Mex-3 RNA binding family member A (MEX3A)/circMPP6 complex promotes colorectal cancer progression by inhibiting autophagy.

RNA-binding proteins (RBPs)-RNA networks have contributed to cancer development. Circular RNAs (circRNAs) are considered as protein recruiters; nevertheless, the patterns of circRNA-protein interactions in colorectal cancer (CRC) are still lacking. Processing bodies (PBs) formed through liquid-liquid phase separation (LLPS) are membrane-less organelles (MLOs) consisting of RBPs and RNA. Previous evidence suggests a connection between PBs dynamics and cancer progression. Despite the increasingly acknowledged crucial role of RBPs and RNA in the accumulation and maintenance of MLOs, there remains a lack of specific research on the interactions between PBs-related RBPs and circRNAs in CRC. Herein, we identify that MEX-3 RNA binding family member A (MEX3A), frequently upregulated in CRC tissues, predicts poorer patient survival. Elevated MEX3A accelerates malignance and inhibits autophagy of CRC cells. Importantly, MEX3A undergoes intrinsically disordered regions (IDRs)-dependent LLPS in the cytoplasm. Specifically, circMPP6 acts as a scaffold to facilitate the interaction between MEX3A and PBs proteins. The MEX3A/circMPP6 complex modulates PBs dynamic and promotes UPF-mediated phosphodiesterase 5A (PDE5A) mRNA degradation, consequently leading to the aggressive properties of CRC cells. Clinically, CRC patients exhibiting high MEX3A expression and low PDE5A expression have the poorest overall survival. Our findings reveal a collaboration between MEX3A and circMPP6 in the regulation of mRNA decay through triggering the PBs aggregation, which provides prognostic markers and/or therapeutic targets for CRC.

LncRNA RGMB-AS1 inhibits HMOX1 ubiquitination and NAA10 activation to induce ferroptosis in non-small cell lung cancer.

Ferroptosis, an iron-dependent regulated cell death caused by excessive lipid peroxide accumulation, has emerged as a promising therapeutic target in various cancers, including non-small cell lung cancer (NSCLC). In this study, we identified the long non-coding RNA RGMB-AS1 as a key regulator of ferroptosis in NSCLC. Mechanistically, RGMB-AS1 interacted with heme oxygenase 1 (HMOX1) and prevented its ubiquitination by the E3 ligase TRC8, leading to increased HMOX1 stability and enhanced ferroptosis. Additionally, RGMB-AS1 bound to the 82-87 amino acid region of N-alpha-acetyltransferase 10 (NAA10), stimulating its acetyltransferase activity and promoting the conversion of acetyl-CoA to HMG-CoA, further contributing to ferroptosis. The RGMB-AS1-HMOX1 and RGMB-AS1-NAA10 axes synergistically inhibited NSCLC growth both in vitro and in vivo. Clinically, low RGMB-AS1 expression was associated with advanced tumor stage and poor overall survival in NSCLC patients. Furthermore, adeno-associated virus-mediated RGMB-AS1 overexpression significantly suppressed tumor growth in mouse xenograft models. Our findings uncover a novel lncRNA-mediated regulatory mechanism of ferroptosis and highlight the potential of RGMB-AS1 as a prognostic biomarker and therapeutic target in NSCLC.

Targeted Ultrasound Nanobubbles Therapy for Prostate Cancer via Immuno-Sonodynamic Effect.

