Convolutional Neural Network Predictions Research Articles

Detecting Racial/Ethnic Health Disparities Using Deep Learning From Frontal Chest Radiography.

The aim of this study was to assess racial/ethnic and socioeconomic disparities in the difference between atherosclerotic vascular disease prevalence measured by a multitask convolutional neural network (CNN) deep learning model using frontal chest radiographs (CXRs) and the prevalence reflected by administrative hierarchical condition category codes in two cohorts of patients with coronavirus disease2019 (COVID-19). A CNN model, previously published, was trained to predict atherosclerotic disease from ambulatory frontal CXRs. The model was then validated on two cohorts of patients with COVID-19: 814 ambulatory patients from a suburban location (presenting from March 14, 2020, to October 24, 2020, the internal ambulatory cohort) and 485 hospitalized patients from an inner-city location (hospitalized from March 14, 2020, to August 12, 2020, the external hospitalized cohort). The CNN model predictions were validated against electronic health record administrative codes in both cohorts and assessed using the area under the receiver operating characteristic curve (AUC). The CXRs from the ambulatory cohort were also reviewed by two board-certified radiologists and compared with the CNN-predicted values for the same cohort to produce a receiver operating characteristic curve and the AUC. The atherosclerosis diagnosis discrepancy, Δvasc, referring to the difference between the predicted value and presence or absence of the vascular disease HCC categorical code, was calculated. Linear regression was performed to determine the association of Δvasc with the covariates of age, sex, race/ethnicity, language preference, and social deprivation index. Logistic regression was used to look for an association between the presence of any hierarchical condition category codes with Δvasc and other covariates. The CNN prediction for vascular disease from frontal CXRs in the ambulatory cohort had an AUC of 0.85 (95% confidence interval, 0.82-0.89) and in the hospitalized cohort had an AUC of 0.69 (95% confidence interval, 0.64-0.75) against the electronic health record data. In the ambulatory cohort, the consensus radiologists' reading had an AUC of 0.89 (95% confidence interval, 0.86-0.92) relative to the CNN. Multivariate linear regression of Δvasc in the ambulatory cohort demonstrated significant negative associations with non-English-language preference (β=-0.083, P < .05) and Black or Hispanic race/ethnicity (β=-0.048, P < .05) and positive associations with age (β= 0.005, P < .001) and sex (β= 0.044, P < .05). For the hospitalized cohort, age was also significant (β= 0.003, P < .01), as was social deprivation index (β= 0.002, P < .05). The Δvasc variable (odds ratio [OR], 0.34), Black or Hispanic race/ethnicity (OR, 1.58), non-English-language preference (OR, 1.74), and site (OR, 0.22) were independent predictors of having one or more hierarchical condition category codes (P < .01 for all) in the combined patient cohort. A CNN model was predictive of aortic atherosclerosis in two cohorts (one ambulatory and one hospitalized) with COVID-19. The discrepancy between the CNN model and the administrative code, Δvasc, was associated with language preference in the ambulatory cohort; in the hospitalized cohort, this discrepancy was associated with social deprivation index. The absence of administrative code(s) was associated with Δvasc in the combined cohorts, suggesting that Δvasc is an independent predictor of health disparities. This may suggest that biomarkers extracted from routine imaging studies and compared with electronic health record data could play a role in enhancing value-based health care for traditionally underserved or disadvantaged patients for whom barriers to care exist.

Deep Learning for Patient-Specific Quality Assurance: Predicting Gamma Passing Rates for IMRT Based on Delivery Fluence Informed by log Files.

