was observed between maternal weight and SV, MAP and HR (all P<0.01). As maternal weight increased, SV, MAP and HR showed an upward trend. Furthermore, there was an inverse association between maternal age and SVR (P<0.001). (3) There was a significant nonlinear association observed between TFC and body mass index during pregnancy (P<0.05). Additionally, a nonlinear relationship was found between SVR and MAP in relation to maternal age (all P<0.05). Notably, when the age exceeded 31 years old, there was an evident upward trend observed in both SVR and MAP. Conclusions: The hemodynamic parameters of normal pregnant women are influenced by their height, body weight, and age. It is advisable to maintain a reasonable weight during pregnancy and give birth at an appropriate age.

Maf1 controls retinal neuron number by both RNA Pol III- and Pol II-dependent mechanisms.

The generation of appropriate numbers and types of neurons is a prerequisite for assembling functional neural circuits. However, the molecular basis regulating retinal neuron number remains poorly understood. Here, we report that inactivation of the RNA polymerase (Pol) III inhibitor gene Maf1 in mice results in decreased retinal thickness and neuron number that cause attenuated electroretinogram (ERG) responses. Its absence causes aberrant differentiation of all retinal neuron types primarily by an RNA Pol II-dependent mechanism while promoting retinal progenitor cell proliferation via both Pol III- and Pol II-dependent mechanisms. Chromatin profiling and transcription assay reveal that Maf1 binds widely to the genome to regulate the expression of a large set of Pol II-transcribed genes involved in retinal cell proliferation, differentiation, and/or survival. Together, our data suggest that Maf1 may control retinal neuron number by a balanced regulation of cell proliferation, differentiation, and death via both Pol III-dependent and Pol II-dependent mechanisms.

[Mid-term outcome of deep layer repair with the long head of the biceps autograft bridging for Kim classification type â A delaminiated rotator cuff tear].

Objective: To investigate the mid-term clinical outcome of deep layer repair with the long head of the biceps autograft bridging for Kim classification type ⅠA delaminated rotator cuff tear. Methods: A follow-up study. The clinical data of 42 consecutive patients with Kim classification type ⅠA delaminated rotator cuff tear admitted to the First Affiliated Hospital of Jinan University from January 2018 to June 2019 were retrospectively included. All patients underwent shoulder arthroscopic surgery. During the operation, the autogenous long head of the biceps tendon was transferred to repair the deep layer of delaminated rotator cuff tear. The preoperative and postoperative (last follow-up) visual analogue scale (VAS) for pain, University of California Los Angeles (UCLA) score, Constant-Murley shoulder score, range of motion (ROM) of the shoulder and radiographic results were statistically analyzed. Results: A total of 42 patients were included in this study. There were 18 males and 24 females, with an average age of (64.5±15.2) years and a mean follow-up of (43.9±7.1) months. At the last follow-up, ROM of abduction increased from 80.8°±26.5° to 154.2°±14.3°, and ROM of external rotation increased from 18.2°±13.6° to 31.8°±7.8°; the VAS score of pain decreased from (5.5±1.3) points to (0.7±0.7) points, the UCLA score increased from (21.3±3.7) points to (29.9±2.1) points, and the Constant-Murley score increased from (45.4±10.0) points to (87.2±4.8) points; the differences were all statistically significant (all P<0.001). The X-ray films showed that there were no upward of the humeral head in all the patients. MRI results indicated that rotator cuff re-teared in one case (Sugaya classification type Ⅲ), and healed in other cases (Sugaya classification type Ⅰ-Ⅱ). No complications such as upper limb nerve injury was found in all cases. Conclusion: Deep layer repair with the long head of the biceps autograft bridging can significantly alleviate the pain and improve the function of patients with Kim classification type ⅠA delaminated rotator cuff tear, and the incidence of retear is low.

Maf1 is an intrinsic suppressor against spontaneous neural repair and functional recovery after ischemic stroke.

