Sequential CD7 CAR T-Cell Therapy and Allogeneic HSCT without GVHD Prophylaxis.

BACKGROUND: Patients with relapsed or refractory hematologic cancers have a poor prognosis. Chimeric antigen receptor (CAR) T-cell therapy as a bridge to allogeneic hematopoietic stem-cell transplantation (HSCT) has the potential for long-term tumor elimination. However, pre-HSCT myeloablation and graft-versus-host disease (GVHD) prophylaxis agents have toxic effects and could eradicate residual CAR T cells and compromise antitumor effects. Whether the integration of CAR T-cell therapy and allogeneic HSCT can preserve CAR T-cell function and improve tumor control is unclear. METHODS: We tested a novel "all-in-one" strategy consisting of sequential CD7 CAR T-cell therapy and haploidentical HSCT in 10 patients with relapsed or refractory CD7-positive leukemia or lymphoma. After CAR T-cell therapy led to complete remission with incomplete hematologic recovery, patients received haploidentical HSCT without pharmacologic myeloablation or GVHD prophylaxis drugs. Toxic effects and efficacy were closely monitored. RESULTS: After CAR T-cell therapy, all 10 patients had complete remission with incomplete hematologic recovery and grade 4 pancytopenia. After haploidentical HSCT, 1 patient died on day 13 of septic shock and encephalitis, 8 patients had full donor chimerism, and 1 patient had autologous hematopoiesis. Three patients had grade 2 HSCT-associated acute GVHD. The median follow-up was 15.1 months (range, 3.1 to 24.0) after CAR T-cell therapy. Six patients remained in minimal residual disease-negative complete remission, 2 had a relapse of CD7-negative leukemia, and 1 died of septic shock at 3.7 months. The estimated 1-year overall survival was 68% (95% confidence interval [CI], 43 to 100), and the estimated 1-year disease-free survival was 54% (95% CI, 29 to 100). CONCLUSIONS: Our findings suggest that sequential CD7 CAR T-cell therapy and haploidentical HSCT is safe and effective, with remission and serious but reversible adverse events. This strategy offers a feasible approach for patients with CD7-positive tumors who are ineligible for conventional allogeneic HSCT. (Funded by the National Natural Science Foundation of China and the Key Project of Science and Technology Department of Zhejiang Province; ClinicalTrials.gov numbers, NCT04599556 and NCT04538599.).

Roseinatronobacter alkalisoli sp. nov., an alkaliphilic bacterium isolated from soda soil, and genome-based reclassification of the genera Rhodobaca and Roseinatronobacter.

The genera Rhodobaca and Roseinatronobacter are phylogenetically related genera within the family Paracoccaceae. Species of these genera were described using 16S rRNA gene-based phylogeny and phenotypic characteristics. However, the 16S rRNA gene identity and phylogeny reveal the controversy of the taxonomic status of these two genera. In this work, we examined the taxonomic positions of members of both genera using 16S rRNA gene phylogeny, phylogenomic analysis and further validated using overall genome-related indexes, including digital DNA-DNA hybridization, average nucleotide identity, average amino acid identity and percentage of conserved proteins. Based on phylogenetic and phylogenomic results, the current four species of the two genera clustered tightly into one clade with high bootstrap values, suggesting that the genus Rhodobaca should be merged with Roseinatronobacter. In addition, a novel species isolated from a soda soil sample collected from Anda City, PR China, and designated as HJB301T was also described. Phenotypic, chemotaxonomic, genomic and phylogenetic properties suggested that strain HJB301T (=CCTCC AB 2021113T=KCTC 82977T) represents a novel species of the genus Roseinatronobacter, for which the name Roseinatronobacter alkalisoli sp. nov. is proposed.

Apoptosis inhibition enhances induced pluripotent stem cell generation during T cell reprogramming.

