Moisture-enabled self-charging and voltage stabilizing supercapacitor.

Supercapacitor is highly demanded in emerging portable electronics, however, which faces frequent charging and inevitable rapid self-discharging of huge inconvenient. Here, we present a flexible moisture-powered supercapacitor (mp-SC) that capable of spontaneously moisture-enabled self-charging and persistently voltage stabilizing. Based on the synergy effect of moisture-induced ions diffusion of inner polyelectrolyte-based moist-electric generator and charges storage ability of inner graphene electrochemical capacitor, this mp-SC demonstrates the self-charged high areal capacitance of 138.3 mF cm-2 and ~96.6% voltage maintenance for 120 h. In addition, a large-scale flexible device of 72 mp-SC units connected in series achieves a self-charged 60 V voltage in air, efficiently powering various commercial electronics in practical applications. This work will provide insight into the design self-powered and ultra-long term stable supercapacitors and other energy storage devices.

Short-term efficacy and safety of neoadjuvant pyrotinib plus taxanes for early HER2-positive breast cancer: a single-arm exploratory phase II trial.

Background: The effectiveness and safety of pyrotinib have been substantiated in human epidermal growth factor receptor 2 (HER2)-positive advanced breast cancer (BC). However, the role of pyrotinib as a single HER2 blockade in neoadjuvant setting among BC patients has not been studied. The objective of this study was to evaluate the efficacy and tolerability of pyrotinib plus taxanes as a novel neoadjuvant regimen in patients with HER2-positive early or locally advanced BC. Methods: In this single-arm exploratory phase II trial, patients with treatment-naïve HER2-positive BC (stage IIA-IIIC) received pyrotinib 400 mg once daily and taxanes [docetaxel 75 mg/m2 or nanoparticle albumin-bound (nab)-paclitaxel 260 mg/m2 every 3 weeks, or paclitaxel 80 mg/m2 weekly] for a total of four 21-day cycles before surgery. Efficacy assessment was based on pathological and clinical measurements. The primary endpoint of this study was the total pathological complete response (tpCR) rate. The secondary endpoints included breast pCR (bpCR) rate, investigator-assessed objective response rate (ORR) and adverse events (AEs) profiles. Results: From 1 September 2021 to 30 December 2022, a total of 31 patients were enrolled. One patient was withdrawn due to unbearable skin rash after the second cycle of neoadjuvant therapy. The majority of the intention-to-treat (ITT) population was premenopausal (54.8%), had large tumors (90.3%) and metastatic nodes (58.1%) at diagnosis and hormone-receptor positive tumors (64.5%). Most participants used nab-paclitaxel (74.2%) and received mastectomy (67.7%) after neoadjuvant treatment. The tpCR and bpCR rates were 48.4% [95% confidence interval (CI): 30.8-66%] and 51.6% (95% CI: 34-69.2%), respectively. Grade ≥3 treatment-related AEs were observed in 16.1% (5/31) of the ITT population, including diarrhea (n=2, 6.5%), hand and foot numbness (n=1, 3.2%), loss of appetite (n=1, 3.2%), and skin rash (n=1, 3.2%). AE related dose reduction or pyrotinib interruption was not required. Conclusions: In female patients with HER2-positive non-metastatic BC, neoadjuvant pyrotinib monotherapy plus taxanes appears to show promising clinical benefit and controllable AEs [Chinese Clinical Trial Registry (ChiCTR2100050870)]. The long-term efficacy and safety of this regime warrant further verification.

A Deep Learning Decision Support Tool to Improve Risk Stratification and Reduce Unnecessary Biopsies in BI-RADS 4 Mammograms.