Background: Prostate cancer (PCa) poses a significant global health threaten. Immunotherapy has emerged as a novel strategy to augment the inhibition of tumor proliferation. However, the sole use of anti-PD-L1 Ab for PCa has not yielded improvements, mirroring outcomes observed in other tumor types. Methods: This study employed the thin film hydration method to develop lipid nanobubbles (NBs) encapsulating chlorin e6 (Ce6) and anti-PD-L1 Ab (Ce6@aPD-L1 NBs). Our experimental approach included cellular assays and mouse immunization, providing a comprehensive evaluation of Ce6@aPD-L1 NBs' impact. Results: The Ce6@aPD-L1 NBs effectively induced reactive oxygen species generation, leading to tumor cells death. In mice, they demonstrated a remarkable enhancement of immune responses compared to control groups. These immune responses encompassed immunogenic cell death induced by sonodynamic therapy and PD-1/PD-L1 blockade, activating dendritic cells maturation and effectively stimulating CD8+T cells. Conclusion: Ce6@aPD-L1 NBs facilitate tumor-targeted delivery, activating anti-tumor effects through direct sonodynamic therapy action and immune system reactivation in the tumor microenvironment. Ce6@aPD-L1 NBs exhibit substantial potential for achieving synergistic anti-cancer effects in PCa.

Giant Tunability of Intersubband Transitions and Quantum Hall Quartets in Few-Layer InSe Quantum Wells.

A two-dimensional (2D) quantum electron system is characterized by quantized energy levels, or subbands, in the out-of-plane direction. Populating higher subbands and controlling the intersubband transitions have wide technological applications such as optical modulators and quantum cascade lasers. In conventional materials, however, the tunability of intersubband spacing is limited. Here we demonstrate electrostatic population and characterization of the second subband in few-layer InSe quantum wells, with giant tunability of its energy, population, and spin-orbit coupling strength, via the control of not only layer thickness but also the out-of-plane displacement field. A modulation of as much as 350% or over 250 meV is achievable, underscoring the promise of InSe for tunable infrared and THz sources, detectors, and modulators.

[Synthetic phenolic compounds perturb lipid metabolism and induce obesogenic effects].

Given continuous development in society and the economy, obesity has become a global epidemic, arousing great concern. In addition to genetic and dietary factors, exposure to environmental chemicals is associated with the occurrence and development of obesity. Current research has indicated that some chemicals with endocrine-disrupting effects can affect lipid metabolism in vivo, causing elevated lipid storage. These chemicals are called "environmental obesogens". Synthetic phenolic compounds (SPCs) are widely used in industrial and daily products, such as plastic products, disinfectants, pesticides, food additives, and so on. The exposure routes of SPCs to the human body may include food and water consumption, direct skin contact, etc. Their unintended exposure could cause harmful effects on human health. As a type of endocrine disruptor, SPCs interfere with adipogenesis and lipid metabolism, exhibiting the characteristics of environmental obesogens. Because SPCs have similar phenolic structures, gathering information on their influences on lipid metabolism would be helpful to understand their structure-related effects. In this review, three commonly used research methods for screening environmental obesogens, including in vitro testing for molecular interactions, cell adipogenic differentiation models, and in vivo studies on lipid metabolism, are summarized, and the advantages and disadvantages of these methods are compared and discussed. Based on both in vitro and in vivo data, three types of SPCs, including bisphenol A (BPA) and its analogues, alkylphenols (APs), and synthetic phenolic antioxidants (SPAs), are systematically discussed in terms of their ability to disrupt adipogenesis and lipid metabolism by focusing on adipose and hepatic tissues, among others. Common findings on the effects of these SPCs on adipocyte differentiation, lipid storage, hepatic lipid accumulation, and liver steatosis are described. The underlying toxicological mechanisms are also discussed from the aspects of nuclear receptor transactivation, inflammation and oxidative stress regulation, intestinal microenvironment alteration, epigenetic modification, and some other signaling pathways. Future research to increase public knowledge on the obesogenic effects of emerging chemicals of concern is encouraged.

Hippocampal Subregions Volume and Texture for the Diagnosis of Mild Cognitive Impairment.