Objectives: In this study, we propose a deep learning-based approach to predict Intensity-modulated radiation therapy (IMRT) quality assurance (QA) gamma passing rates using delivery fluence informed by log files. Methods: A total of 112 IMRT plans for chest cancers were planned and measured by portal dosimetry equipped on TrueBeam linac. The convolutional neural network (CNN) based learning model was trained using delivery fluence as inputs and gamma passing rates (GPRs) of 4 different criteria (3%/3 mm, 2%/3 mm, 3%/2 mm, and 2%/2 mm) as outputs. Model performance for both validation and test sets was assessed using mean absolute error (MAE), mean squared error (MSE), root MSE (RMSE), Spearman rank correlation coefficients (Sr), and Determination coefficient (R2) between the measured and predicted GPR values. Results: In the test set, the MAE of the prediction model were 0.402, 0.511, 1.724, and 2.530, the MSE were 0.640, 0.986, 6.654, and 9.508, the RMSE were 0.800, 0.993, 2.580, and 3.083, the Sr were 0.643, 0.684, 0.821, and 0.824 (P < .001) and the R2 were 0.4110, 0.4666, 0.6677, and 0.6769 for 3%/3 mm, 3%/2 mm, 2%/3 mm, and 2%/2 mm, respectively. The MAE and RMSE of the prediction model decreased with stricter gamma criteria while the Sr and R2 between measured and predicted GPR values increased. Conclusions: The CNN prediction model based on delivery fluence informed by log files could accurately predict IMRT QA passing rates for different gamma criteria. It could reduce QA workload and improve efficiency in pretreatment QA. Our results suggest that the CNN prediction model based on delivery fluence informed by log files may be a promising tool for the gamma evaluation of IMRT QA.

How to Handle Data Imbalance and Feature Selection Problems in CNN-Based Stock Price Forecasting

Stock market forecasting is a time series problem that aims to predict possible future prices or directions of an index/stock. The stock data contains high uncertainty and is influenced by too many factors; hence it isn’t easy to achieve the goal by traditional time series methods. In literature, the convolutional neural networks (CNN) models were used for stock market forecasting and gave successful results. But, data imbalance due to labeling and feature selection problems were seen when considering these models. Hence, this study proposed a new rule-based labeling algorithm and a new feature selection approach to solve the issues. In addition, a CNN-based model, which was presented to predict the next day’s trade action of stocks in the Dow30 index, was constructed to check the effectiveness of the data labeling and the feature selection approach. Different image-based input variable sets were created using technical indicators, gold, and oil price data to feed the CNN model. The prediction performance of CNN models was compared with other studies in the literature. The experimental results showed that the CNN prediction model, which uses the proposed feature selection and labeling approaches in this study, performs 3-22% higher accuracy than the CNN-based models taking part in other studies. Also, the labeling approach proposed is more successful than Chen and Huang’s data weighting approach to solve the stock data imbalance problem. This algorithm reduced the ratio between labeled data from 15 times to 1.8 times.

Data‐Driven Prediction of Turbulent Flow Statistics Past Bridge Piers in Large‐Scale Rivers Using Convolutional Neural Networks

AbstractPrediction of statistical properties of the turbulent flow in large‐scale rivers is essential for river flow analysis. The large‐eddy simulation (LES) provides a powerful tool for such predictions; however, it requires a very long sampling time and demands significant computing power to calculate the turbulence statistics of riverine flows. In this study, we developed encoder‐decoder convolutional neural networks (CNNs) to predict the first‐ and second‐order turbulence statistics of the turbulent flow of large‐scale meandering rivers using instantaneous LES results. We train the CNNs using a data set obtained from LES of the flood flow in a large‐scale river with three bridge piers—a training testbed. Subsequently, we employed the trained CNNs to predict the turbulence statistics of the flood flow in two different meandering rivers and bridge pier arrangements—validation testbed rivers. The CNN predictions for the validation testbed river flow were compared with the simulation results of a separately done LES to evaluate the performance of the developed CNNs. We show that the trained CNNs can successfully produce turbulence statistics of the flood flow in the large‐scale rivers, that is, the validation testbeds.