Purpose: To evaluate the performance of a biopsy decision support algorithmic model, the intelligent-augmented breast cancer risk calculator (iBRISK), on a multicenter patient dataset. Materials and Methods: iBRISK was previously developed by applying deep learning to clinical risk factors and mammographic descriptors from 9700 patient records at the primary institution and validated using another 1078 patients. All patients were seen from March 2006 to December 2016. In this multicenter study, iBRISK was further assessed on an independent, retrospective dataset (January 2015-June 2019) from three major health care institutions in Texas, with Breast Imaging Reporting and Data System (BI-RADS) category 4 lesions. Data were dichotomized and trichotomized to measure precision in risk stratification and probability of malignancy (POM) estimation. iBRISK score was also evaluated as a continuous predictor of malignancy, and cost savings analysis was performed. Results: The iBRISK model's accuracy was 89.5%, area under the receiver operating characteristic curve (AUC) was 0.93 (95% CI: 0.92, 0.95), sensitivity was 100%, and specificity was 81%. A total of 4209 women (median age, 56 years [IQR, 45-65 years]) were included in the multicenter dataset. Only two of 1228 patients (0.16%) in the "low" POM group had malignant lesions, while in the "high" POM group, the malignancy rate was 85.9%. iBRISK score as a continuous predictor of malignancy yielded an AUC of 0.97 (95% CI: 0.97, 0.98). Estimated potential cost savings were more than $420 million. Conclusion: iBRISK demonstrated high sensitivity in the malignancy prediction of BI-RADS 4 lesions. iBRISK may safely obviate biopsies in up to 50% of patients in low or moderate POM groups and reduce biopsy-associated costs.Keywords: Mammography, Breast, Oncology, Biopsy/Needle Aspiration, Radiomics, Precision Mammography, AI-augmented Biopsy Decision Support Tool, Breast Cancer Risk Calculator, BI-RADS 4 Mammography Risk Stratification, Overbiopsy Reduction, Probability of Malignancy (POM) Assessment, Biopsy-based Positive Predictive Value (PPV3) Supplemental material is available for this article. Published under a CC BY 4.0 license.See also the commentary by McDonald and Conant in this issue.

Tumor-Targeting Extracellular Vesicles Loaded with siS100A4 for Suppressing Postoperative Breast Cancer Metastasis.

Introduction: S100A4 promotes the establishment of tumor microenvironment for malignant cancer cells, and knockdown of S100A4 can inhibit tumorigenesis. However, there is no efficient way to target S100A4 in metastatic tumor tissues. Here, we investigated the role of siS100A4-loaded iRGD-modified extracellular vesicles (siS100A4-iRGD-EVs) in postoperative breast cancer metastasis. Methods: siS100A4-iRGD-EVs nanoparticles were engineered and analyzed using TEM and DLS. siRNA protection, cellular uptake, and cytotoxicity of EV nanoparticles were examined in vitro. Postoperative lung metastasis mouse model was created to investigate the tissue distribution and anti-metastasis roles of nanoparticles in vivo. Results: siS100A4-iRGD-EVs protected siRNA from RNase degradation, enhanced the cellular uptake and compatibility in vitro. Strikingly, iRGD-modified EVs significantly increased tumor organotropism and siRNA accumulation in lung PMNs compared to siS100A4-EVs in vivo. Moreover, siS100A4-iRGD-EVs treatment remarkedly attenuated lung metastases from breast cancer and increased survival rate of mice through suppressing S100A4 expression in lung. Conclusions: siS100A4-iRGD-EVs nanoparticles show more potent anti-metastasis effect in postoperative breast cancer metastasis mouse model. Supplementary Information: The online version contains supplementary material available at 10.1007/s12195-022-00757-5.

Controllable 3D Flower-Like Ag-CF Electrodes as Flexible Marine Electric Field Sensors with High Stability.