The aim was to examine the diagnostic efficacy of hippocampal subregions volume and texture in differentiating amnestic mild cognitive impairment (MCI) from normal aging changes. Ninety MCI subjects and eighty-eight well-matched healthy controls (HCs) were selected. Twelve hippocampal subregions volume and texture features were extracted using Freesurfer and MaZda based on T1 weighted MRI. Then, two-sample t-test and Least Absolute Shrinkage and Selection Operator (LASSO) regression were developed to select a subset of the original features. Support vector machine (SVM) was used to perform the classification task and the area under the curve (AUC), sensitivity, specificity and accuracy were calculated to evaluate the diagnostic efficacy of the model. The volume features with high discriminative power were mainly located in the bilateral CA1 and CA4, while texture feature were gray-level non-uniformity, run length non-uniformity and fraction. Our model based on hippocampal subregions volume and texture features achieved better classification performance with an AUC of 0.90. The volume and texture of hippocampal subregions can be utilized for the diagnosis of MCI. Moreover, we found that the features that contributed most to the model were mainly textural features, followed by volume. These results may guide future studies using structural scans to classify patients with MCI.

Correlation Analysis of KCNQ1 Gene Polymorphism with Blood Indexes and Prognosis of Non-Small Cell Lung Cancer.

BACKGROUND: This study was conducted to investigate the correlation between KCNQ1 rs2237895 A/C gene polymorphism and blood indexes and prognosis in non-small cell lung cancer (NSCLC). METHODS: A total of 260 NSCLC patients were selected and classified into stage I - II (n = 109) and stage III - IV (n = 151) according to by American Joint Committee on Cancer Staging Manual. A control group was established with another 92 healthy subjects. The genotype distribution of rs2237895 was analyzed in all subjects. 2 analysis or Fisher's test was employed to analyze the association between genotype and allele distribution frequencies with carcinoembryonic antigen (CEA), squamous cell carcinoma antigen, and cytokeratin fragment 19 (CyfrA 21-1). Overall survival was compared by genotype stratification using Kaplan-Meier analysis. Univariate and multivariate Cox risk regression analyses were used to determine the prognostic value of allele C in NSCLC. RESULTS: AC/CC genotypes in NSCLC patients were associated with gender, hypertension, smoking, clinical TNM stage, lymph node metastasis, and distant metastasis. C allele was associated with higher risk levels of serum tumor markers. Patients with allele C (AC + CC) had lower overall survival than patients with genotype AA. Finally, clinical stage, lymph node metastasis, higher CEA and CyfrA 21-1 serum levels, and rs2237895 A/C gene poly-morphism were independent prognostic factors of NSCLC. CONCLUSIONS: rs2237895 A/C polymorphism of the KCNQ1 gene can be a prognostic predictor in patients with surgically treated NSCLC.

Low­intensity pulsed ultrasound accelerates diabetic wound healing by ADSC­derived exosomes via promoting the uptake of exosomes and enhancing angiogenesis.

Diabetic wounds remain a great challenge for clinicians globally as a lack of effective radical treatment often results in poor prognosis. Exosomes derived from adipose­derived stem cells (ADSC­Exos) have been explored as an appealing nanodrug delivery system in the treatment of diabetic wounds. However, the short half­life and low utilization efficiency of exosomes limit their therapeutic effects. Low­intensity pulsed ultrasound (LIPUS) provides a non­invasive mechanical stimulus to cells and exerts a number of biological effects such as cavitation and thermal effects. In the present study, whether LIPUS could enhance ADSC­Exo­mediated diabetic wound repair was investigated and its possible mechanism of action was explored. After isolation and characterization, ADSC­Exos were injected into mice with diabetic wounds, then the mice were exposed to LIPUS irradiation. The control mice were subcutaneously injected with PBS. Wound healing assays, laser Doppler perfusion, Masson's staining and angiogenesis assays were used to assess treatment efficiency. Then, ADSC­Exos were cocultured with human umbilical vein endothelial cells (HUVECs), and the proliferation, migration and tube formation of HUVECs were assessed. Moreover, the cellular uptake of ADSC­Exos in vitro and in vivo was assessed to explore the synergistic mechanisms underlying the effects of LIPUS. The in vivo results demonstrated that LIPUS increased the uptake of exosomes and prolonged the residence of exosomes in the wound area, thus enhancing angiogenesis and accelerating wound repair in diabetic mice. The in vitro results further confirmed that LIPUS enhanced the uptake efficiency of ADSC­Exos by 10.93­fold and significantly increased the proliferation, migration and tubular formation of HUVECs. Therefore, the present study indicates that LIPUS is a promising strategy to improve the therapeutic effects of ADSC­Exos in diabetic wounds by promoting the cellular uptake of exosomes and enhancing angiogenesis.

