Stabilization of CsPbBr3 Nanowires Through SU-8 Encapsulation for the Fabrication of Bilayer Microswimmers with Magnetic and Fluorescence Properties.

All-inorganic cesium lead halide (CsPbX3, X = Cl, Br, I) perovskite nanocrystals have drawn great interest because of their excellent photophysical properties and potential applications. However, their poor stability in water greatly limited their use in applications that require stable structures. In this work, a facile approach to stabilize CsPbBr3 nanowires is developed by using SU-8 as a protection medium; thereby creating stable CsPbBr3/SU-8 microstructures. Through photolithography and layer-by-layer deposition, CsPbBr3/SU-8 is used to fabricate bilayer achiral microswimmers (BAMs), which consist of a top CsPbBr3/SU-8 layer and a bottom Fe3O4 magnetic layer. Compared to pure CsPbBr3 nanowires, the CsPbBr3/SU-8 shows long-term structural and fluorescence stability in water against ultrasonication treatment. Due to the magnetic layer, the motion of the microswimmers can be controlled precisely under a rotating magnetic field, allowing them to swim at low Reynolds number and tumble or roll on surfaces. Furthermore, CsPbBr3/SU-8 can be used to fabricate various types of planar microstructures with high throughput, high consistency, and fluorescence properties. This work provides a method for the stabilization of CsPbBr3 and demonstrates the potential to mass fabricate planar microstructures with various shapes, which can be used in different applications such as microrobotics.

A dataset of trophic state index for nation-scale lakes in China from 40-year Landsat observations.

Trophic state index (TSI) serves as a key indicator for quantifying and understanding the lake eutrophication, which has not been fully explored for long-term water quality monitoring, especially for small and medium inland waters. Landsat satellites offer an effective complement to facilitate the temporal and spatial monitoring of multi-scale lakes. Landsat surface reflectance products were utilized to retrieve the annual average TSI for 2693 lakes over 1 km2 in China from 1984 to 2023. Our method first distinguishes lake types by pixels with a decision tree and then derives relationships between trophic state and algal biomass index. Validation with public reports and existing datasets confirmed the good consistency and reliability. The dataset provides reliable annual TSI results and credible trends for lakes under different area scales, which can serve as a reference for further research and provide convenience for lake sustainable management.

Eutrophication evolution of lakes in China: Four decades of observations from space.

The lake eutrophication is highly variable in both time and location, and greatly restricts the sustainable development of water resources. The lack of national eutrophication evaluation for multi-scale lakes limits the pertinent governance and sustainable management of water quality. In this study, a remote sensing approach was developed to capture 40-year dynamics of trophic state index (TSI) for nationwide lakes in China. 32% of lakes (N = 1925) in China were eutrophic and 26% were oligotrophic, and a longitudinal pattern was discovered, with the 40-year average TSI of 62.26 in the eastern plain compared to 23.72 in the Tibetan Plateau. A decreasing trend was further observed in the past four decades with a correlation of -0.16, which was mainly discovered in the Tibetan Plateau lakes (r > -0.90, p < 0.01). The contribution of climate change and human activities was quantified and varied between lake zones, with anthropogenic factors playing a dominant role in the east plain lakes (88%, N = 473) and large lakes are subject to a more complex driving mechanism (≥ 3 driving factors). The study expands the spatiotemporal scale for eutrophication monitoring and provides an important base for strengthening lake management and ecological services.

Tracing sediment sources in a plain river network area by using optimized experimental design and reflectance spectroscopy.