Construction of three-dimensional (3D) flower-like nanostructures with controlled morphologies has emerged as an attractive tool by scientists in the marine electric field sensor research field due to their peculiar structural features. Herein, novel 3D flower-like Ag-CF capacitive composite electrodes have been created by an eco-friendly water-bath strategy via AgNO3 as a sliver source and subsequently compounded with carbon fibers (CFs) pretreated by thermal oxidation. A series of electrode samples with various morphologies obtained by modulating different reaction times and temperatures bring about the dominant formation mechanism of these nanostructures and the influence behavior on the CF electrode in detail. Especially, the 3D flower-like Ag-CF electrode shows a large surface area acquired under the conditions of 80 °C and 15 min, which can provide more electroactive sites in electrochemical analysis and exhibit a maximum areal specific capacitance of 619.75 mF·cm-2 at a scanning speed of 10 mV·s-1. This is mainly due to the synergistic behavior of the unique 3D flower-like morphology and the large specific surface area of CFs. Furthermore, a cylinder-shaped Ag-CF sensor is designed, which delivers a superior potential difference of 33.08 µV, a potential difference drift of 18.62 µV/24 h for 30 days, and a self-noise of 0.92 nV/rt (Hz)@1 Hz. In this work, the intriguing synthesis strategy can be a promising facile approach to manufacture the controllable 3D flower-like Ag-CF electrode for electric field sensor applications.

Research on the Application of Artificial Intelligence in Public Health Management: Leveraging Artificial Intelligence to Improve COVID-19 CT Image Diagnosis.

Since the start of 2020, the outbreak of the Coronavirus disease (COVID-19) has been a global public health emergency, and it has caused unprecedented economic and social disaster. In order to improve the diagnosis efficiency of COVID-19 patients, a number of researchers have conducted extensive studies on applying artificial intelligence techniques to the analysis of COVID-19-related medical images. The automatic segmentation of lesions from computed tomography (CT) images using deep learning provides an important basis for the quantification and diagnosis of COVID-19 cases. For a deep learning-based CT diagnostic method, a few of accurate pixel-level labels are essential for the training process of a model. However, the translucent ground-glass area of the lesion usually leads to mislabeling while performing the manual labeling operation, which weakens the accuracy of the model. In this work, we propose a method for correcting rough labels; that is, to hierarchize these rough labels into precise ones by performing an analysis on the pixel distribution of the infected and normal areas in the lung. The proposed method corrects the incorrectly labeled pixels and enables the deep learning model to learn the infected degree of each infected pixel, with which an aiding system (named DLShelper) for COVID-19 CT image diagnosis using the hierarchical labels is also proposed. The DLShelper targets lesion segmentation from CT images, as well as the severity grading. The DLShelper assists medical staff in efficient diagnosis by providing rich auxiliary diagnostic information (including the severity grade, the proportions of the lesion and the visualization of the lesion area). A comprehensive experiment based on a public COVID-19 CT image dataset is also conducted, and the experimental results show that the DLShelper significantly improves the accuracy of segmentation for the lesion areas and also achieves a promising accuracy for the severity grading task.

Iron-Catalyzed, Directed Benzylic Borylation.

Herein, we describe an iron-catalyzed benzylic C-H borylation reaction. The reaction targets primary and secondary C(sp3)-H bonds to deliver high-value boronic esters under mild conditions with short (7-8 min) reaction times. Functional groups are well tolerated, and complete site selectivity is observed in the presence of multiple benzylic C-H bonds.

Bone Metastasis Initiation Is Coupled with Bone Remodeling through Osteogenic Differentiation of NG2+ Cells.

The bone microenvironment is dynamic and undergoes remodeling in normal and pathologic conditions. Whether such remodeling affects disseminated tumor cells (DTC) and bone metastasis remains poorly understood. Here, we demonstrated that pathologic fractures increase metastatic colonization around the injury. NG2+ cells are a common participant in bone metastasis initiation and bone remodeling in both homeostatic and fractured conditions. NG2+ bone mesenchymal stem/stromal cells (BMSC) often colocalize with DTCs in the perivascular niche. Both DTCs and NG2+ BMSCs are recruited to remodeling sites. Ablation of NG2+ lineage impaired bone remodeling and concurrently diminished metastatic colonization. In cocultures, NG2+ BMSCs, especially when undergoing osteodifferentiation, enhanced cancer cell proliferation and migration. Knockout of N-cadherin in NG2+ cells abolished these effects in vitro and phenocopied NG2+ lineage depletion in vivo. These findings uncover dual roles of NG2+ cells in metastasis and remodeling and indicate that osteodifferentiation of BMSCs promotes metastasis initiation via N-cadherin-mediated cell-cell interaction. SIGNIFICANCE: The bone colonization of cancer cells occurs in an environment that undergoes constant remodeling. Our study provides mechanistic insights into how bone homeostasis and pathologic repair lead to the outgrowth of disseminated cancer cells, thereby opening new directions for further etiologic and epidemiologic studies of tumor recurrences. This article is highlighted in the In This Issue feature, p. 247.