[Retracted] Potentiation of the antitumor activity of adriamycin against osteosarcoma by cannabinoid WIN­55,212­2.

[This retracts the article DOI: 10.3892/ol.2015.3525.].

Neoadjuvant nivolumab with or without platinum-doublet chemotherapy based on PD-L1 expression in resectable NSCLC (CTONG1804): a multicenter open-label phase II study.

This prospective multicenter phase II study evaluated the clinical efficacy of neoadjuvant nivolumab-exclusive (N) and nivolumab-chemotherapy (N/C) combinations based on PD-L1 expression. Eligible patients exhibited resectable clinical stage IIA-IIIB (AJCC 8th edition) NSCLC without EGFR/ALK alterations. Patients received either mono-nivolumab (N) or nivolumab + nab-paclitaxel+ carboplatin (N/C) for three cycles based on PD-L1 expression. The primary endpoint was the major pathological response (MPR). Key secondary endpoints included the pathologic complete response (pCR), objective response rate (ORR), and event-free survival (EFS). Baseline PD-L1 expression and perioperative circulating tumor DNA (ctDNA) status were correlated with pCR and EFS. Fifty-two patients were enrolled, with 46 undergoing surgeries. The MPR was 50.0% (26/52), with 25.0% (13/52) achieving pCR, and 16.7% and 66.7% for patients with PD-L1 ≥ 50% in N and N/C groups, respectively. Thirteen (25.0%) patients experienced grade 3 or higher immune-related adverse events during neoadjuvant treatment. Patients with post-neoadjuvant ctDNA negativity was more likely to have pCR (39.1%) compared with those remained positive (6.7%, odds ratio = 6.14, 95% CI 0.84-Inf, p = 0.077). With a median follow-up of 25.1 months, the 18-month EFS rate was 64.8% (95% CI 51.9-81.0%). For patients with ctDNA- vs. ctDNA + , the 18m-EFS rate was 93.8% vs 47.3% (HR, 0.15; 95% CI 0.04, 0.94; p = 0.005). Immunochemotherapy may serve as an optimal neoadjuvant treatment even for patients with PD-L1 expression ≥ 50%. ctDNA negativity following neoadjuvant treatment and surgery could help identify superior pathological and survival benefits, which requires further confirmation in a prospective clinical trial (NCT04015778).

Single-cell RNA sequencing reveals a hierarchical transcriptional regulatory network of terpenoid biosynthesis in cotton secretory glandular cells.

Plants can synthesize a wide range of terpenoids in response to various environmental cues. However, the specific regulatory mechanisms governing terpenoid biosynthesis at the cellular level remain largely elusive. In this study, we employed single-cell RNA sequencing to comprehensively characterize the transcriptome profile of cotton leaves and established a hierarchical transcriptional network regulating cell-specific terpenoid production. We observed substantial expression levels of genes associated with the biosynthesis of both volatile terpenes (such as ß-caryophyllene and ß-myrcene) and non-volatile gossypol-type terpenoids in secretory glandular cells. Moreover, two novel transcription factors, namely GoHSFA4a and GoNAC42, are identified to function downstream of the Gossypium PIGMENT GLAND FORMATION genes. Both transcription factors could directly regulate the expression of terpenoid biosynthetic genes in secretory glandular cells in response to developmental and environmental stimuli. For convenient retrieval of the single-cell RNA sequencing data generated in this study, we developed a user-friendly web server . Our findings not only offer valuable insights into the precise regulation of terpenoid biosynthesis genes in cotton leaves but also provide potential targets for cotton breeding endeavors.