Soil erosion in a plain river network area with dense rivers, fertile land, and agricultural development is easily causes river siltation, agricultural non-point source pollution, and water eutrophication. Therefore, the negative impact of the sediment on the environment cannot be underestimated. Most traditional sediment fingerprint tracing studies have focused on mountain basins and lack a scheme suitable for plain river network sediment tracing. Here, a typical plain river network in the Taihu Basin was selected as the study area. The flow structure and characteristics were analysed, and a sampling scheme for the stream segment and a two-step model of sediment tracing in a plain river network were proposed to quantitatively distinguish the types of sediment sources. The results indicated that the traditional discriminant function analysis adequately distinguishes the contribution rate of basin soil and has a good validation accuracy (R2 = 0.96, root mean square error of calibration = 5.91 %), whereas Random Forest obtains better discrimination results by mining non-linear information in the soil spectra of different land types, with R2 values of 0.89, 0.83, and 0.80 for farmland, forest, and grassland, respectively. The average proportion of soil in the sediment in the watershed was 23 %, and the proportion of soil in the watershed increased from upstream to downstream. The sediment sources of the Caoqiao, Yincun, and Shaoxiang Rivers mainly came from grassland (44 %), forest (39 %), and farmland (42 %), respectively. Land-use distribution, water conservation facilities, and soil particle size were the main factors affecting these sources. Each river adopts measures to remove the corresponding pollutants, optimise water and soil conservation measures for riverbank green belts and forest, and regularly clean up silt in water conservancy ditches and rivers, which can reduce the pollution impact caused by sediment.

IGF1R Contributes to Cell Proliferation in ALK-Mutated Neuroblastoma with Preference for Activating the PI3K-AKT Signaling Pathway.

Aberrant activation of anaplastic lymphoma kinase (ALK) by activating point mutation or amplification drives 5-12% of neuroblastoma (NB). Previous work has identified the involvement of the insulin-like growth factor 1 receptor (IGF1R) receptor tyrosine kinase (RTK) in a wide range of cancers. We show here that many NB cell lines exhibit IGF1R activity, and that IGF1R inhibition led to decreased cell proliferation to varying degrees in ALK-driven NB cells. Furthermore, combined inhibition of ALK and IGF1R resulted in synergistic anti-proliferation effects, in particular in ALK-mutated NB cells. Mechanistically, both ALK and IGF1R contribute significantly to the activation of downstream PI3K-AKT and RAS-MAPK signaling pathways in ALK-mutated NB cells. However, these two RTKs employ a differential repertoire of adaptor proteins to mediate downstream signaling effects. We show here that ALK signaling led to activation of the RAS-MAPK pathway by preferentially phosphorylating the adaptor proteins GAB1, GAB2, and FRS2, while IGF1R signaling preferentially phosphorylated IRS2, promoting activation of the PI3K-AKT pathway. Together, these findings reveal a potentially important role of the IGF1R RTK in ALK-mutated NB and that co-targeting of ALK and IGF1R may be advantageous in clinical treatment of ALK-mutated NB patients.

Integrating scRNA and bulk-RNA sequencing develops a cell senescence signature for analyzing tumor heterogeneity in clear cell renal cell carcinoma.

Introduction: Cellular senescence (CS) plays a critical role in cancer development, including clear cell renal cell carcinoma (ccRCC). Traditional RNA sequencing cannot detect precise molecular composition changes within tumors. This study aimed to analyze cellular senescence's biochemical characteristics in ccRCC using single RNA sequencing (ScRNA-seq) and traditional RNA sequencing (Bulk RNA-seq). Methods: Researchers analyzed the biochemical characteristics of cellular senescence in ccRCC using ScRNA-seq and Bulk RNA-seq. They combined these approaches to identify differences between malignant and non-malignant phenotypes in ccRCC across three senescence-related pathways. Genes from these pathways were used to identify molecular subtypes associated with senescence, and a new risk model was constructed. The function of the gene DUSP1 in ccRCC was validated through biological experiments. Results: The combined analysis of ScRNA-seq and Bulk RNA-seq revealed significant differences between malignant and non-malignant phenotypes in ccRCC across three senescence-related pathways. Researchers identified genes from these pathways to identify molecular subtypes associated with senescence, constructing a new risk model. Different subgroups showed significant differences in prognosis level, clinical stage and grade, immune infiltration, immunotherapy, and drug sensitivity. Discussion: Senescence signature markers are practical biomarkers and predictors of molecular typing in ccRCC. Differences in prognosis level, clinical stage and grade, immune infiltration, immunotherapy, and drug sensitivity between different subgroups indicate that this approach could provide valuable insights into senescence-related treatment options and prognostic assessment for patients with ccRCC. The function of the gene DUSP1 in ccRCC was validated through biological experiments, confirming its feasibility as a novel biomarker for ccRCC. These findings suggest that targeted therapies based on senescence-related mechanisms could be an effective treatment option for ccRCC.

