Controlling bacterial growth and inactivation using thin film-based surface acoustic waves.

Formation of bacterial films on structural surfaces often leads to severe contamination of medical devices, hospital equipment, implant materials, etc., and antimicrobial resistance of microorganisms has indeed become a global health issue. Therefore, effective therapies for controlling infectious and pathogenic bacteria are urgently needed. Being a promising active method for this purpose, surface acoustic waves (SAWs) have merits such as nanoscale earthquake-like vibration/agitation/radiation, acoustic streaming induced circulations, and localised acoustic heating effect in liquids. However, only a few studies have explored controlling bacterial growth and inactivation behaviour using SAWs. In this study, we proposed utilising piezoelectric thin film-based SAW devices on a silicon substrate for controlling bacterial growth and inactivation with and without using ZnO micro/nanostructures. Effects of SAW powers on bacterial growth for two types of bacteria, i.e., E. coli and S. aureus, were evaluated. Varied concentrations of ZnO tetrapods were also added into the bacterial culture to study their effects and the combined antimicrobial effects along with SAW agitation. Our results showed that when the SAW power was below a threshold (e.g., about 2.55 W in this study), the bacterial growth was apparently enhanced, whereas the further increase of SAW power to a high power caused inactivation of bacteria. Combination of thin film SAWs with ZnO tetrapods led to significantly decreased growth or inactivation for both E. coli and S. aureus, revealing their effectiveness for antimicrobial treatment. Mechanisms and effects of SAW interactions with bacterial solutions and ZnO tetrapods have been systematically discussed.

Flexible surface acoustic wave technology for enhancing transdermal drug delivery.

Transdermal drug delivery provides therapeutic benefits over enteric or injection delivery because its transdermal routes provide more consistent concentrations of drug and avoid issues of drugs affecting kidneys and liver functions. Many technologies have been evaluated to enhance drug delivery through the relatively impervious epidermal layer of the skin. However, precise delivery of large hydrophilic molecules is still a great challenge even though microneedles or other energized (such as electrical, thermal, or ultrasonic) patches have been used, which are often difficult to be integrated into small wearable devices. This study developed a flexible surface acoustic wave (SAW) patch platform to facilitate transdermal delivery of macromolecules with fluorescein isothiocyanates up to 2000 kDa. Two surrogates of human skin were used to evaluate SAW based energized devices, i.e., delivering dextran through agarose gels and across stratum corneum of pig skin into the epidermis. Results showed that the 2000 kDa fluorescent molecules have been delivered up to 1.1 mm in agarose gel, and the fluorescent molecules from 4 to 2000 kDa have been delivered up to 100 µm and 25 µm in porcine skin tissue, respectively. Mechanical agitation, localised streaming, and acousto-thermal effect generated on the skin surface were identified as the main mechanisms for promoting drug transdermal transportation, although micro/nanoscale acoustic cavitation induced by SAWs could also have its contribution. SAW enhanced transdermal drug delivery is dependent on the combined effects of wave frequency and intensity, duration of applied acoustic waves, temperature, and drug molecules molecular weights.

Surface Acoustic Wave Sensors for Wireless Temperature Measurements above 1200 Degree Celsius.

High-temperature wireless sensing is crucial for monitoring combustion chambers and turbine stators in aeroengines, where surface temperatures can reach up to 1200 °C. Surface Acoustic Wave (SAW) temperature sensors are an excellent choice for these measurements. However, at extreme temperatures, they face issues such as agglomeration and recrystallization of electrodes, leading to loss of conductivity and reduced quality factor, hindering effective wireless signal transmission. This study develops an LGS SAW sensor with a Pt-10%Rh/Zr/Pt-10%Rh/Zr/Pt-10%Rh/Zr multilayer composite electrode structure to address these challenges. We demonstrate that the sensor can achieve wireless temperature measurements from room temperature to 1200 °C with an accuracy of 1.59%. The composite electrodes excite a quasi-shear wave on the LGS substrate, maintaining a Q-factor of 3526 at room temperature, providing an initial assurance for the strength of the wireless interrogation echo signal. The sensor operates stably for 2.18 h at 1200 °C before adhesion loss between the composite electrode and the substrate causes a sudden increase in resonant frequency. This study highlights the durability of the proposed electrode materials and structure at extreme temperatures and suggests future research to improve adhesion and extend the sensor's lifespan, thereby enhancing the reliability and effectiveness of high-temperature wireless sensing in aerospace applications.