INTRODUCTION: Spontaneous recovery after CNS injury is often very limited and incomplete, leaving most stroke patients with permanent disability. Maf1 is known as a key growth suppressor in proliferating cells. However, its role in neuronal cells after stroke remains unclear. OBJECTIVE: We aimed to investigate the mechanistic role of Maf1 in spontaneous neural repair and evaluated the therapeutic effect of targeting Maf1 on stroke recovery. METHODS: We used mouse primary neurons to determine the signaling mechanism of Maf1, and the cleavage-under-targets-and-tagmentation-sequencing to map the whole-genome promoter binding sites of Maf1 in isolated mature cortical neurons. Photothrombotic stroke model was used to determine the therapeutic effect on neural repair and functional recovery by AAV-mediated Maf1 knockdown. RESULTS: We found that Maf1 mediates mTOR signaling to regulate RNA polymerase III (Pol III)-dependent rRNA and tRNA transcription in mouse cortical neurons. mTOR regulates neuronal Maf1 phosphorylation and subcellular localization. Maf1 knockdown significantly increases Pol III transcription, neurite outgrowth and dendritic spine formation in neurons. Conversely, Maf1 overexpression suppresses such activities. In response to photothrombotic stroke in mice, Maf1 expression is increased and accumulates in the nucleus of neurons in the peripheral region of infarcted cortex, which is the key region for neural remodeling and repair during spontaneous recovery. Intriguingly, Maf1 knockdown in the peri-infarct cortex significantly enhances neural plasticity and functional recovery. Mechanistically, Maf1 not only interacts with the promoters and represses Pol III-transcribed genes, but also those of CREB-associated genes that are critical for promoting plasticity during neurodevelopment and neural repair. CONCLUSION: These findings indicate Maf1 as an intrinsic neural repair suppressor against regenerative capability of mature CNS neurons, and suggest that Maf1 is a potential therapeutic target for enhancing functional recovery after ischemic stroke and other CNS injuries.

mTOR regulates aerobic glycolysis through NEAT1 and nuclear paraspeckle-mediated mechanism in hepatocellular carcinoma.

Background: Hepatocellular Carcinoma (HCC) is a major form of liver cancer and a leading cause of cancer-related death worldwide. New insights into HCC pathobiology and mechanism of drug actions are urgently needed to improve patient outcomes. HCC undergoes metabolic reprogramming of glucose metabolism from respiration to aerobic glycolysis, a phenomenon known as the 'Warburg Effect' that supports rapid cancer cell growth, survival, and invasion. mTOR is known to promote Warburg Effect, but the underlying mechanism(s) remains poorly defined. The aim of this study is to understand the mechanism(s) and significance of mTOR regulation of aerobic glycolysis in HCC. Methods: We profiled mTORC1-dependent long non-coding RNAs (lncRNAs) by RNA-seq of HCC cells treated with rapamycin. Chromatin immunoprecipitation (ChIP) and luciferase reporter assays were used to explore the transcriptional regulation of NEAT1 by mTORC1. [U-13C]-glucose labeling and metabolomic analysis, extracellular acidification Rate (ECAR) by Seahorse XF Analyzer, and glucose uptake assay were used to investigate the role of mTOR-NEAT1-NONO signaling in the regulation of aerobic glycolysis. RNA immunoprecipitation (RIP) and NONO-binding motif scanning were performed to identify the regulatory mechanism of pre-mRNA splicing by mTOR-NEAT1. Myristoylated AKT1 (mAKT1)/NRASV12-driven HCC model developed by hydrodynamic transfection (HDT) was employed to explore the significance of mTOR-NEAT1 signaling in HCC tumorigenesis and mTOR-targeted therapy. Results: mTOR regulates lncRNA transcriptome in HCC and that NEAT1 is a major mTOR transcriptional target. Interestingly, although both NEAT1_1 and NEAT1_2 are down-regulated in HCC, only NEAT1_2 is significantly correlated with poor overall survival of HCC patients. NEAT1_2 is the organizer of nuclear paraspeckles that sequester the RNA-binding proteins NONO and SFPQ. We show that upon oncogenic activation, mTORC1 suppresses NEAT1_2 expression and paraspeckle biogenesis, liberating NONO/SFPQ, which in turn, binds to U5 within the spliceosome, stimulating mRNA splicing and expression of key glycolytic enzymes. This series of actions lead to enhanced glucose transport, aerobic glycolytic flux, lactate production, and HCC growth both in vitro and in vivo. Furthermore, the paraspeckle-mediated mechanism is important for the anticancer action of US FDA-approved drugs rapamycin/temsirolimus. Conclusions: These findings reveal a molecular mechanism by which mTOR promotes the 'Warburg Effect', which is important for the metabolism and development of HCC, and anticancer response of mTOR-targeted therapy.