Quantitative Assessment of Dimensional Evaluation and Artefacts from Filling Materials with CBCT Using Standard Phantom Roots.

OBJECTIVE: To investigate the accuracy of dimensional evaluation and representation of artefacts generated by different gutta-percha (GP) cones with or without sealer with CBCT using a reproducible, standardised phantom root methodology. METHODS: The reproducible artificial phantom roots with six root canal sizes from #25 to #50 and 0.04 taper were aligned according to the jaw curvature in a stone model for dimensional measurements. Each root was scanned while empty and filled with four types of filling materials. The specimens were scanned using the CS 9300 3D (Carestream Dental, Rochester, NY, USA) (at two different resolutions), 3D Accuitomo (J Morita, Kyoto, Japan) and NewTom VGi (Verona, Italy) CBCT systems. The hyperdense and hypodense axial slice artefacts from root canal sizes #40, #45 and #50 were recorded. RESULTS: Dimensions were significantly smaller and more accurate with CS 9300/0.09 mm voxel size than with other protocols. The hypodense band was found mostly in the CS 9300 3D system with 0.18 mm voxel size, especially in the buccal-lingual (95%) and coronal (64%) sections. The 3D Accuitomo CBCT system showed the lowest presence of the hypodense band. Areas of both light and dark artefacts were significantly larger in the coronal third than in the apical and middle thirds. CONCLUSION: Artefacts in the coronal locations and in buccal-lingual sections were more evident in the CS 9300 3D system with a 0.18-mm voxel size.

Polyphyllin D punctures hypertrophic lysosomes to reverse drug resistance of hepatocellular carcinoma by targeting acid sphingomyelinase.

Hypertrophic lysosomes are critical for tumor progression and drug resistance; however, effective and specific lysosome-targeting compounds for cancer therapy are lacking. Here we conducted a lysosomotropic pharmacophore-based in silico screen in a natural product library (2,212 compounds), and identified polyphyllin D (PD) as a novel lysosome-targeted compound. PD treatment was found to cause lysosomal damage, as evidenced by the blockade of autophagic flux, loss of lysophagy, and the release of lysosomal contents, thus exhibiting anticancer effects on hepatocellular carcinoma (HCC) cell both in vitro and in vivo. Closer mechanistic examination revealed that PD suppressed the activity of acid sphingomyelinase (SMPD1), a lysosomal phosphodieserase that catalyzes the hydrolysis of sphingomyelin to produce ceramide and phosphocholine, by directly occupying its surface groove, with Trp148 in SMPD1 acting as a major binding residue; this suppression of SMPD1 activity irreversibly triggers lysosomal injury and initiates lysosome-dependent cell death. Furthermore, PD-enhanced lysosomal membrane permeabilization to release sorafenib, augmenting the anticancer effect of sorafenib both in vivo and in vitro. Overall, our study suggests that PD can potentially be further developed as a novel autophagy inhibitor, and a combination of PD with classical chemotherapeutic anticancer drugs could represent a novel therapeutic strategy for HCC intervention.

Motion Decoupling Network for Intra-Operative Motion Estimation Under Occlusion.

In recent intelligent-robot-assisted surgery studies, an urgent issue is how to detect the motion of instruments and soft tissue accurately from intra-operative images. Although optical flow technology from computer vision is a powerful solution to the motion-tracking problem, it has difficulty obtaining the pixel-wise optical flow ground truth of real surgery videos for supervised learning. Thus, unsupervised learning methods are critical. However, current unsupervised methods face the challenge of heavy occlusion in the surgical scene. This paper proposes a novel unsupervised learning framework to estimate the motion from surgical images under occlusion. The framework consists of a Motion Decoupling Network to estimate the tissue and the instrument motion with different constraints. Notably, the network integrates a segmentation subnet that estimates the segmentation map of instruments in an unsupervised manner to obtain the occlusion region and improve the dual motion estimation. Additionally, a hybrid self-supervised strategy with occlusion completion is introduced to recover realistic vision clues. Extensive experiments on two surgical datasets show that the proposed method achieves accurate motion estimation for intra-operative scenes and outperforms other unsupervised methods, with a margin of 15% in accuracy. The average estimation error for tissue is less than 2.2 pixels on average for both surgical datasets.