Formation processes and mechanisms of a fault-controlled colluvial landslide in the Qinling-Daba Mountains, China.

Faults play a crucial role in shaping the formation and damage patterns of landslides in the mountainous region, particularly in Qinling-Daba (Qinba) area in China. On 6 October 2022, following a 4-day rainfall event totaling 221.5 mm, a landslide occurred in Hanwang Town, Shaanxi Province. The left boundary of the landslide coincided with a fault, which influence the formation and movement development of the landslide. To further understand and quantified the formation process and damage mechanism of the landslide, a comprehensive study was conducted, incorporating field investigations, local rainfall data, and various methods including unmanned aerial vehicles (UAVs), numerical simulations, and laboratory test. The results indicate that fault dictate the formation of the Lijiaping landslide by influencing the mechanical strength of the rock mass and the catchment topography in the landslide area. Due to fault, the rock mass in the landslide area is high fragmentation, with a softening coefficient of about 0.52. Weathering resulted in numerous residual and slope sediments in the landslide area, providing ample material for the landslide. Meanwhile, the fault activity led to a wedge-shaped topography in the landslide area, with an average Terrain Wetness Index (TWI) of 3.43, significantly higher than the Hanwang Township average of 1.47. This creates a hydrogeological structure favorable for landslides. Numerical simulations revealed that the maximum velocity of the landslide reached 5.05 m/s and the maximum displacement was 53.18 m, both occurring in the central part of the landslide. These findings offer crucial scientific insights for understanding and preventing similar geological hazards.

TRIM35 Negatively Regulates the cGAS-STING-Mediated Signaling Pathway by Attenuating K63-Linked Ubiquitination of STING.

The cGAS-STING-mediated antiviral response plays an important role in the defense against DNA virus infection. Tripartite motif protein 35 (TRIM35), an E3 ubiquitin ligase, was identified as a positive regulator of RLR-mediated antiviral signaling in our previous study, but the effect of TRIM35 on the cGAS-STING signaling pathway has not been elucidated. Herein, we showed that TRIM35 negatively regulates the cGAS-STING signaling pathway by directly targeting STING. TRIM35 overexpression significantly inhibited the cGAMP-triggered phosphorylation of TBK1 and IRF3, attenuating IFN-ß expression and the downstream antiviral response. Mechanistically, TRIM35 colocalized and directly interacted with STING in the cytoplasm. TRM35 removed K63-linked ubiquitin from STING through the C36 and C44 sites in the RING domain, which impaired the interaction of STING with TBK1 or IKKε. In addition, we demonstrated that the RING domain is a key region for the antiviral effects of TIRM35. These results collectively indicate that TRIM35 negatively regulates type I interferon (IFN-I) production by targeting and deubiquitinating STING. TRIM35 may be a potential therapeutic target for controlling viral infection.

Rational Design of a Surface Acoustic Wave Device for Wearable Body Temperature Monitoring.

Continuous monitoring of vital signs based on advanced sensing technologies has attracted extensive attention due to the ravages of COVID-19. A maintenance-free and low-cost passive wireless sensing system based on surface acoustic wave (SAW) device can be used to continuously monitor temperature. However, the current SAW-based passive sensing system is mostly designed at a low frequency around 433 MHz, which leads to the relatively large size of SAW devices and antenna, hindering their application in wearable devices. In this paper, SAW devices with a resonant frequency distributed in the 870 MHz to 960 MHz range are rationally designed and fabricated. Based on the finite-element method (FEM) and coupling-of-modes (COM) model, the device parameters, including interdigital transducer (IDT) pairs, aperture size, and reflector pairs, are systematically optimized, and the theoretical and experimental results show high consistency. Finally, SAW temperature sensors with a quality factor greater than 2200 are obtained for real-time temperature monitoring ranging from 20 to 50 °C. Benefitting from the higher operating frequency, the size of the sensing system can be reduced for human body temperature monitoring, showing its potential to be used as a wearable monitoring device in the future.

Moisture content online detection system based on multi-sensor fusion and convolutional neural network.