Amino acids control blood glucose levels through mTOR signaling.

Amino Acids are not only major nutrient sources, but also serve as chemical signals to control cellular growth. Rab1A recently emerged as a key component in amino acid sensing and signaling to activate the mTOR complex1 (mTORC1). In a recently published study [1], we generated tamoxifen-inducible, conditional whole-body Rab1A knockout in adult mice. These mice are viable but develop hyperglycemia and glucose intolerance. Interestingly, Rab1A ablation selectively reduces insulin expression and pancreatic beta-cell population. Mechanistically, branched chain amino acids (BCAA), through the Rab1A-mTORC1 complex, promote the stability and nuclear localization of Pdx1, a master transcription factor that controls growth, function and identity of pancreatic beta-cells. These findings reveal a role and underlying mechanism by which amino acids control body's glucose level through a beta-cell specific function by the Rab1A-mTORC1-Pdx1 signaling axis, which has implications in both diabetes and cancer.

Nuclear SOD1 in Growth Control, Oxidative Stress Response, Amyotrophic Lateral Sclerosis, and Cancer.

SOD1 is the major superoxide dismutase responsible for catalyzing dismutation of superoxide to hydrogen peroxide and molecular oxygen. It is well known as an essential antioxidant enzyme for maintaining cellular redox homeostasis. SOD1 dysregulation has been associated with many diseases, including amyotrophic lateral sclerosis (ALS), cancer, accelerated aging, and age-related diseases. Recent studies also revealed that SOD1 can serve as a regulatory protein in cell signaling, transcription, and ribosome biogenesis. Notably, SOD1 is localized in the nucleus under both normal and pathological conditions, contributing to oxidative stress response and growth control. Moreover, increasing evidence points to the importance of nuclear SOD1 in the pathogenesis of ALS and cancer.

Phosphorylation of androgen receptor by mTORC1 promotes liver steatosis and tumorigenesis.

BACKGROUND AND AIMS: Androgen receptor (AR) has been reported to play an important role in the development and progression of man's prostate cancer. Hepatocellular carcinoma (HCC) is also male-dominant, but the role of AR in HCC remains poorly understood. Mechanistic target of rapamycin complex 1 (mTORC1) also has been reported to be highly activated in HCC. In this study, we aimed to explore the role of AR phosphorylation and its relationship with mTORC1 in hepatocarcinogenesis. APPROACH AND RESULTS: In vitro experiment, we observed that mTORC1 interacts with hepatic AR and phosphorylates it at S96 in response to nutrient and mitogenic stimuli in HCC cells. S96 phosphorylation promotes the stability, nuclear localization, and transcriptional activity of AR, which enhances de novo lipogenesis and proliferation in hepatocytes and induces liver steatosis and hepatocarcinogenesis in mice independently and cooperatively with androgen. Furthermore, high ARS96 phosphorylation is observed in human liver steatotic and HCC tissues and is associated with overall survival and disease-free survival, which has been proven as an independent survival predictor for patients with HCC. CONCLUSIONS: AR S96 phosphorylation by mTORC1 drives liver steatosis and HCC development and progression independently and cooperatively with androgen, which not only explains why HCC is man-biased but also provides a target molecule for prevention and treatment of HCC and a potential survival predictor in patients with HCC.

mTORC1 Promotes ARID1A Degradation and Oncogenic Chromatin Remodeling in Hepatocellular Carcinoma.