Semi-Supervised Medical Image Segmentation Guided by Bi-Directional Constrained Dual-Task Consistency.

BACKGROUND: Medical image processing tasks represented by multi-object segmentation are of great significance for surgical planning, robot-assisted surgery, and surgical safety. However, the exceptionally low contrast among tissues and limited available annotated data makes developing an automatic segmentation algorithm for pelvic CT challenging. METHODS: A bi-direction constrained dual-task consistency model named PICT is proposed to improve segmentation quality by leveraging free unlabeled data. First, to learn more unmarked data features, it encourages the model prediction of the interpolated image to be consistent with the interpolation of the model prediction at the pixel, model, and data levels. Moreover, to constrain the error prediction of interpolation interference, PICT designs an auxiliary pseudo-supervision task that focuses on the underlying information of non-interpolation data. Finally, an effective loss algorithm for both consistency tasks is designed to ensure the complementary manner and produce more reliable predictions. RESULTS: Quantitative experiments show that the proposed PICT achieves 87.18%, 96.42%, and 79.41% mean DSC score on ACDC, CTPelvic1k, and the individual Multi-tissue Pelvis dataset with gains of around 0.8%, 0.5%, and 1% compared to the state-of-the-art semi-supervised method. Compared to the baseline supervised method, the PICT brings over 3-9% improvements. CONCLUSIONS: The developed PICT model can effectively leverage unlabeled data to improve segmentation quality of low contrast medical images. The segmentation result could improve the precision of surgical path planning and provide input for robot-assisted surgery.

Erlotinib versus gemcitabine plus cisplatin as neoadjuvant treatment of stage IIIA-N2 EGFR-mutant non-small-cell lung cancer: final overall survival analysis of the EMERGING-CTONG 1103 randomised phase II trial.

EMERGING-CTONG 1103 showed improved progression-free survival (PFS) with neoadjuvant erlotinib vs. chemotherapy for patients harbouring EGFR sensibility mutations and R0 resected stage IIIA-N2 non-small cell lung cancer (NSCLC) (NCT01407822). Herein, we report the final results. Recruited patients were randomly allocated 1:1 to the erlotinib group (150 mg/day orally; neoadjuvant phase for 42 days and adjuvant phase to 12 months) or to the GC group (gemcitabine 1250 mg/m2 plus cisplatin 75 mg/m2 intravenously; 2 cycles in neoadjuvant phase and 2 cycles in adjuvant phase). Objective response rate (ORR), complete pathologic response (pCR), PFS, and overall survival (OS) were assessed along with safety. Post hoc analysis was performed for subsequent treatments after disease recurrence. Among investigated 72 patients (erlotinib, n = 37; GC, n = 35), the median follow-up was 62.5 months. The median OS was 42.2 months (erlotinib) and 36.9 months (GC) (hazard ratio [HR], 0.83; 95% confidence interval [CI], 0.47-1.47; p = 0.513). The 3- and 5-year OS rates were 58.6% and 40.8% with erlotinib and 55.9% and 27.6% with GC (p3-y = 0.819, p5-y = 0.252). Subsequent treatment was administered in 71.9% and 81.8% of patients receiving erlotinib and GC, respectively; targeted therapy contributed mostly to OS (HR, 0.35; 95% CI, 0.18-0.70). After disease progression, the ORR was 53.3%, and the median PFS was 10.9 months during the EGFR-TKI rechallenge. During postoperative therapy, grade 3 or 4 adverse events (AEs) were 13.5% in the erlotinib group and 29.4% in the GC group. No serious adverse events were observed. Erlotinib exhibited clinical feasibility for resectable IIIA-N2 NSCLC over chemotherapy in the neoadjuvant setting.