Vertical distribution analysis and total mass estimation of nitrogen and phosphorus in large shallow lakes.

Analysing the vertical distribution of nutrient salts and estimating the total mass of lake nutrients is helpful for the management of lake nutrient status and the formulation of drainage standards in basins. However, studies on nitrogen (N) and phosphorus (P) in lakes have focused on obtaining measures of N and P concentrations, but no understanding exists on the vertical distribution of N and P in the entire water column. The present study proposes algorithms for estimating the total masses of N/P per unit water column (ALGO-TNmass/ALGO-TPmass) for shallow eutrophic lakes. Using Lake Taihu as an example, the total masses of nutrients in Lake Taihu in the historical period were obtained, and the algorithm performance was discussed. The results showed that the vertical distribution of nutrients decreased with increasing depth and exhibited a quadratic distribution. Surface nutrients and chlorophyll-a concentrations play important roles in the vertical distribution of nutrients. Based on conventional surface water quality indicators, algorithms for the vertical nutrient concentration in Lake Taihu were proposed. Both algorithms had good accuracy (ALGO-TNmass R2 > 0.75, RMSE <0.57; ALGO-TPmass R2 > 0.80, RMSE ≤0.50), the ALGO-TPmass had better applicability than the ALGO-TNmass, and had good accuracy in other shallow lakes. Therefore, deducing the TPmass using conventional water quality indicators in surface water, which not only simplifies the sampling process but also provides an opportunity for remote sensing technology to monitor the total masses of nutrients, is feasible. The long-term average total mass of N was 11,727 t, showing a gradual downward trend before 2010, after which it stabilised. The maximum and minimum intra-annual total N masses were observed in May and November, respectively. The long-term average total mass of P was 512 t, showing a gradual downward trend before 2010, and a slow upward trend thereafter. The maximum and minimum intra-annual total masses of P occurred in August and February or May, respectively. The correlation between the total mass of N and meteorological conditions was not obvious, whereas some influence on the total mass of P was evident, particularly water level and wind speed.

Horizontal and vertical migration of cyanobacterial blooms in two eutrophic lakes observed from the GOCI satellite.

Under the variations of natural conditions (temperature, wind speed, light, et al.) and self-regulation of buoyancy, cyanobacterial blooms can change rapidly in a short time. The Geostationary Ocean Color Imager (GOCI) can provide hourly monitoring of the dynamics of algal blooms (eight times per day), and has potential in observing the horizontal and vertical movement of cyanobacterial blooms. Based on the fractional floating algae cover (FAC), the diurnal dynamics and migration of floating algal blooms were evaluated, and the horizontal and vertical migration speed of phytoplankton was estimated from the proposed algorithm in two eutrophic lakes, Lake Taihu and Lake Chaohu in China. The locations, number, and area of algal bloom patches showed the hotspots and horizontal movement of bloom patches. The spatial and seasonal variations of the vertical velocities indicated that both the rising and sinking speed were higher in summer and autumn than those in spring and winter. The factors affecting diurnal horizontal and vertical migrations of phytoplankton were analyzed. Diffuse horizontal irradiance (DHI), direct normal irradiance (DNI), and temperature had significant positive relationships with FAC in the morning. Wind speed contributed 18.3 and 15.1% to the horizontal movement speed in Lake Taihu and Lake Chaohu, respectively. The rising speed was more related to DNI and DHI in Lake Taihu and Lake Chaohu with contribution of 18.1 and 16.6%. The horizontal and vertical movement of algae provide important information for understanding phytoplankton dynamics and the prediction and warning of algal blooms in lake management.