An emotion analysis in learning environment based on theme-specified drawing by convolutional neural network.

Emotion in the learning process can directly influence the learner's attention, memory, and cognitive activities. Several literatures indicate that hand-drawn painting could reflect the learner's emotional status. But, such an evaluation of emotional status, manually conducted by the psychologist, is usually subjective and inefficient for clinical practice. To address the issues of subjectivity and inefficiency in the painting based emotional analysis, we conducted an exploration of a painting based emotional analysis in learning environment by using convolutional neural network model. A painting image of 100 × 100 pixels was used as input for the model. The instant emotional statue of the learner was collected by filling out a questionnaire and was reviewed by a psychologist and then used as the label for training the convolutional neural network model. With the completion of convolutional, full-connected, and classification operations, the features of the painting image were learned from the underlying pixel matrix to the high-level semantic feature mapping. Then the emotional classification of the painting image could be made to reflect the learner's emotional status. Finally, the classification result by the model was compared with the result manually conducted by a psychologist to validate the model accuracy. We conducted an experiment in a university at Hangzhou, and 2,103 learners joined in the experiment. The learner was required to first fill out a questionnaire reporting emotional status in the learning process, and then to complete a theme-specified painting. Two thousand valid paintings were received and divided into training dataset (1,600) and test dataset (400). The experimental result indicated that the model achieved the accuracy of 72.1%, which confirmed the effectiveness of the model for emotional analysis.

Transfer learning enhanced water-enabled electricity generation in highly oriented graphene oxide nanochannels.

Harvesting energy from spontaneous water flow within artificial nanochannels is a promising route to meet sustainable power requirements of the fast-growing human society. However, large-scale nanochannel integration and the multi-parameter coupling restrictive influence on electric generation are still big challenges for macroscale applications. In this regard, long-range (1 to 20 cm) ordered graphene oxide assembled framework with integrated 2D nanochannels have been fabricated by a rotational freeze-casting method. The structure can promote spontaneous absorption and directional transmission of water inside the channels to generate considerable electric energy. A transfer learning strategy is implemented to address the complicated multi-parameters coupling problem under limited experimental data, which provides highly accurate performance optimization and efficiently guides the design of 2D water flow enabled generators. A generator unit can produce ~2.9 V voltage or ~16.8 µA current in a controllable manner. High electric output of ~12 V or ~83 µA is realized by connecting several devices in series or parallel. Different water enabled electricity generation systems have been developed to directly power commercial electronics like LED arrays and display screens, demonstrating the material's potential for development of water enabled clean energy.

An AI-driven clinical care pathway to reduce 30-day readmission for chronic obstructive pulmonary disease (COPD) patients.

Healthcare regulatory agencies have mandated a reduction in 30-day hospital readmission rates and have targeted COPD as a major contributor to 30-day readmissions. We aimed to develop and validate a simple tool deploying an artificial neural network (ANN) for early identification of COPD patients with high readmission risk. Using COPD patient data from eight hospitals within a large urban hospital system, four variables were identified, weighted and validated. These included the number of in-patient admissions in the previous 6 months, the number of medications administered on the first day, insurance status, and the Rothman Index on hospital day one. An ANN model was trained to provide a predictive algorithm and validated on an additional dataset from a separate time period. The model was implemented in a smartphone app (Re-Admit) incorporating four input risk factors, and a clinical care plan focused on high-risk readmission candidates was then implemented. Subsequent readmission data was analyzed to assess impact. The areas under the curve of receiver operating characteristics predicting readmission with ANN is 0.77, with sensitivity 0.75 and specificity 0.67 on the separate validation data. Readmission rates in the COPD high-risk subgroup after app and clinical intervention implementation saw a significant 48% decline. Our studies show the efficacy of ANN model on predicting readmission risks for COPD patients. The AI enabled Re-Admit smartphone app predicts readmission risk on day one of the patient's admission, allowing for early implementation of medical, hospital, and community resources to optimize and improve clinical care pathways.