Modeling of H2S solubility in ionic liquids using deep learning: A chemical structure-based approach

Hydrogen sulfide (H2S) is a toxic, flammable, corrosive, and acidic gas that can have harmful impacts on the environment. Ionic liquids (ILs) are relatively new solutions that have desirable characteristics such as high thermal and electrochemical stabilities, negligible vapor pressure, non-combustibility, and high solubility, allowing them to be an appropriate choice of liquid solvents in gas removal processes. In this study, deep learning (DL) approaches were utilized to predict the H2S solubility in ionic liquids (ILs). The proposed models include convolutional neural network (CNN), recurrent neural networks (RNNs), deep belief networks (DBN), and deep neural network initialization with decision trees (DJINN). To this end, a large databank in a broad range of pressure and temperature encompassing 1516 data points of H2S solubility in 37 various ILs was used for developing models. In these models, the chemical structure of ILs, temperature, and pressure were considered as the model’s inputs, while H2S solubility was the model’s output. The results reveal that the CNN model provides more accurate, rapid, flexible, and inexpensive estimations for H2S solubility in ILs with an average absolute percent relative error (AAPRE) of 2.92% and a determination coefficient (R2) of 0.99. The results were then compared to those of previously proposed techniques in the literature. According to the results of the sensitivity analysis, it was found that pressure and OH (as substructure) impose the highest positive and the highest negative impacts on the solubility of H2S in ILs, respectively. Also, based on sensitivity analysis, temperature has a reverse effect on the solubility of H2S and with increasing temperature, the H2S solubility decreases. The Taylor diagram illustrated that the CNN approach is very efficient, accurate, and realistic for predicting the solubility of H2S in diverse ILs based on the chemical structure, pressure, and temperature. Finally, trend analysis revealed that the trends of CNN predictions are in great agreement with the measured data of H2S solubility in ILs as a function of pressure. The findings of this work may be used not only to overcome the difficulties of calculating H2S solubility in ILs, but also to develop novel and accurate forecasting algorithms for a large dataset that cover a broad range of temperatures and pressures.

Time-Averaged Wind Turbine Wake Flow Field Prediction Using Autoencoder Convolutional Neural Networks

A convolutional neural network (CNN) autoencoder model has been developed to generate 3D realizations of time-averaged velocity in the wake of the wind turbines at the Sandia National Laboratories Scaled Wind Farm Technology (SWiFT) facility. Large-eddy simulations (LES) of the SWiFT site are conducted using an actuator surface model to simulate the turbine structures to produce training and validation datasets of the CNN. The simulations are validated using the SpinnerLidar measurements of turbine wakes at the SWiFT site and the instantaneous and time-averaged velocity fields from the training LES are used to train the CNN. The trained CNN is then applied to predict 3D realizations of time-averaged velocity in the wake of the SWiFT turbines under flow conditions different than those for which the CNN was trained. LES results for the validation cases are used to evaluate the performance of the CNN predictions. Comparing the validation LES results and CNN predictions, we show that the developed CNN autoencoder model holds great potential for predicting time-averaged flow fields and the power production of wind turbines while being several orders of magnitude computationally more efficient than LES.

A Deep Convolutional Neural Network for Prediction of Peptide Collision Cross Sections in Ion Mobility Spectrometry.

Most frequently, the identification of peptides in mass spectrometry-based proteomics is carried out using high-resolution tandem mass spectrometry. In order to increase the accuracy of analysis, additional information on the peptides such as chromatographic retention time and collision cross section in ion mobility spectrometry can be used. An accurate prediction of the collision cross section values allows erroneous candidates to be rejected using a comparison of the observed values and the predictions based on the amino acids sequence. Recently, a massive high-quality data set of peptide collision cross sections was released. This opens up an opportunity to apply the most sophisticated deep learning techniques for this task. Previously, it was shown that a recurrent neural network allows for predicting these values accurately. In this work, we present a deep convolutional neural network that enables us to predict these values more accurately compared with previous studies. We use a neural network with complex architecture that contains both convolutional and fully connected layers and comprehensive methods of converting a peptide to multi-channel 1D spatial data and vector. The source code and pre-trained model are available online.