CRISPR screens reveal genetic determinants of PARP inhibitor sensitivity and resistance in prostate cancer.

Prostate cancer harboring BRCA1/2 mutations are often exceptionally sensitive to PARP inhibitors. However, genomic alterations in other DNA damage response genes have not been consistently predictive of clinical response to PARP inhibition. Here, we perform genome-wide CRISPR-Cas9 knockout screens in BRCA1/2-proficient prostate cancer cells and identify previously unknown genes whose loss has a profound impact on PARP inhibitor response. Specifically, MMS22L deletion, frequently observed (up to 14%) in prostate cancer, renders cells hypersensitive to PARP inhibitors by disrupting RAD51 loading required for homologous recombination repair, although this response is TP53-dependent. Unexpectedly, loss of CHEK2 confers resistance rather than sensitivity to PARP inhibition through increased expression of BRCA2, a target of CHEK2-TP53-E2F7-mediated transcriptional repression. Combined PARP and ATR inhibition overcomes PARP inhibitor resistance caused by CHEK2 loss. Our findings may inform the use of PARP inhibitors beyond BRCA1/2-deficient tumors and support reevaluation of current biomarkers for PARP inhibition in prostate cancer.

Fast instruments and tissues segmentation of micro-neurosurgical scene using high correlative non-local network.

Surgical scene segmentation provides critical information for guidance in micro-neurosurgery. Segmentation of instruments and critical tissues contributes further to robot assisted surgery and surgical evaluation. However, due to the lack of relevant scene segmentation dataset, scale variation and local similarity, micro-neurosurgical segmentation faces many challenges. To address these issues, a high correlative non-local network (HCNNet), is proposed to aggregate multi-scale feature by optimized non-local mechanism. HCNNet adopts two-branch design to generate features of different scale efficiently, while the two branches share common weights in shallow layers. Several short-term dense concatenate (STDC) modules are combined as the backbone to capture both semantic and spatial information. Besides, a high correlative non-local module (HCNM) is designed to guide the upsampling process of the high-level feature by modeling global context generated from the low-level feature. It filters out confused pixels of different classes in the non-local correlation map. Meanwhile, a large segmentation dataset named NeuroSeg is constructed, which contains 15 types of instruments and 3 types of tissues that appear in meningioma resection surgery. The proposed HCNNet achieves the state-of-the-art performance on NeuroSeg, it reaches an inference speed of 54.85 FPS with the highest accuracy of 59.62% mIoU, 74.7% Dice, 70.55% mAcc and 87.12% aAcc.

Learning Skill Characteristics From Manipulations.

Percutaneous coronary intervention (PCI) has increasingly become the main treatment for coronary artery disease. The procedure requires high experienced skills and dexterous manipulations. However, there are few techniques to model PCI skill so far. In this study, a learning framework with local and ensemble learning is proposed to learn skill characteristics of different skill-level subjects from their PCI manipulations. Ten interventional cardiologists (four experts and six novices) were recruited to deliver a medical guidewire to two target arteries on a porcine model for in vivo studies. Simultaneously, translation and twist manipulations of thumb, forefinger, and wrist are acquired with electromagnetic (EM) and fiber-optic bend (FOB) sensors, respectively. These behavior data are then processed with wavelet packet decomposition (WPD) under 1-10 levels for feature extraction. The feature vectors are further fed into three candidate individual classifiers in the local learning layer. Furthermore, the local learning results from different manipulation behaviors are fused in the ensemble learning layer with three rule-based ensemble learning algorithms. In subject-dependent skill characteristics learning, the ensemble learning can achieve 100% accuracy, significantly outperforming the best local result (90%). Furthermore, ensemble learning can also maintain 73% accuracy in subject-independent schemes. These promising results demonstrate the great potential of the proposed method to facilitate skill learning in surgical robotics and skill assessment in clinical practice.