Magnetically-actuated hydrogel-based achiral planar microswimmers for SERS detection: In situ coprecipitation for continuous loading of iron oxide nanoparticles.

Ultraviolet lithography is a very promising technology used for the batch fabrication of biomedical microswimmers. However, creating microswimmers that can swim at low Reynolds number using biocompatible materials while retaining strong magnetic properties and excellent biomedical functionality is a great challenge. Most of the previously reported biomedical microswimmers possess either strong magnetic properties by using non-biocompatible nickel coating or good biocompatibility by using iron oxide particle-embedded hydrogel with weak magnetism, but not both. Alternatively, iron oxide nanoparticles can be coated on the surface of microswimmers to improve magnetic properties; however, this method limited the usability of the microswimmers' surfaces. To address these shortcomings, this work utilized an in situ synthesis technique to generate high magnetic content inside hydrogel-based achiral planar microswimmers while leaving their surfaces free to be functionalized for SERS detection. The hydrogel matrices of the magnetically actuated hydrogel-based microswimmers were first prepared by ultraviolet lithography. Then, the high concentration of iron oxide was achieved through multiple continuous in situ coprecipitation cycles. Finally, the SERS detection capability of magnetically actuated hydrogel-based microswimmers was enabled by uniformly growing silver nanoparticles on the surface of the microswimmers. In the motion control tests, the microswimmers showed a high swimming efficiency, high step-out frequency, and consistent synchronized motion. Furthermore, the magnetically actuated hydrogel-based microswimmers were able to improve the detection efficiency of analytes under magnetic guidance.

Source appointment at large-scale and ungauged catchment using physically-based model and dynamic export coefficient.

Data scarcity has caused enormous problems in non-point pollution predictions and the related source apportionment. In this study, a new framework was developed to undertake the source apportionment at a large-scale and ungauged catchment, by integrating the physically-based model and a surrogate model. The improvements were made, in terms of the application of a physically-based model in an ungauged area for the transfer process and the parametric transplantation process. The new framework was then tested in the Chaohu Lake basin, China. The result suggested that there has been a good match between simulated and observed data. Although the planting industry was the largest emission source with 48.16% of nitrogen (N), itonly contributed 12.61% of N flux to the Chaohu Lake. The ungauged catchments surrounding the Chaohu Lake were identified as non-negligible sources with 8.46% of phosphorus (P) contribution. The rainfall conditions could have great impacts on source apportionment results; e.g., the planting industry contributed from 68.17t of P in dry year to 436.02t in wet year. The new framework could be extended to other large-scale watersheds for source apportionment with data limitations.

Fabrication of Bilayer Magnetically Actuated L-Shaped Microrobot Based on Chitosan via Photolithography.

Magnetically actuated microrobots showed increasing potential in various fields, especially in the biomedical area, such as invasive surgery, targeted cargo delivery, and treatment. However, it remains a challenge to incorporate biocompatible natural polymers that are favorable for practical biomedical applications. In this work, bilayer magnetic microrobots with an achiral planar design were fabricated using a biocompatible natural polymer and Fe3O4 nanoparticles through the photolithography by applying the layer-by-layer method. The microrobots consisted of a magnetic bottom layer and a photo-crosslinked chitosan top layer. The SEM results showed that the microrobot processed the L-shaped planar structure with the average width, length, and thickness of 99.18 ± 5.11 µm, 189.56 ± 11.37 µm, and 23.56 ± 4.08 µm, respectively. Moreover, microrobots actuated using a three-dimensional (3D) Helmholtz coil system was characterized and reached up to an average maximum velocity of 325.30 µm/s and a step-out frequency of 14 Hz. Furthermore, the microrobots exhibited excellent cell biocompatibility towards L929 cells in the CCK-8 assay. Therefore, the development of bi-layered chitosan-based microrobots offers a general solution for using magnetic microrobots in biomedical applications by providing an easy-to-fabricate, highly mobile microrobotic platform with the incorporation of biocompatible natural polymers for enhanced biocompatibility.