The widespread adoption of chimeric antigen receptor (CAR)-T cell therapy has been hindered by its complex and costly manufacturing process. Induced pluripotent stem cells (iPSCs) have shown promise as a cellular immunotherapy alternative, due to their unlimited self-renewal potential in culture and capacity to differentiate into functional immune cell types. However, it is imperative to carefully select the original cell for iPSC seed preparation, as iPSCs have been found to retain the epigenetic imprint of the original somatic cells. Additionally, the efficiency of reprogramming terminal differentiated cells for immunotherapy must be addressed. Our research highlights the superiority of lymphocyte-origin cells over embryonic stem cells in functional immune cell differentiation. Furthermore, blocking Fas-FasL induced apoptosis in T cells significantly improves iPSC generation. Interestingly, transient Fas suppression in T cells does not alter the expression of Fas in the resulting iPSCs or affect their differentiation potential. This finding brings up new avenues in the field of cellular immunotherapy and provides a solution for creating high-quality and suitable iPSCs for lymphocyte differentiation for immunotherapy purposes.

Real-time detection of laryngopharyngeal cancer using an artificial intelligence-assisted system with multimodal data.

BACKGROUND: Laryngopharyngeal cancer (LPC) includes laryngeal and hypopharyngeal cancer, whose early diagnosis can significantly improve the prognosis and quality of life of patients. Pathological biopsy of suspicious cancerous tissue under the guidance of laryngoscopy is the gold standard for diagnosing LPC. However, this subjective examination largely depends on the skills and experience of laryngologists, which increases the possibility of missed diagnoses and repeated unnecessary biopsies. We aimed to develop and validate a deep convolutional neural network-based Laryngopharyngeal Artificial Intelligence Diagnostic System (LPAIDS) for real-time automatically identifying LPC in both laryngoscopy white-light imaging (WLI) and narrow-band imaging (NBI) images to improve the diagnostic accuracy of LPC by reducing diagnostic variation among on-expert laryngologists. METHODS: All 31,543 laryngoscopic images from 2382 patients were categorised into training, verification, and test sets to develop, validate, and internal test LPAIDS. Another 25,063 images from five other hospitals were used as external tests. Overall, 551 videos were used to evaluate the real-time performance of the system, and 200 randomly selected videos were used to compare the diagnostic performance of the LPAIDS with that of laryngologists. Two deep-learning models using either WLI (model W) or NBI (model N) images were constructed to compare with LPAIDS. RESULTS: LPAIDS had a higher diagnostic performance than models W and N, with accuracies of 0·956 and 0·949 in the internal image and video tests, respectively. The robustness and stability of LPAIDS were validated in external sets with the area under the receiver operating characteristic curve values of 0·965-0·987. In the laryngologist-machine competition, LPAIDS achieved an accuracy of 0·940, which was comparable to expert laryngologists and outperformed other laryngologists with varying qualifications. CONCLUSIONS: LPAIDS provided high accuracy and stability in detecting LPC in real-time, which showed great potential for using LPAIDS to improve the diagnostic accuracy of LPC by reducing diagnostic variation among on-expert laryngologists.

Regulating Catalytic Activity of DNA-Templated Silver Nanoclusters Based on their Differential Interactions with DNA Structures and Stimuli-Responsive Structural Transition.

This work reports exquisite engineering of catalytic activity of DNA-templated silver nanoclusters (DNA-AgNCs) based on unique adsorption phenomena of DNAs on DNA-AgNCs and reversible transition between double and triple-stranded DNAs. Four DNA homopolymers exhibit different inhibition effects on the catalytic activity of DNA-AgNCs, poly adenine (polyA) > poly guanine (polyG) > poly cytosine (polyC) > poly thymine (polyT), demonstrating that polyA strands have the strongest adsorption affinity on DNA-AgNCs. Through the formation of T-A•T triplex DNAs, catalytic activity of DNA-AgNCs is restored from the deactivated state by double or single-stranded DNAs, indicating the participation of N7 groups of adenine bases in binding to DNA-AgNCs and blocking active sites. Accordingly, reversibly regulating catalytic activity of DNA-AgNCs can be realized based on DNA input-stimulated transition between duplex and triplex structures. In the end, two low-cost and facile biosensing methods are presented, which are derived from the activity-switchable platform. It is worthy to anticipate that the DNA-AgNCs with controlled catalytic activity will inspire researchers to devise more functionalized nanocatalysts and contribute to the exploration of intelligent biomedicine in the future.