The SWI/SNF chromatin remodeling complexes control accessibility of chromatin to transcriptional and coregulatory machineries. Chromatin remodeling plays important roles in normal physiology and diseases, particularly cancer. The ARID1A-containing SWI/SNF complex is commonly mutated and thought to be a key tumor suppressor in hepatocellular carcinoma (HCC), but its regulation in response to oncogenic signals remains poorly understood. mTOR is a conserved central controller of cell growth and an oncogenic driver of HCC. Remarkably, cancer mutations in mTOR and SWI/SNF complex are mutually exclusive in human HCC tumors, suggesting that they share a common oncogenic function. Here, we report that mTOR complex 1 (mTORC1) interact with ARID1A and regulates ubiquitination and proteasomal degradation of ARID1A protein. The mTORC1-ARID1A axis promoted oncogenic chromatin remodeling and YAP-dependent transcription, thereby enhancing liver cancer cell growth in vitro and tumor development in vivo. Conversely, excessive ARID1A expression counteracted AKT-driven liver tumorigenesis in vivo. Moreover, dysregulation of this axis conferred resistance to mTOR-targeted therapies. These findings demonstrate that the ARID1A-SWI/SNF complex is a regulatory target for oncogenic mTOR signaling, which is important for mTORC1-driven hepatocarcinogenesis, with implications for therapeutic interventions in HCC. SIGNIFICANCE: mTOR promotes oncogenic chromatin remodeling by controlling ARID1A degradation, which is important for liver tumorigenesis and response to mTOR- and YAP-targeted therapies in hepatocellular carcinoma.See related commentary by Pease and Fernandez-Zapico, p. 5608.

SOD1 regulates ribosome biogenesis in KRAS mutant non-small cell lung cancer.

SOD1 is known as the major cytoplasmic superoxide dismutase and an anticancer target. However, the role of SOD1 in cancer is not fully understood. Herein we describe the generation of an inducible Sod1 knockout in KRAS-driven NSCLC mouse model. Sod1 knockout markedly reduces tumor burden in vivo and blocks growth of KRAS mutant NSCLC cells in vitro. Intriguingly, SOD1 is enriched in the nucleus and notably in the nucleolus of NSCLC cells. The nuclear and nucleolar, not cytoplasmic, form of SOD1 is essential for lung cancer cell proliferation. Moreover, SOD1 interacts with PeBoW complex and controls its assembly necessary for pre-60S ribosomal subunit maturation. Mechanistically, SOD1 regulates co-localization of PeBoW with and processing of pre-rRNA, and maturation of cytoplasmic 60S ribosomal subunits in KRAS mutant lung cancer cells. Collectively, our study unravels a nuclear SOD1 function essential for ribosome biogenesis and proliferation in KRAS-driven lung cancer.

Amino acids-Rab1A-mTORC1 signaling controls whole-body glucose homeostasis.

Rab1A is a small GTPase known for its role in vesicular trafficking. Recent evidence indicates that Rab1A is essential for amino acids (aas) sensing and signaling to regulate mTORC1 in normal and cancer cells. However, Rab1A's in vivo function in mammals is not known. Here, we report the generation of tamoxifen (TAM)-induced whole body Rab1A knockout (Rab1A-/-) in adult mice. Rab1A-/- mice are viable but become hyperglycemic and glucose intolerant due to impaired insulin transcription and ß-cell proliferation and maintenance. Mechanistically, Rab1A mediates AA-mTORC1 signaling, particularly branched chain amino acids (BCAA), to regulate the stability and localization of the insulin transcription factor Pdx1. Collectively, these results reveal a physiological role of aa-Rab1A-mTORC1 signaling in the control of whole-body glucose homeostasis in mammals. Intriguingly, Rab1A expression is reduced in ß-cells of type 2 diabetes (T2D) patients, which is correlated with loss of insulin expression, suggesting that Rab1A downregulation contributes to T2D progression.