A Machine-Learning-Enhanced Simultaneous and Multimodal Sensor Based on Moist-Electric Powered Graphene Oxide.

Simultaneous multimodal monitoring can greatly perceive intricately multiple stimuli, which is important for the understanding and development of a future human-machine fusion world. However, the integrated multisensor networks with cumbersome structure, huge power consumption, and complex preparation process have heavily restricted practical applications. Herein, a graphene oxide single-component multimodal sensor (GO-MS) is developed, which enables simultaneous monitoring of multiple environmental stimuli by a single unit with unique moist-electric self-power supply. This GO-MS can generate a sustainable moist-electric potential by spontaneously adsorbing water molecules in air, which has a characteristic response behavior when exposed to different stimuli. As a result, the simultaneous monitoring and decoupling of the changes of temperature, humidity, pressure, and light intensity are achieved by this single GO-MS with machine-learning (ML) assistance. Of practical importance, a moist-electric-powered human-machine interaction wristband based on GO-MS is constructed to monitor pulse signals, body temperature, and sweating in a multidimensional manner, as well as gestures and sign language commanding communication. This ML-empowered moist-electric GO-MS provides a new platform for the development of self-powered single-component multimodal sensors, showing great potential for applications in the fields of health detection, artificial electronic skin, and the Internet-of-Things.

A personalized image-guided intervention system for peripheral lung cancer on patient-specific respiratory motion model.

PURPOSE: Due to respiratory motion, precise tracking of lung nodule movement is a persistent challenge for guiding percutaneous lung biopsy during image-guided intervention. We developed an automated image-guided system incorporating effective and robust tracking algorithms to address this challenge. Accurate lung motion prediction and personalized image-guided intervention are the key technological contributions of this work. METHODS: A patient-specific respiratory motion model is developed to predict pulmonary movements of individual patients. It is based on the relation between the artificial 4D CT and corresponding positions tracked by position sensors attached on the chest using an electromagnetic (EM) tracking system. The 4D CT image of the thorax during breathing is calculated through deformable registration of two 3D CT scans acquired at inspiratory and expiratory breath-hold. The robustness and accuracy of the image-guided intervention system were assessed on a static thorax phantom under different clinical parametric combinations. RESULTS: Real 4D CT images of ten patients were used to evaluate the accuracy of the respiratory motion model. The mean error of the model in different breathing phases was 1.59 ± 0.66 mm. Using a static thorax phantom, we achieved an average targeting accuracy of 3.18 ± 1.2 mm across 50 independent tests with different intervention parameters. The positive results demonstrate the robustness and accuracy of our system for personalized lung cancer intervention. CONCLUSIONS: The proposed system integrates a patient-specific respiratory motion compensation model to reduce the effect of respiratory motion during percutaneous lung biopsy and help interventional radiologists target the lesion efficiently. Our preclinical studies indicate that the image-guided system has the ability to accurately predict and track lung nodules of individual patients and has the potential for use in the diagnosis and treatment of early stage lung cancer.

Image-to-image translation of label-free molecular vibrational images for a histopathological review using the UNet+/seg-cGAN model.

Translating images generated by label-free microscopy imaging, such as Coherent Anti-Stokes Raman Scattering (CARS), into more familiar clinical presentations of histopathological images will help the adoption of real-time, spectrally resolved label-free imaging in clinical diagnosis. Generative adversarial networks (GAN) have made great progress in image generation and translation, but have been criticized for lacking precision. In particular, GAN has often misinterpreted image information and identified incorrect content categories during image translation of microscopy scans. To alleviate this problem, we developed a new Pix2pix GAN model that simultaneously learns classifying contents in the images from a segmentation dataset during the image translation training. Our model integrates UNet+ with seg-cGAN, conditional generative adversarial networks with partial regularization of segmentation. Technical innovations of the UNet+/seg-cGAN model include: (1) replacing UNet with UNet+ as the Pix2pix cGAN's generator to enhance pattern extraction and richness of the gradient, and (2) applying the partial regularization strategy to train a part of the generator network as the segmentation sub-model on a separate segmentation dataset, thus enabling the model to identify correct content categories during image translation. The quality of histopathological-like images generated based on label-free CARS images has been improved significantly.