Catalytical feature of optical nanoprobes of boron nitride quantum dots in the presence of Cu2+ for the determination of dopamine.

Monitoring the concentration of dopamine (DA) is vital for preventing and diagnosing DA related diseases. In contrast to the traditional sensing methods for DA, in which direct or indirect effects on the optical probes are often recorded, a novel sensing concept is disclosed based on as a result of the in situ formation of polydopamine (PDA) originating from the synergetic effect between boron nitride quantum dots (BNQDs) and Cu2+. In the co-presence of BNQDs and Cu2+, DA was catalytically oxidized to PDA, accompanied by an obvious color change from colorless to brown. In contrast to previous reports, in which BNQDs have been employed as an optical probe, herein, the BNQDs not only acted as the optical energy donor, but also as the catalysts for the formation of PDA. The quenching efficiency resulting from the inner filter effect and the electron transfer between the BNQDs and PDA was directly proportional to the concentration of DA, ranging linearly from 2 to 80 µM with a limit of detection of 0.49 µM. The present system exhibited an outstanding selectivity for DA among other interfering coexisting biomolecules. Furthermore, the practical application of the proposed platform was verified by assaying DA in human plasma samples, and satisfactory recoveries ranging from 101.24% to 111.98% were obtained. With the satisfactory reliability, repeatability and stability, the proposed simple sensor showed significant potential for use in DA detection in other biomedical applications.

Exploration of intramolecular split G-quadruplex and its analytical applications.

Distinct from intermolecular split G-quadruplex (Inter-SG), intramolecular split G-quadruplex (Intra-SG) which could be generated in a DNA spacer-inserted G-quadruplex strand has not been systematically explored. Not only is it essential for the purpose of simplicity of DNA-based bioanalytical applications, but also it will give us hints how to design split G-quadruplex-based system. Herein, comprehensive information is provided about influences of spacer length and split mode on the formation of Intra-SG, how to adjust its thermodynamic stability, and selection of optimal Intra-SG for bioanalysis. For instances, non-classical Intra-SG (e.g. 2:10, 4:8 and 5:7) displays lower stability than classical split strands (3:9, 6:6 and 9:3), which is closely related to integrity of consecutive guanine tract; as compared to regular Intra-SG structures, single-thymine capped ones have reduced melting temperature, providing an effective approach to adjustment of stability. It is believed that the disclosed rules in this study will contribute to the effective application of split G-quadruplex in the field of DNA technology in the future.

Fine-tuning ERNIE for chest abnormal imaging signs extraction.

Chest imaging reports describe the results of chest radiography procedures. Automatic extraction of abnormal imaging signs from chest imaging reports has a pivotal role in clinical research and a wide range of downstream medical tasks. However, there are few studies on information extraction from Chinese chest imaging reports. In this paper, we formulate chest abnormal imaging sign extraction as a sequence tagging and matching problem. On this basis, we propose a transferred abnormal imaging signs extractor with pretrained ERNIE as the backbone, named EASON (fine-tuning ERNIE with CRF for Abnormal Signs ExtractiON), which can address the problem of data insufficiency. In addition, to assign the attributes (the body part and degree) to corresponding abnormal imaging signs from the results of the sequence tagging model, we design a simple but effective tag2relation algorithm based on the nature of chest imaging report text. We evaluate our method on the corpus provided by a medical big data company, and the experimental results demonstrate that our method achieves significant and consistent improvement compared to other baselines.

Reciprocal regulation of Akt and Oct4 promotes the self-renewal and survival of embryonal carcinoma cells.