To monitor the moisture content of agricultural products in the drying process in real time, this study applied a model combining multi-sensor fusion and convolutional neural network (CNN) to moisture content online detection. This study built a multi-sensor data acquisition platform and established a CNN prediction model with the raw monitoring data of load sensor, air velocity sensor, temperature sensor, and the tray position as input and the weight of the material as output. The model's predictive performance was compared with that of the linear partial least squares regression (PLSR) and nonlinear support vector machine (SVM) models. A moisture content online detection system was established based on this model. Results of the model performance comparison showed that the CNN prediction model had the optimal prediction effect, with the determination coefficient (R2) and root mean square error (RMSE) of 0.9989 and 6.9, respectively, which were significantly better than those of the other two models. Results of validation experiments showed that the detection system met the requirements of moisture content online detection in the drying process of agricultural products. The R2 and RMSE were 0.9901 and 1.47, respectively, indicating the good performance of the model combining multi-sensor fusion and CNN in moisture content online detection for agricultural products in the drying process. The moisture content online detection system established in this study is of great significance for researching new drying processes and realizing the intelligent development of drying equipment. It also provides a reference for online detection of other indexes in the drying process of agricultural products.

A phase 3 randomized, open-label study evaluating the immunogenicity and safety of concomitant and staggered administration of a live, pentavalent rotavirus vaccine and an inactivated poliomyelitis vaccine in healthy infants in China.

This open-label, randomized, phase 3 study in China (V260-074; NCT04481191) evaluated the immunogenicity and safety of concomitant and staggered administration of three doses of an oral, live, pentavalent rotavirus vaccine (RV5) and three doses of an intramuscular, inactivated poliomyelitis vaccine (IPV) in 400 healthy infants. The primary objective was the non-inferiority of neutralizing antibody (nAb) responses in the concomitant- versus the staggered-use groups. Antibody responses were measured at baseline and 1-month post-dose 3 (PD3). Parents/legal guardians recorded adverse events for 30 or 15 d after study vaccinations in the concomitant-use or staggered-use groups, respectively. At PD3, >98% of participants seroconverted to all three poliovirus types, and the primary objective was met as lower bounds of the two-sided 95% CI for between-group difference in nAb seroconversion percentages ranged from - 4.3% to - 1.6%, for all poliovirus types, p < .001. At PD3, geometric mean titers (GMTs) of nAb responses to poliovirus types 1, 2, and 3 in the concomitant-use group and the staggered-use group were comparable; 100% of participants had nAb titers ≥1:8 and ≥1:64 for all poliovirus types. Anti-rotavirus serotype-specific IgA GMTs and participants with ≥3-fold rise in postvaccination titers from baseline were comparable between groups. Administration of RV5 and IPV was well tolerated with comparable safety profiles in both groups. The immunogenicity of IPV in the concomitant-use group was non-inferior to the staggered-use group and RV5 was immunogenic in both groups. No safety concerns were identified. These data support the concomitant use of RV5 and IPV in healthy Chinese infants.

Quality and Process Optimization of Infrared Combined Hot Air Drying of Yam Slices Based on BP Neural Network and Gray Wolf Algorithm.

In this paper, the effects on drying time (Y1), the color difference (Y2), unit energy consumption (Y3), polysaccharide content (Y4), rehydration ratio (Y5), and allantoin content (Y6) of yam slices were investigated under different drying temperatures (50-70 °C), slice thicknesses (2-10 mm), and radiation distances (80-160 mm). The optimal drying conditions were determined by applying the BP neural network wolf algorithm (GWO) model based on response surface methodology (RMS). All the above indices were significantly affected by drying conditions (p < 0.05). The drying rate and effective water diffusion coefficient of yam slices accelerated with increasing temperature and decreasing slice thickness and radiation distance. The selection of lower temperature and slice thickness helped reduce the energy consumption and color difference. The polysaccharide content increased and then decreased with drying temperature, slice thickness, and radiation distance, and it was highest at 60 °C, 6 mm, and 120 mm. At 60 °C, lower slice thickness and radiation distance favored the retention of allantoin content. Under the given constraints (minimization of drying time, unit energy consumption, color difference, and maximization of rehydration ratio, polysaccharide content, and allantoin content), BP-GWO was found to have higher coefficients of determination (R2 = 0.9919 to 0.9983) and lower RMSEs (reduced by 61.34% to 80.03%) than RMS. Multi-objective optimization of BP-GWO was carried out to obtain the optimal drying conditions, as follows: temperature 63.57 °C, slice thickness 4.27 mm, radiation distance 91.39 mm, corresponding to the optimal indices, as follows: Y1 = 133.71 min, Y2 = 7.26, Y3 = 8.54 kJ·h·kg-1, Y4 = 20.73 mg/g, Y5 = 2.84 kg/kg, and Y6 = 3.69 µg/g. In the experimental verification of the prediction results, the relative error between the actual and predicted values was less than 5%, proving the model's reliability for other materials in the drying technology process research to provide a reference.