The therapeutic effect of Taijiquan combined with acupoint pressing on the treatment of anxiety insomnia in college students: A study protocol for a randomized controlled trial.

Introduction: Sleep health is an important part of health and has become a common concern of society. For anxiety insomnia, the commonly used clinical therapies have limitations. Alternative and complementary therapy is gradually rising and showing remarkable effect in clinical practice. This is the first study to evaluate the therapeutic effect of Taijiquan combined with acupoint pressing in the treatment of anxiety insomnia in college students and to compare the difference in intervention before and after sleep, to choose the best treatment time. Methods and analysis: This is a multicenter, single-blind, randomized controlled trial. A total of 126 eligible subjects who have passed the psychological evaluation and met inclusion criteria by completing a psychometric scale will be randomly divided into treatment group A (treat before sleep), treatment group B (treat after sleep) and control group C (waiting list group) in a ratio of 1:1:1. All the three groups will receive regular psychological counseling during the trial, and the treatment groups will practice 24-style Taijiquan and do meridian acupuncture at Baihui (DU20), Shenting (DU24), Yintang (EX-HN3), Shenmen (HT7) and Sanyinjiao (SP6). This RCT includes a 2-week baseline period, a 12-week intervention period, and a 12-week follow-up period. The main results will be measured by changes in the Pittsburgh sleep quality index (PSQI) and Hamilton anxiety scale (HAMA). The secondary results will be measured by the generalized anxiety scale (GAD-7) and insomnia severity index (ISI). The safety of the intervention will be evaluated at each assessment. The statistical analysis of data will be carried out by SPSSV.26.0 software. Discussion: We expect this trial to explore the effectiveness of Taijiquan combined with acupoint pressing in the treatment of anxiety insomnia in college students and choose the best treatment time by comparison. Clinical trial registration: [www.ClinicalTrials.gov], identifier [ChiCTR2200057003].

Stop-Flow Lithography for the Continuous Production of Degradable Hydrogel Achiral Crescent Microswimmers.

The small size of robotic microswimmers makes them suitable for performing biomedical tasks in tiny, enclosed spaces. Considering the effects of potentially long-term retention of microswimmers in biological tissues and the environment, the degradability of microswimmers has become one of the pressing issues in this field. While degradable hydrogel was successfully used to prepare microswimmers in previous reports, most hydrogel microswimmers could only be fabricated using two-photon polymerization (TPP) due to their 3D structures, resulting in costly robotic microswimmers solution. This limits the potential of hydrogel microswimmers to be used in applications where a large number of microswimmers are needed. Here, we proposed a new type of preparation method for degradable hydrogel achiral crescent microswimmers using a custom-built stop-flow lithography (SFL) setup. The degradability of the hydrogel crescent microswimmers was quantitatively analyzed, and the degradation rate in sodium hydroxide solution (NaOH) of different concentrations was investigated. Cytotoxicity assays showed the hydrogel crescent microswimmers had good biocompatibility. The hydrogel crescent microswimmers were magnetically actuated using a 3D Helmholtz coil system and were able to obtain a swimming efficiency on par with previously reported microswimmers. The results herein demonstrated the potential for the degradable hydrogel achiral microswimmers to become a candidate for microscale applications.

Development of remote sensing algorithm for total phosphorus concentration in eutrophic lakes: Conventional or machine learning?