Signaling via the Akt serine/threonine protein kinase plays critical roles in the self-renewal of embryonic stem cells and their malignant counterpart, embryonal carcinoma cells (ECCs). Here we show that in ECCs, Akt phosphorylated the master pluripotency factor Oct4 at threonine 235, and that the levels of phosphorylated Oct4 in ECCs correlated with resistance to apoptosis and tumorigenic potential. Phosphorylation of Oct4 increased its stability and facilitated its nuclear localization and its interaction with Sox2, which promoted the transcription of the core stemness genes POU5F1 and NANOG. Furthermore, in ECCs, unphosphorylated Oct4 bound to the AKT1 promoter and repressed its transcription. Phosphorylation of Oct4 by Akt resulted in dissociation of Oct4 from the AKT1 promoter, which activated AKT1 transcription and promoted cell survival. Therefore, a site-specific, posttranslational modification of the Oct4 protein orchestrates the regulation of its stability, subcellular localization, and transcriptional activities, which collectively promotes the survival and tumorigenicity of ECCs.

Chinese clinical named entity recognition with radical-level feature and self-attention mechanism.

Named entity recognition is a fundamental and crucial task in medical natural language processing problems. In medical fields, Chinese clinical named entity recognition identifies boundaries and types of medical entities from unstructured text such as electronic medical records. Recently, a composition model of bidirectional Long Short-term Memory Networks (BiLSTMs) and conditional random field (BiLSTM-CRF) based character-level semantics has achieved great success in Chinese clinical named entity recognition tasks. But this method can only capture contextual semantics between characters in sentences. However, Chinese characters are hieroglyphics, and deeper semantic information is hidden inside, the BiLSTM-CRF model failed to get this information. In addition, some of the entities in the sentence are dependent, but the Long Short-term Memory (LSTM) does not capture long-term dependencies perfectly between characters. So we propose a BiLSTM-CRF model based on the radical-level feature and self-attention mechanism to solve these problems. We use the convolutional neural network (CNN) to extract radical-level features, aims to capture the intrinsic and internal relevances of characters. In addition, we use self-attention mechanism to capture the dependency between characters regardless of their distance. Experiments show that our model achieves F1-score 93.00% and 86.34% on CCKS-2017 and TP_CNER dataset respectively.

A narrow open tubular column for high efficiency liquid chromatographic separation.

We report a great feature of open tubular liquid chromatography when it is run using an extremely narrow (e.g., 2 µm inner diameter) open tubular column: more than 10 million plates per meter can be achieved in less than 10 min and under an elution pressure of ca. 20 bar. The column is coated with octadecylsilane and both isocratic and gradient separations are performed. We reveal a focusing effect that may be used to interpret the efficiency enhancement. We also demonstrate the feasibility of using this technique for separating complex peptide samples. This high-resolution and fast separation technique is promising and can lead to a powerful tool for trace sample analysis.

Tunable Electroosmosis-Based Femto-Liter Pipette: A Promising Tool toward Living-Cell Surgery.

Single-cell analysis has attracted increasing attention because of cell heterogeneities. Various strategies have been developed for analyzing single cells, but most of these analytical processes kill the cells. Tools that can qualitatively and quantitatively measure the cellular contents without killing the cell are highly demanding because they enable us to conduct single-cell time-course studies (e.g., to examine how a cell responds to a therapy before, during, and after a treatment). Here we develop a femto-liter (fL) pipet to serve this purpose. To ensure that we can accurately and precisely pipet fL solutions, we fill all conduits with liquid and use an electroosmotic pump (EOP) as the driving force to facilitate withdrawal of cellular contents from single cells. We tentatively term this device an EOP-driven pipette or EDP. We characterize the EDP for accurately and precisely withdrawing solution from ∼250 fL to 80 nL; a volume range that covers the applications for most types of cells. To demonstrate the feasibility of utilizing the EDP for a single-cell time-course study, we utilize the EDP to take the cellular contents out at different times during the course of a zebrafish embryo development for cholesterol measurements. More than 50% of the embryos survive after each pipetting and analysis step, and this number will increase considerably as we improve our cell manipulation skills and reduce the pipet-tip diameter. We expect this EDP to become an effective tool for single-cell time-course studies.