miR-26a exerts broad-spectrum antiviral effects via the enhancement of RIG-I-mediated type I interferon response by targeting USP15.

IMPORTANCE: miR-26a serves as a potent positive regulator of type I interferon (IFN) responses. By inhibiting USP15 expression, miR-26a promotes RIG-I K63-ubiquitination to enhance type I IFN responses, resulting in an active antiviral state against viruses. Being an intricate regulatory network, the activation of type I IFN responses could in turn suppress miR-26a expression to avoid the disordered activation that might result in the so-called "type I interferonopathy." The knowledge gained would be essential for the development of novel antiviral strategies against viral infection.

Drug repurposing screen identifies vidofludimus calcium and pyrazofurin as novel chemical entities for the development of hepatitis E interventions.

Hepatitis E virus (HEV) infection can cause severe complications and high mortality, particularly in pregnant women, organ transplant recipients, individuals with pre-existing liver disease and immunosuppressed patients. However, there are still unmet needs for treating chronic HEV infections. Herein, we screened a best-in-class drug repurposing library consisting of 262 drugs/compounds. Upon screening, we identified vidofludimus calcium and pyrazofurin as novel anti-HEV entities. Vidofludimus calcium is the next-generation dihydroorotate dehydrogenase (DHODH) inhibitor in the phase 3 pipeline to treat autoimmune diseases or SARS-CoV-2 infection. Pyrazofurin selectively targets uridine monophosphate synthetase (UMPS). Their anti-HEV effects were further investigated in a range of cell culture models and human liver organoids models with wild type HEV strains and ribavirin treatment failure-associated HEV strains. Encouragingly, both drugs exhibited a sizeable therapeutic window against HEV. For instance, the IC50 value of vidofludimus calcium is 4.6-7.6-fold lower than the current therapeutic doses in patients. Mechanistically, their anti-HEV mode of action depends on the blockage of pyrimidine synthesis. Notably, two drugs robustly inhibited ribavirin treatment failure-associated HEV mutants (Y1320H, G1634R). Their combination with IFN-α resulted in synergistic antiviral activity. In conclusion, we identified vidofludimus calcium and pyrazofurin as potent candidates for the treatment of HEV infections. Based on their antiviral potency, and also the favorable safety profile identified in clinical studies, our study supports the initiation of clinical studies to repurpose these drugs for treating chronic hepatitis E.

Immune response and safety of inactivated SARS-CoV-2 vaccines during pregnancy: a real-world observational study.

BACKGROUND: As pregnant women are excluded from clinical trials of inactivated SARS-CoV-2 vaccines, it is important to assess the immune response in women receiving the vaccination while unknowingly pregnant. METHODS: In a multicenter cross-sectional study, we enrolled 873 pregnant women aged 18-45 years. Serum antibody levels induced by inactivated vaccines were determined. Adverse events were collected by self-reported survey after vaccination. Logistic regression model and restricted cubic spline model were used to investigate the association of factors with antibody positivity. RESULTS: As the doses of the vaccine increase, neutralizing antibody (NAb) positivity was 98.3%, 39.5%, and 9.5% in pregnant women, respectively. The dose of vaccine and duration since vaccination were associated with NAb positivity. The OR of two and three doses of vaccines were 7.20 and 458.33 (P < 0.05). NAb levels and duration since vaccination showed a linear relationship in pregnant women vaccinated two doses, with a decrease to a near seropositivity threshold at 22 weeks. Adverse events were mainly mild or moderate after vaccinated during pregnancy, with no increase in incidence compared with whom vaccinated during pre-pregnancy. CONCLUSIONS: The use of inactivated vaccines during pregnancy induced favorable immune persistence, and the incidence of adverse events did not increase.