Synthesizing Disparate LiDAR and Satellite Datasets through Deep Learning to Generate Wall-to-Wall Regional Inventories for the Complex, Mixed-Species Forests of the Eastern United States

Light detection and ranging (LiDAR) has become a commonly-used tool for generating remotely-sensed forest inventories. However, LiDAR-derived forest inventories have remained uncommon at a regional scale due to varying parameters among LiDAR data acquisitions and the availability of sufficient calibration data. Here, we present a model using a 3-D convolutional neural network (CNN), a form of deep learning capable of scanning a LiDAR point cloud, combined with coincident satellite data (spectral, phenology, and disturbance history). We compared this approach to traditional modeling used for making forest predictions from LiDAR data (height metrics and random forest) and found that the CNN had consistently lower uncertainty. We then applied the CNN to public data over six New England states in the USA, generating maps of 14 forest attributes at a 10 m resolution over 85% of the region. Aboveground biomass estimates produced a root mean square error of 36 Mg ha−1 (44%) and were within the 97.5% confidence of independent county-level estimates for 33 of 38 or 86.8% of the counties examined. CNN predictions for stem density and percentage of conifer attributes were moderately successful, while predictions for detailed species groupings were less successful. The approach shows promise for improving the prediction of forest attributes from regional LiDAR data and for combining disparate LiDAR datasets into a common framework for large-scale estimation.

A Novel Method for Predicting Local Site Amplification Factors Using 1-D Convolutional Neural Networks

The analysis of site seismic amplification characteristics is one of the important tasks of seismic safety evaluation. Owing to the high computational cost and complex implementation of numerical simulations, significant differences exist in the prediction of seismic ground motion amplification in engineering problems. In this paper, a novel prediction method for the amplification characteristics of local sites was proposed, using a state-of-the-art convolutional neural network (CNN) combined with real-time seismic signals. The amplification factors were computed by the standard spectral ratio method according to the observed records of seven stations in the Lower Hutt Valley, New Zealand. Based on the geological exploration data from the seven stations and the geological hazard information of the Lower Hutt Valley, eight parameters related to the seismic information were presumed to influence the amplification characteristics of the local site. The CNN method was used to establish the relationship between the amplification factors of local sites and the eight parameters, and the training samples and testing samples were generated through the observed and geological data other than the estimated values. To analyze the CNN prediction ability for amplification factors on unrecorded domains, two CNN models were established for comparison. One CNN model used about 80% of the data from 44 seismic events of the seven stations for training and the remaining data for testing. The other CNN model used the data of six stations to train and the remaining station’s data to test the CNN. The results showed that the CNN method based on the observation data can provide a powerful tool for predicting the amplification factors of local sites both for recorded positions and for unrecorded positions, while the traditional standard spectral ratio method only predicts the amplification factors for recorded positions. The comparison of the two CNN models showed that both can effectively predict the amplification factors of local ground motion without records, and the accuracy and stability of predictions can meet the requirements. With increasing seismic records, the CNN method becomes practical and effective for prediction purposes in earthquake engineering.

Abstract 14443: Deep Learning to Identify Atrial Fibrillation Phenotypes in Individual Patientsbased on Rate, Variability, Electrogram Shape

Introduction: Automatic identification of atrial fibrillation (AF) from Flutter or Tachycardia (AFL/AT) is key to therapy, yet devices often group them as ‘high-rate events’. Hypothesis: We hypothesized that (a) convolutional neural networks (CNN) may separate AF from AFL/AT better than Electrogram (EGM) rate or rhythm, and (b) The physiology represented by CNN could be probed using novel methods including computer modelling. Methods: We developed custom CNN on N= 450,960 EGM tracings (4 sec), in 110 patients with AF(n=55) or AFL/AT (n=55) at ablation (31.8% female, 65±11Y). Training used 88 patients; validation used 22. We systematically probed CNN with 140,360 EGM sequences controlling rate (cycle length, CL), beat-to-beat variability (%) or shape (autocorrelation, %) (figs A-B) alone or in combination. Results: CNN differentiated AF from AFL with a c-statistic of 0.95. Individual variations in rate (N=572), CL variability (N=418), shape (N=462), or in combination (N=26,620) decreased the c-statistic of the CNN predictions to 0.66, 0.84, 0.88, 0.58 ( P &lt;0.05, Delong test) respectively, suggesting rate as the primary determinant of CNN decision (Figure 1, C). Conclusions: Machine learning of intracardiac EGMs separates AF from AFL/AT significantly better than electrogram rate, regularity or shape. Opening the ‘black-box’ of CNN showed that additional indices from Electrograms identify the physiology of AF in individual patients. Studies on the physiological signatures for AF between patients may identify important clinical phenotypes.