Phosphorus is a limiting nutrient in freshwater ecosystems. Therefore, the estimation of total phosphorus (TP) concentration in eutrophic water using remote sensing technology is of great significance for lake environmental management. However, there is no TP remote sensing model for lake groups, and thus far, specific models have been used for specific lakes. To address this issue, this study proposes a framework for TP estimation. First, three algorithm development frameworks were compared and applied to the development of an algorithm for Lake Taihu, which has complex water environment characteristics and is a representative of eutrophic lakes. An Extremely Gradient Boosting (BST) machine learning framework was proposed for developing the Taihu TP algorithm. The machine learning algorithm could mine the relationship between FAI and TP in Lake Taihu, where the optical properties of the water body are dominated by phytoplankton. The algorithm exhibited robust performance with an R2 value of 0.6 (RMSE = 0.07 mg/L, MRE = 43.33%). Then, a general TP algorithm (R2 = 0.64, RMSE = 0.06 mg/L, MRE = 34.13%) was developed using the proposed framework and tested in seven other lakes using synchronous image data. The algorithm accuracy was found to be affected by aquatic vegetation and enclosure aquaculture. Third, compared with field investigations in other studies on Lake Taihu, the Taihu TP algorithm showed good performance for long-term TP estimation. Therefore, the machine learning framework developed in this study has application potential in large-scale spatio-temporal TP estimation in eutrophic lakes.

Predicting EGFR mutation status in lung adenocarcinoma presenting as ground-glass opacity: utilizing radiomics model in clinical translation.

OBJECTIVES: This study aimed to establish a non-invasive radiomics model based on computed tomography (CT), with favorable sensitivity and specificity to predict EGFR mutation status in GGO-featured lung adenocarcinoma subsequently guiding the administration of targeted therapy. METHODS: Clinical-pathological information and preoperative CT images of 636 lung adenocarcinoma patients (464, 100, and 72 in the training, internal, and external validation sets, respectively) that underwent GGO lesions resection were included. A total of 1476 radiomics features were extracted with gradient boosting decision tree (GBDT). RESULTS: The established radiomics model containing 102 selected features showed an encouraging discrimination performance of EGFR mutation status (mutant or wild type), and the predictive ability was superior to that of the clinical model (AUC: 0.838 vs. 0.674, 0.822 vs. 0.730, and 0.803 vs. 0.746 for the training, internal validation, and external validation sets, respectively). The combined radiomics plus clinical model showed no additional benefit over the radiomics model in predicting EGFR status (AUC: 0.846 vs. 0.838, 0.816 vs. 0.822, and 0.811 vs. 0.803, respectively, in three cohorts). Uniquely, this model was validated in a cohort of lung adenocarcinoma patients who have undertaken adjuvant EGFR-TKI treatment and harbored unresected GGOs during the medication, leading to a significantly improved potency of EGFR-TKIs (response rate: 25.9% vs. 53.8%, p = 0.006; before and after prediction, respectively). CONCLUSION: This presented radiomics model can be served as a non-invasive and time-saving approach for predicting the EGFR mutation status in lung adenocarcinoma presenting as GGO. KEY POINTS: • We developed a GGO-specific radiomics model containing 102 radiomics features for EGFR mutation status differentiation. • An AUC of 0.822 and 0.803 in the internal and external validation cohorts, respectively, were achieved. • The radiomics model was utilized in clinical translation in an adjuvant EGFR-TKI treatment cohort with unresected GGOs. A significant improvement in the potency of EGFR-TKIs was achieved (response rate: 25.9% vs. 53.8%, p = 0.006; before and after prediction).

Mix-and-Interpolate: A Training Strategy to Deal With Source-Biased Medical Data.