Effect of Combined Infrared Hot Air Drying on Yam Slices: Drying Kinetics, Energy Consumption, Microstructure, and Nutrient Composition.

Using hot air drying (HAD) and combined infrared hot air drying (IR-HAD) test devices, the drying kinetics, unit energy consumption, color difference values, rehydration rate, microstructure, and changes in polysaccharide and allantoin contents of yam slices were examined at various temperatures (50 °C, 55 °C, 60 °C, 65 °C, and 70 °C). The findings demonstrated that each of the aforementioned parameters was significantly impacted by the drying temperature. IR-HAD dries quicker and takes less time to dry than HAD. The Deff of IR-HAD is higher than that of HAD at the same temperature and increases with the increase in temperature. The activation energy required for IR-HAD (26.35 kJ/mol) is lower than that required for HAD (32.53 kJ/mol). HAD uses more energy per unit than IR-HAD by a factor of greater than 1.3. Yam slices treated with IR-HAD had higher microscopic porosity, better rehydration, lower color difference values, and higher polysaccharide and allantoin levels than HAD-treated yam slices. The IR-HAD at 60 °C had the greatest comprehensive rating after a thorough analysis of the dried yam slices using the coefficient of variation method. Three statistical indicators were used to evaluate six thin-layer drying models, and the Weibull model was most applicable to describe the variation of drying characteristics of yam slices.

Improving Image Classification of Knee Radiographs: An Automated Image Labeling Approach.

Large numbers of radiographic images are available in musculoskeletal radiology practices which could be used for training of deep learning models for diagnosis of knee abnormalities. However, those images do not typically contain readily available labels due to limitations of human annotations. The purpose of our study was to develop an automated labeling approach that improves the image classification model to distinguish normal knee images from those with abnormalities or prior arthroplasty. The automated labeler was trained on a small set of labeled data to automatically label a much larger set of unlabeled data, further improving the image classification performance for knee radiographic diagnosis. We used BioBERT and EfficientNet as the feature extraction backbone of the labeler and imaging model, respectively. We developed our approach using 7382 patients and validated it on a separate set of 637 patients. The final image classification model, trained using both manually labeled and pseudo-labeled data, had the higher weighted average AUC (WA-AUC 0.903) value and higher AUC values among all classes (normal AUC 0.894; abnormal AUC 0.896, arthroplasty AUC 0.990) compared to the baseline model (WA-AUC = 0.857; normal AUC 0.842; abnormal AUC 0.848, arthroplasty AUC 0.987), trained using only manually labeled data. Statistical tests show that the improvement is significant on normal (p value < 0.002), abnormal (p value < 0.001), and WA-AUC (p value = 0.001). Our findings demonstrated that the proposed automated labeling approach significantly improves the performance of image classification for radiographic knee diagnosis, allowing for facilitating patient care and curation of large knee datasets.

Inhibition of human adenovirus replication by TRIM35-mediated degradation of E1A.

Human adenovirus (HAdV) is ubiquitous in the human population, constituting a significant burden of global respiratory diseases. Children and individuals with low immunity are at risk of developing severe infections without approved antiviral treatment for HAdV. Our study demonstrated that TRIM35 inhibited HAdV-C5 early gene transcription, early protein expression, genome replication, and infectious virus progeny production. Furthermore, TRIM35 was found to inhibit HAdV replication by attenuating E1A expression. Mechanistically, TRIM35 interacts with and degrades E1A by promoting its K48-linked ubiquitination. Additionally, K253 and K285 are the key sites necessary for TRIM35 degradation. Moreover, an oncolytic adenovirus carrying shTRIM35 was constructed and observed to exhibit improved oncolysis in vivo, providing new ideas for clinical tumor treatment. Our results expand the broad antiviral activity of TRIM35 and mechanically support its application as a HAdV replication inhibitor. IMPORTANCE E1A is an essential human adenovirus (HAdV) protein responsible for the early replication of adenovirus while interacting with multiple host proteins. Understanding the interaction between HAdV E1A and TRIM35 helps identify effective antiviral therapeutic targets. The viral E1A protein is a crucial activator and regulator of viral transcription during the early infection stages. We first reported that TRIM35 interacts with E1A to resist adenovirus infection. Our study demonstrated that TRIM35 targets E1A to resist adenovirus, indicating the applicability of targeting virus-dependent host factors as a suitable antiviral strategy.