Interactive Analysis of CNN Robustness

AbstractWhile convolutional neural networks (CNNs) have found wide adoption as state‐of‐the‐art models for image‐related tasks, their predictions are often highly sensitive to small input perturbations, which the human vision is robust against. This paper presents Perturber, a web‐based application that allows users to instantaneously explore how CNN activations and predictions evolve when a 3D input scene is interactively perturbed. Perturber offers a large variety of scene modifications, such as camera controls, lighting and shading effects, background modifications, object morphing, as well as adversarial attacks, to facilitate the discovery of potential vulnerabilities. Fine‐tuned model versions can be directly compared for qualitative evaluation of their robustness. Case studies with machine learning experts have shown that Perturber helps users to quickly generate hypotheses about model vulnerabilities and to qualitatively compare model behavior. Using quantitative analyses, we could replicate users’ insights with other CNN architectures and input images, yielding new insights about the vulnerability of adversarially trained models.

Fast EUV lithography simulation using convolutional neural network

Background: Thin mask model has been conventionally used in optical lithography simulation. In extreme ultraviolet (EUV) lithography thin mask model is not valid because the absorber thickness is comparable to the mask pattern size. Rigorous electromagnetic (EM) simulations have been used to calculate the thick mask amplitudes. However, these simulations are highly time consuming. Aim: Proposing a prototype of a convolutional neural network (CNN) which reduces the calculation time of rigorous EM simulations in a small mask area with specific mask patterns. Approach: We construct a CNN which reproduces the results of the EM simulation. We define mask 3D amplitude as the difference between the thick mask amplitude and the thin mask amplitude. The mask 3D amplitude of each diffraction order is approximated using three parameters which represent the on-axis and the off-axis mask 3D effects. The mask 3D parameters of all diffraction orders are trained by a CNN. Results: The input and the targets of the CNN are a cut-mask pattern and mask 3D parameters calculated by the EM simulation, respectively. After the training with 199,900 random cut-mask patterns, the CNN successfully predicts the mask 3D parameters of new cut-mask patterns. Conclusions: We construct a CNN which predicts the diffraction amplitudes from 2D EUV mask patterns. After the training, the CNN successfully reproduces the mask 3D amplitude. CNN prediction is 5000 times faster than the rigorous EM simulation. Next challenge is to construct a practical CNN which covers a large area with general mask patterns.

Performance of a 3D convolutional neural network in the detection of hypoperfusion at CT pulmonary angiography in patients with chronic pulmonary embolism: a feasibility study

BackgroundChronic pulmonary embolism (CPE) is a life-threatening disease easily misdiagnosed on computed tomography. We investigated a three-dimensional convolutional neural network (CNN) algorithm for detecting hypoperfusion in CPE from computed tomography pulmonary angiography (CTPA).MethodsPreoperative CTPA of 25 patients with CPE and 25 without pulmonary embolism were selected. We applied a 48%–12%–40% training-validation-testing split (12 positive and 12 negative CTPA volumes for training, 3 positives and 3 negatives for validation, 10 positives and 10 negatives for testing). The median number of axial images per CTPA was 335 (min–max, 111–570). Expert manual segmentations were used as training and testing targets. The CNN output was compared to a method in which a Hounsfield unit (HU) threshold was used to detect hypoperfusion. Receiver operating characteristic area under the curve (AUC) and Matthew correlation coefficient (MCC) were calculated with their 95% confidence interval (CI).ResultsThe predicted segmentations of CNN showed AUC 0.87 (95% CI 0.82–0.91), those of HU-threshold method 0.79 (95% CI 0.74–0.84). The optimal global threshold values were CNN output probability ≥ 0.37 and ≤ -850 HU. Using these values, MCC was 0.46 (95% CI 0.29–0.59) for CNN and 0.35 (95% CI 0.18–0.48) for HU-threshold method (average difference in MCC in the bootstrap samples 0.11 (95% CI 0.05–0.16). A high CNN prediction probability was a strong predictor of CPE.ConclusionsWe proposed a deep learning method for detecting hypoperfusion in CPE from CTPA. This model may help evaluating disease extent and supporting treatment planning.