Till March 31st, 2021, the coronavirus disease 2019 (COVID-19) had reportedly infected more than 127 million people and caused over 2.5 million deaths worldwide. Timely diagnosis of COVID-19 is crucial for management of individual patients as well as containment of the highly contagious disease. Having realized the clinical value of non-contrast chest computed tomography (CT) for diagnosis of COVID-19, deep learning (DL) based automated methods have been proposed to aid the radiologists in reading the huge quantities of CT exams as a result of the pandemic. In this work, we address an overlooked problem for training deep convolutional neural networks for COVID-19 classification using real-world multi-source data, namely, the data source bias problem. The data source bias problem refers to the situation in which certain sources of data comprise only a single class of data, and training with such source-biased data may make the DL models learn to distinguish data sources instead of COVID-19. To overcome this problem, we propose MIx-aNd-Interpolate (MINI), a conceptually simple, easy-to-implement, efficient yet effective training strategy. The proposed MINI approach generates volumes of the absent class by combining the samples collected from different hospitals, which enlarges the sample space of the original source-biased dataset. Experimental results on a large collection of real patient data (1,221 COVID-19 and 1,520 negative CT images, and the latter consisting of 786 community acquired pneumonia and 734 non-pneumonia) from eight hospitals and health institutions show that: 1) MINI can improve COVID-19 classification performance upon the baseline (which does not deal with the source bias), and 2) MINI is superior to competing methods in terms of the extent of improvement.

Predicting Tyrosine Kinase Inhibitor Treatment Response in Stage IV Lung Adenocarcinoma Patients With EGFR Mutation Using Model-Based Deep Transfer Learning.

BACKGROUND: For stage IV patients harboring EGFR mutations, there is a differential response to the first-line TKI treatment. We constructed three-dimensional convolutional neural networks (CNN) with deep transfer learning to stratify patients into subgroups with different response and progression risks. MATERIALS AND METHODS: From 2013 to 2017, 339 patients with EGFR mutation receiving first-line TKI treatment were included. Progression-free survival (PFS) time and progression patterns were confirmed by routine follow-up and restaging examinations. Patients were divided into two subgroups according to the median PFS (<=9 months, > 9 months). We developed a PFS prediction model and a progression pattern classification model using transfer learning from a pre-trained EGFR mutation classification 3D CNN. Clinical features were fused with the 3D CNN to build the final hybrid prediction model. The performance was quantified using area under receiver operating characteristic curve (AUC), and model performance was compared by AUCs with Delong test. RESULTS: The PFS prediction CNN showed an AUC of 0.744 (95% CI, 0.645-0.843) in the independent validation set and the hybrid model of CNNs and clinical features showed an AUC of 0.771 (95% CI, 0.676-0.866), which are significantly better than clinical features-based model (AUC, 0.624, P<0.01). The progression pattern prediction model showed an AUC of 0.762(95% CI, 0.643-0.882) and the hybrid model with clinical features showed an AUC of 0.794 (95% CI, 0.681-0.908), which can provide compensate information for clinical features-based model (AUC, 0.710; 95% CI, 0.582-0.839). CONCLUSION: The CNN exhibits potential ability to stratify progression status in patients with EGFR mutation treated with first-line TKI, which might help make clinical decisions.

Predicting Malignancy and Invasiveness of Pulmonary Subsolid Nodules on CT Images Using Deep Learning.

BACKGROUND: To develop and validate a deep learning-based model on CT images for the malignancy and invasiveness prediction of pulmonary subsolid nodules (SSNs). MATERIALS AND METHODS: This study retrospectively collected patients with pulmonary SSNs treated by surgery in our hospital from 2012 to 2018. Postoperative pathology was used as the diagnostic reference standard. Three-dimensional convolutional neural network (3D CNN) models were constructed using preoperative CT images to predict the malignancy and invasiveness of SSNs. Then, an observer reader study conducted by two thoracic radiologists was used to compare with the CNN model. The diagnostic power of the models was evaluated with receiver operating characteristic curve (ROC) analysis. RESULTS: A total of 2,614 patients were finally included and randomly divided for training (60.9%), validation (19.1%), and testing (20%). For the benign and malignant classification, the best 3D CNN model achieved a satisfactory AUC of 0.913 (95% CI: 0.885-0.940), sensitivity of 86.1%, and specificity of 83.8% at the optimal decision point, which outperformed all observer readers' performance (AUC: 0.846±0.031). For pre-invasive and invasive classification of malignant SSNs, the 3D CNN also achieved satisfactory AUC of 0.908 (95% CI: 0.877-0.939), sensitivity of 87.4%, and specificity of 80.8%. CONCLUSION: The deep-learning model showed its potential to accurately identify the malignancy and invasiveness of SSNs and thus can help surgeons make treatment decisions.