Toward improving androgen receptor-targeted therapies in male-dominant hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) is the predominant form of liver cancer and a leading cause of cancer deaths worldwide. HCC is a male-dominant cancer with a male:female ratio of up to 7:1. The androgen receptor (AR) is the male hormone receptor known as a major oncogenic driver of prostate cancer. Although AR has been linked to the sexual dimorphism of HCC, clinical trials with AR-targeted agents failed to generate survival benefits. Recent studies provide new insights into the role of AR in liver tumorigenesis and therapeutic responses. Herein, we review current understanding of AR signaling in HCC and feedback mechanisms that limit response to AR blockade. New AR-targeting strategies that might improve outcomes in HCC therapies are also discussed.

[The diagnostic value of chromosome microarray analysis technique in the genetic causes of children with intellectual disability or global developmental delay].

Objective: To study the value of chromosome microarray analysis (CMA) application in children with developmental delay (DD), intellectual disability (ID), autistic spectrum disorder (ASD) and multiple congenital anomalies (MCA). Methods: Genomic DNA was extracted from peripheral blood samples. Array-based comparative genomic hybridization (array-CGH) analysis and single nucleotide polymorphism array (SNP-array) were performed in 1 320 children with DD/ID, ASD, with or without epilepsy and MCA who were admitted to Peking University First Hospital from 2014 to 2019. The results of genetic etiology test of CMA in children with mental retardation or global DD was summarized. Results: Of 1 320 samples, there were 10 cases of aneuploid abnormality, 6 cases of uniparental disomy and one case of mosaicism, respectively. Pathogenic copy number variations (CNVs) were found in 320 cases and pathogenic CNVs were detected in 23 cases, with a combined detection rate of 26% (343/1 320). CNVs of uncertain clinical significance occurred in 107 cases, accounting for 8.1% (107/1 320). There were 25 cases of possible benign CNVs, accounting for 2% (25/1 320), while benign CNVs were reported in 20 cases, accounting for 1.5% (20/1 320). The detection rate of MCA with DD/ID was 39.8% (130/327). Conclusions: CMA has the advantages of high resolution and covering the whole genome. It can detect the chromosomal abnormalities, microdeletions and duplications seen under the microscope, thus the genetic etiology of children with mental retardation or global DD can be diagnosed.

[Effects of percutaneous coronary intervention on cardiac function in ischemic cardiomyopathy patients with different left ventricular ejection fraction and SYNTAX score≤22].