Programmed pH-Driven Reversible Association and Dissociation of Interconnected Circular DNA Dimer Nanostructures.

The switchable pH-driven reversible assembly and dissociation of interlocked circular DNA dimers is presented. The circular DNA dimers are interconnected by pH-responsive nucleic acid bridges. In one configuration, the two-ring nanostructure is separated at pH = 5.0 to individual rings by reconfiguring the interlocking bridges into C-G·C(+) triplex units, and the two-ring assembly is reformed at pH = 7.0. In the second configuration, the dimer of circular DNAs is bridged at pH = 7.0 by the T-A·T triplex bridging units that are separated at pH = 10.0, leading to the dissociation of the dimer to single circular DNA nanostructures. The two circular DNA units are also interconnected by two pH-responsive locks. The pH-programmed opening of the locks at pH = 5.0 or pH = 10.0 yields two isomeric dimer structures composed of two circular DNAs. The switchable reconfigured states of the circular DNA nanostructures are followed by time-dependent fluorescence changes of fluorophore/quencher labeled systems and by complementary gel electrophoresis experiments. The dimer circular DNA structures are further implemented as scaffolds for the assembly of Au nanoparticle dimers exhibiting controlled spatial separation.

Versatile G-quadruplex-mediated strategies in label-free biosensors and logic systems.

G-quadruplex (G4), as one of the significant functional nucleic acids (FNAs), has attracted researchers' wide attention, and in particular has been employed for the construction of label-free molecular sensors and logic systems based on the peroxidase-like activity of the G4-hemin complex and G4-enhanced luminescence of G4-binding organic dyes. Its cation-dependent conformation and stability provide opportunities for the recognition of metal ion inputs and application of a split G4 strategy. Moreover, coupling the G4 sequence with other FNAs, e.g. metal ion-dependent DNAzymes and aptamers, has prominently broadened the range of possible targets from metal ions and DNA to diverse proteins and cells. Although there are limitations, such as a low ability of anti-interference and multiplex analysis, the excellent advantages (e.g. simplicity and low cost) endow the G4-mediated strategy with tremendous potential to be further exploited for practical bioanalysis and complicated DNA computing.

Topological effect of an intramolecular split G-quadruplex on thioflavin T binding and fluorescence light-up.

In this work, the topological effect on binding interaction between a G-quadruplex and thioflavin T (ThT) ligand was systematically investigated on a platform of an intramolecular split G-quadruplex (Intra-SG). Distinct fluorescence changes from ThT were presented in the presence of distinct split modes of Intra-SG structures and an intriguing phenomenon of target-induced fluorescence light-up occurred for split modes 2 : 10, 5 : 7 and 8 : 4. It was validated that hybridization between the Intra-SG spacer and target did not unfold the G-quadruplex, but facilitated the ThT binding. Moreover, the 3' guanine-rich fragment of Intra-SG was very susceptible to topology variation produced by the bound target strand. Additionally, a bioanalytical method was developed for ultrasensitive gene detection, confirming the utility of the ThT/Intra-SG complex as a universal signal transducer. It is believed that the results and disclosed rules will inspire researchers to develop many new DNA-based signal transducers in the future.

Dual-signal readout sensing of ATP content in single dental pulp stem cells during differentiation via functionalized glass nanopipettes.