Feasibility of predicting a screening digital breast tomosynthesis recall using features extracted from the electronic medical record.

PURPOSE: Tools to predict a screening mammogram recall at the time of scheduling could improve patient care. We extracted patient demographic and breast care history information within the electronic medical record (EMR) for women undergoing digital breast tomosynthesis (DBT) to identify which factors were associated with a screening recall recommendation. METHOD: In 2018, 21,543 women aged 40 years or greater who underwent screening DBT at our institution were identified. Demographic information and breast care factors were extracted automatically from the EMR. The primary outcome was a screening recall recommendation of BI-RADS 0. A multivariable logistic regression model was built and included age, race, ethnicity groups, family breast cancer history, personal breast cancer history, surgical breast cancer history, recall history, and days since last available screening mammogram. RESULTS: Multiple factors were associated with a recall on the multivariable model: history of breast cancer surgery (OR: 2.298, 95% CI: 1.854, 2.836); prior recall within the last five years (vs no prior, OR: 0.768, 95% CI: 0.687, 0.858); prior screening mammogram within 0-18 months (vs no prior, OR: 0.601, 95% CI: 0.520, 0.691), prior screening mammogram within 18-30 months (vs no prior, OR: 0.676, 95% CI: 0.520, 0.691); and age (normalized OR: 0.723, 95% CI: 0.690, 0.758). CONCLUSIONS: It is feasible to predict a DBT screening recall recommendation using patient demographics and breast care factors that can be extracted automatically from the EMR.

Drying Temperature Precision Control System Based on Improved Neural Network PID Controller and Variable-Temperature Drying Experiment of Cantaloupe Slices.

A drying temperature precision control system was studied to provide technical support for developing and further proving the superiority of the variable-temperature drying process. In this study, an improved neural network (INN) proportional-integral-derivative (PID) controller (INN-PID) was designed. The dynamic performance of the PID, neural network PID (NN-PID) and INN-PID controllers was simulated with unit step signals as an input in MATLAB software. A drying temperature precision control system was set up in an air impingement dryer, and the drying temperature control experiment was carried out to verify the performance of the three controllers. Linear variable-temperature (LVT) and constant-temperature drying experiments of cantaloupe slices were carried out based on the system. Moreover, the experimental results were evaluated comprehensively with the brightness (L value), colour difference (ΔE), vitamin C content, chewiness, drying time and energy consumption (EC) as evaluation indexes. The simulation results show that the INN-PID controller outperforms the other two controllers in terms of control accuracy and regulation time. In the drying temperature control experiment at 50 °C-55 °C, the peak time of the INN-PID controller is 237.37 s, the regulation time is 134.91 s and the maximum overshoot is 4.74%. The INN-PID controller can quickly and effectively regulate the temperature of the inner chamber of the air impingement dryer. Compared with constant-temperature drying, LVT is a more effective drying mode as it ensures the quality of the material and reduces the drying time and EC. The drying temperature precision control system based on the INN-PID controller meets the temperature control requirements of the variable-temperature drying process. This system provides practical and effective technical support for the variable-temperature drying process and lays the foundation for further research. The LVT drying experiments of cantaloupe slices also show that variable-temperature drying is a better process than constant-temperature drying and is worthy of further study to be applied in production.

Multistep Automated Data Labelling Procedure (MADLaP) for thyroid nodules on ultrasound: An artificial intelligence approach for automating image annotation.