Objective: To evaluate the clinical impact of percutaneous coronary intervention (PCI) on left ventricular myocardial remodeling and main adverse cardiovascular and cerebrovascular events (MACE) in ischemic cardiomyopathy patients with different left ventricular ejection fraction and SYNTAX score≤22. Methods: A total of 191 ischemic cardiomyopathy patients who underwent PCI in Department of Cardiology from May 2017 to October 2018 were enrolled in this study, and they were divided into three groups according to preoperative left ventricular ejection fraction (≥50% group, 36%~49% group and ≤35% group). The main outcomes and left ventricular ejection fraction, left ventricular end-diastolic volume were analyzed at 12 months follow-up. The main outcomes were the recurrence of acute left ventricular failure, recurrent angina, restenosis, revascularization, non-fatal myocardial infarction, cardiovascular death and non-cardiovascular death. Results: The incidence of MACE was 32.6% (15 cases) in ≥50% group, 32.0% (31 cases) in 36%-49% group, 45.8% (22 cases) in ≤35% group, respectively, which was lower in the first two groups than in ≤35% group, but there was no statistically significant difference among the 3 groups (P=0.231). The incidence of acute left ventricular failure in the three groups was 2.2%, 12.4% and 22.9%, respectively, and there was statistically significant difference among the 3 groups (P= 0.01). Multivariate analysis indicated that preoperative left ventricular ejection fraction ≤35% was an independent predictor of acute left ventricular failure (OR=2.696, 95%CI: 1.099-6.612, P=0.030). Compared with baseline data, left ventricular end-diastolic volume ((62±4) mm vs (56±5) mm, P<0.001), left atrium ((42±6) mm vs (40±6) mm, P<0.001) decreased significantly 1 year after PCI. However, left ventricular ejection fraction ((43±10)% vs (51±13)%, P<0.001) increased significantly. At 1 year, left ventricular remodeling related parameters were detected in 3 groups, and there was statistically significant difference in left ventricular end-diastolic volume ((53.1±0.6) mm vs (55.1±0.5) mm vs (59.1±0.7) mm, P<0.001), left ventricular ejection fraction ((62.1±1.1)% vs (51.4±1.0)% vs (37.0±1.5)%, P<0.001) among the 3 groups. Conclusions: Coronary vascular reopening with PCI in patients with ischaemic cardiomyopathy and SYNTAX score≤22, can improve prognosis of patients with preoperative left ventricular ejection fraction>35% significantly, but not in those with preoperative left ventricular ejection fraction≤35%. Preoperative left ventricular ejection fraction may be an independent predictor of acute left ventricular failure in patients with ischemic cardiomyopathy and SYNTAX score≤22, postoperative left ventricular remodeling and left ventricular systolic function correlate with preoperative left ventricular ejection fraction.

TGFß1I1 suppressed cell migration and invasion in colorectal cancer by inhibiting the TGF-ß pathway and EMT progress.

OBJECTIVE: Colorectal cancer (CRC) is the fourth leading cause of death worldwide and there is a need for more specific therapeutic targets and biomarkers for the disease. Transforming growth factor ß1-induced transcript 1 (TGFΒ1I1) was reported to be downregulated in CRC tissues; however, the precise roles of TGFΒ1I1 in CRC remain unclear. PATIENTS AND METHODS: The expression of TGFΒ1I1 in CRC cell lines and tissues was assessed by quantitative Polymerase Chain Reaction (qPCR). TGFΒ1I1 was overexpressed in SW620 and RKO cells. Cell viability was analyzed by a CCK-8 assay. The proportion of apoptotic cells was analyzed by flow cytometry. The EdU cell proliferation assay of SW620 and RKO cells after transfection was performed via flow cytometry. The migration potency of SW620 and RKO cells was analyzed using a cell migration assay. A wound healing assay was performed to assess the migration potency of SW620 and RKO cells. The invasion potency of SW620 and RKO cells after TGFΒ1I1 overexpression was analyzed. The protein levels of VEGF, TGF-ß, MMP9, p-Smad2/3, N-cadherin, and E-cadherin were analyzed by Western blot. RESULTS: Decreased expression of TGFΒ1I1 was found in CRC tissues and cell lines. Overexpression of TGFΒ1I1 inhibited the proliferation and induced the apoptosis of CRC cells. The overexpression of TGFΒ1I1 inhibited the migration and invasion of CRC cells. We also found that the overexpression of TGFΒ1I1 in CRC cells inhibited the TGF-ß pathway and epithelial-mesenchymal transition (EMT) progress. CONCLUSIONS: TGFΒ1I1 suppressed cell migration and invasion in CRC by inhibiting the TGF-ß pathway and EMT progress.

[Establishment and clinical testing of pancreatic cancer Faster R-CNN AI system based on fast regional convolutional neural network].