Adenosine triphosphate (ATP) is regarded as the "energy currency" in living cells, so real-time quantification of content variation of intracellular ATP is highly desired for understanding some important physiological processes. Due to its single-molecule readout ability, nanopipette sensing has emerged as a powerful technique for molecular sensing. In this study, based on the effect of targeting-aptamer binding on ionic current, and fluorescence resonance energy transfer (FRET), we reported a dual-signal readout nanopipette sensing system for monitoring ATP content variation at the subcellular level. In the presence of ATP, the complementary DNA-modified gold nanoparticles (cDNAs-AuNPs) were released from the inner wall of the nanopipette, which leads to sensitive response variations in ionic current rectification and fluorescence intensity. The developed nanopipette sensor was capable of detecting ATP in single cells, and the fluctuation of ATP content in the differentiation of dental pulp stem cells (DPSCs) was further quantified with this method. The study provides a more reliable nanopipette sensing platform due to the introduction of fluorescence readout signals. Significantly, the study of energy fluctuation during cell differentiation from the perspective of energy metabolism is helpful for differentiation regulation and cell therapy.

Building and validating an artificial intelligence model to identify tracheobronchopathia osteochondroplastica by using bronchoscopic images.

BACKGROUND: Given the rarity of tracheobronchopathia osteochondroplastica (TO), many young doctors in primary hospitals are unable to identify TO based on bronchoscopy findings. OBJECTIVES: To build an artificial intelligence (AI) model for differentiating TO from other multinodular airway diseases by using bronchoscopic images. DESIGN: We designed the study by comparing the imaging data of patients undergoing bronchoscopy from January 2010 to October 2022 by using EfficientNet. Bronchoscopic images of 21 patients with TO at Anhui Chest Hospital from October 2019 to October 2022 were collected for external validation. METHODS: Bronchoscopic images of patients with multinodular airway lesions (including TO, amyloidosis, tumors, and inflammation) and without airway lesions in the First Affiliated Hospital of Guangzhou Medical University were collected. The images were randomized (4:1) into training and validation groups based on different diseases and utilized for deep learning by convolutional neural networks (CNNs). RESULTS: We enrolled 201 patients with multinodular airway disease (38, 15, 75, and 73 patients with TO, amyloidosis, tumors, and inflammation, respectively) and 213 without any airway lesions. To find multinodular lesion images for deep learning, we utilized 2183 bronchoscopic images of multinodular lesions (including TO, amyloidosis, tumor, and inflammation) and compared them with images without any airway lesions (1733). The accuracy of multinodular lesion identification was 98.9%. Further, the accuracy of TO detection based on the bronchoscopic images of multinodular lesions was 89.2%. Regarding external validation (using images from 21 patients with TO), all patients could be diagnosed with TO; the accuracy was 89.8%. CONCLUSION: We built an AI model that could differentiate TO from other multinodular airway diseases (mainly amyloidosis, tumors, and inflammation) by using bronchoscopic images. The model could help young physicians identify this rare airway disease.

Enantioselective synthesis of chiral α,α-dialkyl indoles and related azoles by cobalt-catalyzed hydroalkylation and regioselectivity switch.

General, catalytic and enantioselective construction of chiral α,α-dialkyl indoles represents an important yet challenging objective to be developed. Herein we describe a cobalt catalyzed enantioselective anti-Markovnikov alkene hydroalkylation via the remote stereocontrol for the synthesis of α,α-dialkyl indoles and other N-heterocycles. This asymmetric C(sp3)-C(sp3) coupling features high flexibility in introducing a diverse set of alkyl groups at the α-position of chiral N-heterocycles. The utility of this methodology has been demonstrated by late-stage functionalization of drug molecules, asymmetric synthesis of bioactive molecules, natural products and functional materials, and identification of a class of molecules exhibiting anti-apoptosis activities in UVB-irradiated HaCaT cells. Ligands play a vital role in controlling the reaction regioselectivity. Changing the ligand from bi-dentate L6 to tridentate L12 enables CoH-catalyzed Markovnikov hydroalkylation. Mechanistic studies disclose that the anti-Markovnikov hydroalkylation involves a migratory insertion process while the Markovnikov hydroalkylation involves a MHAT process.

Predicting the severity of white matter lesions among patients with cerebrovascular risk factors based on retinal images and clinical laboratory data: a deep learning study.