Application of convolutional neural networks for prediction of strip flatness in tandem cold rolling process

The problems of the strip flatness defects are always severe in the tandem cold rolling process. It is of great significance to predict flatness for flatness control according to the process conditions of products. A prediction model based on convolutional neural network (CNN) was developed in this paper, and it can accurately predict strip flatness under various conditions. According to the distribution characteristics of industrial data collected in the rolling process, the isolated forest algorithm was used to eliminate outliers. Considering the special requirements of CNN on the dimension of input features, the data folding method was used to process the input features. Additionally, since strip flatness data is a vector rather than a scalar, and the length of this vector varies with strip width, which decreases the network's training accuracy. To deal with the problems, the loss function was modified. Taking the Inception module as the basic network structure and inspired by Wide & Deep learning, a strip flatness prediction model with high accuracy was developed. The optimal architecture and parameters of our network were determined through a lot of experimental explorations. The performances of BPNN (Back Propagation Neural Network), DNN (Deep Neural Network), and the proposed model were compared by mean square error (MSE) and coefficient of determination (R2). The result indicates that the proposed model has the highest prediction accuracy and better adaptability. It has the lowest MSE, 0.9891, and the highest R2, 0.9555. Finally, the fitting coefficients of Legendre polynomials were used to further prove the excellent prediction performance of the proposed model for strip flatness. Compared with other prediction models, it can obtain the lowest prediction error for the first quadratic and quartic components of strip flatness and it can be well-applied to tandem cold rolling production.

Cyanobacteria cell prediction using interpretable deep learning model with observed, numerical, and sensing data assemblage

Massive cyanobacterial blooms in river water causes adverse impacts on aquatic ecosystems and water quality. Complex and diverse data sources are available to investigate the cyanobacteria phenomena, including in situ data, synthetic measurements, and remote sensing images. Deep learning attention models can process these intricate sources to forecast cyanobacteria by identifying important variables in the data sources. However, deep learning attention models for predicting cyanobacteria have rarely been studied using an assemblage of various datasets. Thus, in this study, a convolutional neural network (CNN) model with a convolutional block attention module (CNNan) was developed to predict cyanobacterial cell concentrations by using the observed cell data from field monitoring, chlorophyll-a distribution map from hyperspectral image sensing, and simulated water quality outputs from a hydrodynamic model. Then, the prediction performance of the CNNan model was compared to an environmental fluid dynamics code (EFDC) simulation and a CNN model without an attention network. The seasonal variations of the predicted cyanobacteria that was obtained from CNNan showed the best agreement with the observed variations with Nash–Sutcliffe efficiency values higher than 0.76 when compared to the EFDC and CNN predictions. The daily hydrodynamic outputs allowed the prediction of cyanobacteria cells, while the rich information of the chlorophyll-a map contributed to the improvement of the prediction performance at certain periods. Moreover, the attention network visualized the importance of the additional chlorophyll-a map and improved the CNNan model prediction performance by refining the input features. Therefore, this study demonstrated that a deep learning model with data assemblage is practically feasible for predicting the presence of harmful algae in inland water.

A One-Dimensional Probabilistic Convolutional Neural Network for Prediction of Breast Cancer Survivability

Abstract Today, machine learning plays a major role in different branches of the healthcare industry, from prognosis and diagnosis to drug development providing a significant perspective on the medical landscape for disease prevention or treatment and the improvement of human life. Recently, the use of deep neural networks in different machine learning applications has shown a great contribution to the improvement of the accuracy of predictions. In this paper, a novel application of convolutional neural networks on medical prognosis is presented. The proposed method employs a one-dimensional convolutional neural network (1D-CNN) to predict the survivability of breast cancer patients. After further examining the network architecture, a number of 8, 14 and 24 convolutional filters were considered within three layers, respectively, followed by a max-pooling layer after the second and third layers. In addition, regarding the probabilistic nature of the survivability prediction problem, an extra layer was added to the network in order to calculate the probability of the patient survivability. To train the developed 1D-CNN machine, the SEER database as the most reliable repository of cancer survivability was used to retrieve the required training set. After a pre-processing to remove unusable records, a set of 50 000 breast cancer cases including 35 features was prepared for training the machine. Based on the results obtained in this study, the developed machine could reach an accuracy of 85.84%. This accuracy is the highest level of accuracy compared to the previous prediction methods. Furthermore, the mean squared error of the calculated probability was 0.112, which is an acceptable value of error for a probability calculation machine. The output of the developed machine can be used reliably by physicians to make decision about the most appropriate treatment strategy.