Moisture adsorption-desorption full cycle power generation.

Environment-adaptive power generation can play an important role in next-generation energy conversion. Herein, we propose a moisture adsorption-desorption power generator (MADG) based on porous ionizable assembly, which spontaneously adsorbs moisture at high RH and desorbs moisture at low RH, thus leading to cyclic electric output. A MADG unit can generate a high voltage of ~0.5 V and a current of 100 µA at 100% relative humidity (RH), delivers an electric output (~0.5 V and ~50 µA) at 15 ± 5% RH, and offers a maximum output power density approaching to 120 mW m-2. Such MADG devices could conduct enough power to illuminate a road lamp in outdoor application and directly drive electrochemical process. This work affords a closed-loop pathway for versatile moisture-based energy conversion.

A Time-Series Feature-Based Recursive Classification Model to Optimize Treatment Strategies for Improving Outcomes and Resource Allocations of COVID-19 Patients.

This paper presents a novel Lasso Logistic Regression model based on feature-based time series data to determine disease severity and when to administer drugs or escalate intervention procedures in patients with coronavirus disease 2019 (COVID-19). Advanced features were extracted from highly enriched and time series vital sign data of hospitalized COVID-19 patients, including oxygen saturation readings, and with a combination of patient demographic and comorbidity information, as inputs into the dynamic feature-based classification model. Such dynamic combinations brought deep insights to guide clinical decision-making of complex COVID-19 cases, including prognosis prediction, timing of drug administration, admission to intensive care units, and application of intervention procedures like ventilation and intubation. The COVID-19 patient classification model was developed utilizing 900 hospitalized COVID-19 patients in a leading multi-hospital system in Texas, United States. By providing mortality prediction based on time-series physiologic data, demographics, and clinical records of individual COVID-19 patients, the dynamic feature-based classification model can be used to improve efficacy of the COVID-19 patient treatment, prioritize medical resources, and reduce casualties. The uniqueness of our model is that it is based on just the first 24 hours of vital sign data such that clinical interventions can be decided early and applied effectively. Such a strategy could be extended to prioritize resource allocations and drug treatment for futurepandemic events.

Artificial intelligence-augmented, label-free molecular imaging method for tissue identification, cancer diagnosis, and cancer margin detection.

Label-free high-resolution molecular and cellular imaging strategies for intraoperative use are much needed, but not yet available. To fill this void, we developed an artificial intelligence-augmented molecular vibrational imaging method that integrates label-free and subcellular-resolution coherent anti-stokes Raman scattering (CARS) imaging with real-time quantitative image analysis via deep learning (artificial intelligence-augmented CARS or iCARS). The aim of this study was to evaluate the capability of the iCARS system to identify and differentiate the parathyroid gland and recurrent laryngeal nerve (RLN) from surrounding tissues and detect cancer margins. This goal was successfully met.

The effect of platelet G proteins on platelet extravasation and tumor growth in the murine model of ovarian cancer.