Background and purpose: As one common feature of cerebral small vascular disease (cSVD), white matter lesions (WMLs) could lead to reduction in brain function. Using a convenient, cheap, and non-intrusive method to detect WMLs could substantially benefit to patient management in the community screening, especially in the settings of availability or contraindication of magnetic resonance imaging (MRI). Therefore, this study aimed to develop a useful model to incorporate clinical laboratory data and retinal images using deep learning models to predict the severity of WMLs. Methods: Two hundred fifty-nine patients with any kind of neurological diseases were enrolled in our study. Demographic data, retinal images, MRI, and laboratory data were collected for the patients. The patients were assigned to the absent/mild and moderate-severe WMLs groups according to Fazekas scoring system. Retinal images were acquired by fundus photography. A ResNet deep learning framework was used to analyze the retinal images. A clinical-laboratory signature was generated from laboratory data. Two prediction models, a combined model including demographic data, the clinical-laboratory signature, and the retinal images and a clinical model including only demographic data and the clinical-laboratory signature, were developed to predict the severity of WMLs. Results: Approximately one-quarter of the patients (25.6%) had moderate-severe WMLs. The left and right retinal images predicted moderate-severe WMLs with area under the curves (AUCs) of 0.73 and 0.94. The clinical-laboratory signature predicted moderate-severe WMLs with an AUC of 0.73. The combined model showed good performance in predicting moderate-severe WMLs with an AUC of 0.95, while the clinical model predicted moderate-severe WMLs with an AUC of 0.78. Conclusion: Combined with retinal images from conventional fundus photography and clinical laboratory data are reliable and convenient approach to predict the severity of WMLs and are helpful for the management and follow-up of WMLs patients.

Distinguishing bronchoscopically observed anatomical positions of airway under by convolutional neural network.

Background: Artificial intelligence (AI) technology has been used for finding lesions via gastrointestinal endoscopy. However, there were few AI-associated studies that discuss bronchoscopy. Objectives: To use convolutional neural network (CNN) to recognize the observed anatomical positions of the airway under bronchoscopy. Design: We designed the study by comparing the imaging data of patients undergoing bronchoscopy from March 2022 to October 2022 by using EfficientNet (one of the CNNs) and U-Net. Methods: Based on the inclusion and exclusion criteria, 1527 clear images of normal anatomical positions of the airways from 200 patients were used for training, and 475 clear images from 72 patients were utilized for validation. Further, 20 bronchoscopic videos of examination procedures in another 20 patients with normal airway structures were used to extract the bronchoscopic images of normal anatomical positions to evaluate the accuracy for the model. Finally, 21 respiratory doctors were enrolled for the test of recognizing corrected anatomical positions using the validating datasets. Results: In all, 1527 bronchoscopic images of 200 patients with nine anatomical positions of the airway, including carina, right main bronchus, right upper lobe bronchus, right intermediate bronchus, right middle lobe bronchus, right lower lobe bronchus, left main bronchus, left upper lobe bronchus, and left lower lobe bronchus, were used for supervised machine learning and training, and 475 clear bronchoscopic images of 72 patients were used for validation. The mean accuracy of recognizing these 9 positions was 91% (carina: 98%, right main bronchus: 98%, right intermediate bronchus: 90%, right upper lobe bronchus: 91%, right middle lobe bronchus 92%, right lower lobe bronchus: 83%, left main bronchus: 89%, left upper bronchus: 91%, left lower bronchus: 76%). The area under the curves for these nine positions were >0.98. In addition, the accuracy of extracting the images via the video by the trained model was 94.7%. We also conducted a deep learning study to segment 10 segment bronchi in right lung, and 8 segment bronchi in Left lung. Because of the problem of radial depth, only segment bronchi distributions below right upper bronchus and right middle bronchus could be correctly recognized. The accuracy of recognizing was 84.33 ± 7.52% by doctors receiving interventional pulmonology education in our hospital over 6 months. Conclusion: Our study proved that AI technology can be used to distinguish the normal anatomical positions of the airway, and the model we trained could extract the corrected images via the video to help standardize data collection and control quality.