Objective: To investigate the effectiveness of an enhanced CT automatic recognition system based on Faster R-CNN for pancreatic cancer and its clinical value. Methods: In this study, 4 024 enhanced CT imaging sequences of 315 patients with pancreatic cancer from January 2013 to May 2016 at the Affiliated Hospital of Qingdao University were collected retrospectively, and 2 614 imaging sequences were input into the faster R-CNN system as training dataset to create an automatic image recognition model, which was then validated by reading 1 410 enhanced CT images of 135 cases of pancreatic cancer.In order to identify its effectiveness, 3 750 CT images of 150 patients with pancreatic lesions were read and a followed-up was carried out.The accuracy and recall rate in detecting nodules were recorded and regression curves were generated.In addition, the accuracy, sensitivity and specificity of Faster R-CNN diagnosis were analyzed, the ROC curves were generated and the area under the curves were calculated. Results: Based on the enhanced CT images of 135 cases, the area under the ROC curve was 0.927 calculated by Faster R-CNN. The accuracy, specificity and sensitivity were 0.902, 0.913 and 0.801 respectively.After the data of 150 patients with pancreatic cancer were verified, 893 CT images showed positive and 2 857 negative.Ninety-eight patients with pancreatic cancer were diagnosed by Faster R-CNN.After the follow-up, it was found that 53 cases were post-operatively proved to be pancreatic ductal carcinoma, 21 cases of pancreatic cystadenocarcinoma, 12 cases of pancreatic cystadenoma, 5 cases of pancreatic cyst, and 7 cases were untreated.During 5 to 17 months after operation, 6 patients died of abdominal tumor infiltration, liver and lung metastasis.Of the 52 patients who were diagnosed negative by Faster R-CNN, 9 were post-operatively proved to be pancreatic ductal carcinoma. Conclusion: Faster R-CNN system has clinical value in helping imaging physicians to diagnose pancreatic cancer.

[Application of convolutional neural network to risk evaluation of positive circumferential resection margin of rectal cancer by magnetic resonance imaging].

Objective: To explore the feasibility of using faster regional convolutional neural network (Faster R-CNN) to evaluate the status of circumferential resection margin (CRM) of rectal cancer in the magnetic resonance imaging (MRI). Methods: This study was registered in the Chinese Clinical Trial Registry (ChiCTR-1800017410). Case inclusion criteria: (1) the positive area of CRM was located between the plane of the levator ani, anal canal and peritoneal reflection; (2) rectal malignancy was confirmed by electronic colonoscopy and histopathological examination; (3) positive CRM was confirmed by postoperative pathology or preoperative high-resolution MRI. Exclusion criteria: patients after neoadjuvant therapy, recurrent cancer after surgery, poor quality images, giant tumor with extensive necrosis and tissue degeneration, and rectal tissue construction changes in previous pelvic surgery. According to the above criteria, MRI plain scan images of 350 patients with rectal cancer and positive CRM in The Affiliated Hospital of Qingdao University from July 2016 to June 2019 were collected. The patients were classified by gender and tumor position, and randomly assigned to the training group (300 cases) and the validation group (50 cases) at a ratio of 6:1 by computer random number method. The CRM positive region was identified on the T2WI image using the LabelImg software. The identified training group images were used to iteratively train and optimize parameters of the Faster R-CNN model until the network converged to obtain the best deep learning model. The test set data were used to evaluate the recognition performance of the artificial intelligence platform. The selected indicators included accuracy, sensitivity, positive predictive value, receiver operating characteristic (ROC) curves, areas under the ROC curves (AUC), and the time taken to identify a single image. Results: The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the CRM status determined by the trained Faster R-CNN artificial intelligence approach were 0.884, 0.857, 0.898, 0.807, and 0.926, respectively; the AUC was 0.934 (95% CI: 91.3% to 95.4%). The Faster R-CNN model's automatic recognition time for a single image was 0.2 s. Conclusion: The artificial intelligence model based on Faster R-CNN for the identification and segmentation of CRM-positive MRI images of rectal cancer is established, which can complete the risk assessment of CRM-positive areas caused by in-situ tumor invasion and has the application value of preliminary screening.