Machine learning (ML) for diagnosis of thyroid nodules on ultrasound is an active area of research. However, ML tools require large, well-labeled datasets, the curation of which is time-consuming and labor-intensive. The purpose of our study was to develop and test a deep-learning-based tool to facilitate and automate the data annotation process for thyroid nodules; we named our tool Multistep Automated Data Labelling Procedure (MADLaP). MADLaP was designed to take multiple inputs including pathology reports, ultrasound images, and radiology reports. Using multiple step-wise 'modules' including rule-based natural language processing, deep-learning-based imaging segmentation, and optical character recognition, MADLaP automatically identified images of a specific thyroid nodule and correctly assigned a pathology label. The model was developed using a training set of 378 patients across our health system and tested on a separate set of 93 patients. Ground truths for both sets were selected by an experienced radiologist. Performance metrics including yield (how many labeled images the model produced) and accuracy (percentage correct) were measured using the test set. MADLaP achieved a yield of 63 % and an accuracy of 83 %. The yield progressively increased as the input data moved through each module, while accuracy peaked part way through. Error analysis showed that inputs from certain examination sites had lower accuracy (40 %) than the other sites (90 %, 100 %). MADLaP successfully created curated datasets of labeled ultrasound images of thyroid nodules. While accurate, the relatively suboptimal yield of MADLaP exposed some challenges when trying to automatically label radiology images from heterogeneous sources. The complex task of image curation and annotation could be automated, allowing for enrichment of larger datasets for use in machine learning development.

Artificial Neural Network Modeling and Genetic Algorithm Multiobjective Optimization of Process of Drying-Assisted Walnut Breaking.

This study combined an artificial neural network (ANN) with a genetic algorithm (GA) to obtain the model and optimal process parameters of drying-assisted walnut breaking. Walnuts were dried at different IR temperatures (40 °C, 45 °C, 50 °C, and 55 °C) and air velocities (1, 2, 3, and 4 m/s) to different moisture contents (10%, 15%, 20%, and 25%) by using air-impingement technology. Subsequently, the dried walnuts were broken in different loading directions (sutural, longitudinal, and vertical). The drying time (DT), specific energy consumption (SEC), high kernel rate (HR), whole kernel rate (WR), and shell-breaking rate (SR) were determined as response variables. An ANN optimized by a GA was applied to simulate the influence of IR temperature, air velocity, moisture content, and loading direction on the five response variables, from which the objective functions of DT, SEC, HR, WR, and SR were developed. A GA was applied for the simultaneous maximization of HR, WR, and SR and minimization of DT and SEC to determine the optimized process parameters. The ANN model had a satisfactory prediction ability, with the coefficients of determination of 0.996, 0.998, 0.990, 0.991, and 0.993 for DT, SEC, HR, WR, and SR, respectively. The optimized process parameters were found to be 54.9 °C of IR temperature, 3.66 m/s of air velocity, 10.9% of moisture content, and vertical loading direction. The model combining an ANN and a GA was proven to be an effective method for predicting and optimizing the process parameters of walnut breaking. The predicted values under optimized process parameters fitted the experimental data well, with a low relative error value of 2.51-3.96%. This study can help improve the quality of walnut breaking, processing efficiency, and energy conservation. The ANN modeling and GA multiobjective optimization method developed in this study provide references for the process optimization of walnut and other similar commodities.

Deep learning for classification of thyroid nodules on ultrasound: validation on an independent dataset.

OBJECTIVES: The purpose is to apply a previously validated deep learning algorithm to a new thyroid nodule ultrasound image dataset and compare its performances with radiologists. METHODS: Prior study presented an algorithm which is able to detect thyroid nodules and then make malignancy classifications with two ultrasound images. A multi-task deep convolutional neural network was trained from 1278 nodules and originally tested with 99 separate nodules. The results were comparable with that of radiologists. The algorithm was further tested with 378 nodules imaged with ultrasound machines from different manufacturers and product types than the training cases. Four experienced radiologists were requested to evaluate the nodules for comparison with deep learning. RESULTS: The Area Under Curve (AUC) of the deep learning algorithm and four radiologists were calculated with parametric, binormal estimation. For the deep learning algorithm, the AUC was 0.69 (95% CI: 0.64-0.75). The AUC of radiologists were 0.63 (95% CI: 0.59-0.67), 0.66 (95% CI:0.61-0.71), 0.65 (95% CI: 0.60-0.70), and 0.63 (95%CI: 0.58-0.67). CONCLUSION: In the new testing dataset, the deep learning algorithm achieved similar performances with all four radiologists. The relative performance difference between the algorithm and the radiologists is not significantly affected by the difference of ultrasound scanner.