We and other investigators have shown that platelets promote metastasis and the growth of tumors. Our rationale for conducting this study is that platelets' prometastatic and progrowth effects depend on a close encounter between platelets and cancer cells. This interaction occurs inside blood vessels with circulating tumor cells and outside blood vessels with cancer cells residing in the tumor parenchyma. Our hypothesis was that platelet extravasation is required for the effect of platelets on tumor growth. Platelets respond to environmental stimuli by activation of G protein-coupled receptors on their surface. We investigated the impact of various platelet G proteins on the growth of ovarian cancer tumors and platelet extravasation. We used mice with platelet-specific deficiency of Gαi2 (Gi), Gα13 (G13), or Gαq (Gq) in a syngeneic ovarian cancer model. We measured the total weight of tumor nodules resected from tumor-bearing mice. We developed methods for automated whole-slide image acquisition and unbiased computerized image analysis to quantify extravasated platelets. We compared the number of platelets inside tumor nodules of platelet G protein-deficient tumor-bearing mice. We found that deficiency of Gi and G13, but not Gq, in platelets resulted in smaller tumors compared with those in corresponding littermates. Deficiency of Gi and G13 in platelets reduced the number of extravasated platelets by >90%, but deficiency of Gq did not reduce the number of extravasated platelets significantly. The lack of Gi or G13 in platelets reduced platelet extravasation into the tumor and tumor growth.

Bilayer of polyelectrolyte films for spontaneous power generation in air up to an integrated 1,000 V output.

Environmentally adaptive power generation is attractive for the development of next-generation energy sources. Here we develop a heterogeneous moisture-enabled electric generator (HMEG) based on a bilayer of polyelectrolyte films. Through the spontaneous adsorption of water molecules in air and induced diffusion of oppositely charged ions, one single HMEG unit can produce a high voltage of ~0.95 V at low (25%) relative humidity (RH), and even jump to 1.38 V at 85% RH. A sequentially aligned stacking strategy is created for large-scale integration of HMEG units, to offer a voltage of more than 1,000 V under ambient conditions (25% RH, 25 °C). Using origami assembly, a small section of folded HMEGs renders an output of up to 43 V cm-3. Such integration devices supply sufficient power to illuminate a lamp bulb of 10 W, to drive a dynamic electronic ink screen and to control the gate voltage for a self-powered field effect transistor.

An imageomics and multi-network based deep learning model for risk assessment of liver transplantation for hepatocellular cancer.

INTRODUCTION: Liver transplantation (LT) is an effective treatment for hepatocellular carcinoma (HCC), the most common type of primary liver cancer. Patients with small HCC (<5 cm) are given priority over others for transplantation due to clinical allocation policies based on tumor size. Attempting to shift from the prevalent paradigm that successful transplantation and longer disease-free survival can only be achieved in patients with small HCC to expanding the transplantation option to patients with HCC of the highest tumor burden (>5 cm), we developed a convergent artificial intelligence (AI) model that combines transient clinical data with quantitative histologic and radiomic features for more objective risk assessment of liver transplantation for HCC patients. METHODS: Patients who received a LT for HCC between 2008-2019 were eligible for inclusion in the analysis. All patients with post-LT recurrence were included, and those without recurrence were randomly selected for inclusion in the deep learning model. Pre- and post-transplant magnetic resonance imaging (MRI) scans and reports were compressed using CapsNet networks and natural language processing, respectively, as input for a multiple feature radial basis function network. We applied a histological image analysis algorithm to detect pathologic areas of interest from explant tissue of patients who recurred. The multilayer perceptron was designed as a feed-forward, supervised neural network topology, with the final assessment of recurrence risk. We used area under the curve (AUC) and F-1 score to assess the predictability of different network combinations. RESULTS: A total of 109 patients were included (87 in the training group, 22 in the testing group), of which 20 were positive for cancer recurrence. Seven models (AUC; F-1 score) were generated, including clinical features only (0.55; 0.52), magnetic resonance imaging (MRI) only (0.64; 0.61), pathological images only (0.64; 0.61), MRI plus pathology (0.68; 0.65), MRI plus clinical (0.78, 0.75), pathology plus clinical (0.77; 0.73), and a combination of clinical, MRI, and pathology features (0.87; 0.84). The final combined model showed 80 % recall and 89 % precision. The total accuracy of the implemented model was 82 %. CONCLUSION: We validated that the deep learning model combining clinical features and multi-scale histopathologic and radiomic image features can be used to discover risk factors for recurrence beyond tumor size and biomarker analysis. Such a predictive, convergent AI model has the potential to alter the LT allocation system for HCC patients and expand the transplantation treatment option to patients with HCC of the highest tumor burden.