Magnetic tri-bead microrobot assisted near-infrared triggered combined photothermal and chemotherapy of cancer cells.

Magnetic micro/nanorobots attracted much attention in biomedical fields because of their precise movement, manipulation, and targeting abilities. However, there is a lack of research on intelligent micro/nanorobots with stimuli-responsive drug delivery mechanisms for cancer therapy. To address this issue, we developed a type of strong covalently bound tri-bead drug delivery microrobots with NIR photothermal response azobenzene molecules attached to their carboxylic surface groups. The tri-bead microrobots are magnetic and showed good cytocompatibility even when their concentration is up to 200 µg/mL. In vitro photothermal experiments demonstrated fast NIR-responsive photothermal property; the microrobots were heated to 50 °C in 4 min, which triggered a significant increase in drug release. Motion control of the microrobots inside a microchannel demonstrated the feasibility of targeted therapy on tumor cells. Finally, experiments with lung cancer cells demonstrated the effectiveness of targeted chemo-photothermal therapy and were validated by cell viability assays. These results indicated that tri-bead microrobots have excellent potential for targeted chemo-photothermal therapy for lung cancer cell treatment.

Nanorobots-Assisted Natural Computation for Multifocal Tumor Sensitization and Targeting.

We have proposed a new tumor sensitization and targeting (TST) framework, named in vivo computation, in our previous investigations. The problem of TST for an early and microscopic tumor is interpreted from the computational perspective with nanorobots being the "natural" computing agents, the high-risk tissue being the search space, the tumor targeted being the global optimal solution, and the tumor-triggered biological gradient field (BGF) providing the aided knowledge for fitness evaluation of nanorobots. This natural computation process can be seen as on-the-fly path planning for nanorobot swarms with an unknown target position, which is different from the traditional path planning methods. Our previous works are focusing on the TST for a solitary lesion, where we proposed the weak priority evolution strategy (WP-ES) to adapt to the actuating mode of the homogeneous magnetic field used in the state-of-the-art nanorobotic platforms, and some in vitro validations were performed. In this paper, we focus on the problem of TST for multifocal tumors, which can be seen as a multimodal optimization problem for the "natural" computation. To overcome this issue, we propose a sequential targeting strategy (Se-TS) to complete TST for the multiple lesions with the assistance of nanorobot swarms, which are maneuvered by the external actuating and tracking devices according to the WP-ES. The Se-TS is used to modify the BGF landscape after a tumor is detected by a nanorobot swarm with the gathered BGF information around the detected tumor. Next, another nanorobot swarm will be employed to find the second tumor according to the modified BGF landscape without being misguided to the previous one. In this way, all the tumor lesions will be detected one by one. In other words, the paths of nanorobots to find the targets can be generated successively with the sequential modification of the BGF landscape. To demonstrate the effectiveness of the proposed Se-TS, we perform comprehensive simulation studies by enhancing the WP-ES based swarm intelligence algorithms using this strategy considering the realistic in-body constraints. The performance is compared against that of the "brute-force" search, which corresponds to the traditional systemic tumor targeting, and also against that of the standard swarm intelligence algorithms from the algorithmic perspective. Furthermore, some in vitro experiments are performed by using Janus microparticles as magnetic nanorobots, a two-dimensional microchannel network as the human vasculature, and a magnetic nanorobotic control system as the external actuating and tracking system. Results from the in silico simulations and in vitro experiments verify the effectiveness of the proposed Se-TS for two representative BGF landscapes.