Artificial intelligence for melanoma diagnosis.

Convolutional neural networks (CNN) have shown unprecedented accuracy in digital image analysis, which can be harnessed for melanoma recognition through automated evaluation of clinical and dermatoscopic images. In experimental studies, modern CNN architectures perform single image analysis at the level of dermatologists and domain-experts, also for multiclass predictions including a multitude of possible diagnoses. This may not necessarily translate to good clinical performance, and reliable randomized controlled prospective clinical trials for modern CNNs are essentially missing. Weaknesses of CNNs are that limitations of available training image datasets propagate to limitations of CNN predictions, and they cannot provide a reliable estimate of uncertainty. Recent research focuses on human-computer collaboration, where gains in accuracy were measured even with imperfect CNNs. With missing academic and clinical agreement on equivocal melanocytic lesions, fully automating histologic assessment of them with CNNs appear problematic, and applications in the near future are probably limited to supporting, referencing or recommendation roles.

A CNN Prediction Method for Belt Grinding Tool Wear in a Polishing Process Utilizing 3-Axes Force and Vibration Data

This paper presents a tool wear monitoring methodology on the abrasive belt grinding process using vibration and force signatures on a convolutional neural network (CNN). A belt tool typically has a random orientation of abrasive grains and grit size variation for coarse or fine material removal. Degradation of the belt condition is a critical phenomenon that affects the workpiece quality during grinding. This work focuses on the identifation and the study of force and vibrational signals taken from sensors along an axis or combination of axes that carry important information of the contact conditions, i.e., belt wear. Three axes of the two sensors are aligned and labelled as X-axis (parallel to the direction of the tool during the abrasive process), Y-axis (perpendicular to the direction of the tool during the abrasive process) and Z-axis (parallel to the direction of the tool during the retract movement). The grinding process was performed using a customized abrasive belt grinder attached to a multi-axis robot on a mild-steel workpiece. The vibration and force signals along three axes (X, Y and Z) were acquired for four discrete sequential belt wear conditions: brand-new, 5-min cycle time, 15-min cycle time, and worn-out. The raw signals that correspond to the sensor measurement along the different axes were used to supervisedly train a 10-Layer CNN architecture to distinguish the belt wear states. Different possible combinations within the three axes of the sensors (X, Y, Z, XY, XZ, YZ and XYZ) were fed as inputs to the CNN model to sort the axis (or combination of axes) in the order of distinct representation of the belt wear state. The CNN classification results revealed that the combination of the XZ-axes and YZ-axes of the accelerometer sensor provides more accurate predictions than other combinations, indicating that the information from the Z-axis of the accelerometer is significant compared to the other two axes. In addition, the CNN accuracy of the XY-axes combination of dynamometer outperformed that of other combinations.

P0235 - Application of deep learning convolutional neural network in prediction of stone location, skin to stone distance and composition in renal lithiasis: A single center pilot study

P0235 - Application of deep learning convolutional neural network in prediction of stone location, skin to stone distance and composition in renal lithiasis: A single center pilot study

Identification and Localization of Endotracheal Tube on Chest Radiographs Using a Cascaded Convolutional Neural Network Approach

Rapid and accurate assessment of endotracheal tube (ETT) location is essential in the intensive care unit (ICU) setting, where timely identification of a mispositioned support device may prevent significant patient morbidity and mortality. This study proposes a series of deep learning-based algorithms which together iteratively identify and localize the position of an ETT relative to the carina on chest radiographs. Using the open-source MIMIC Chest X-Ray (MIMIC-CXR) dataset, a total of 16,000 patients were identified (8000 patients with an ETT and 8000 patients without an ETT). Three different convolutional neural network (CNN) algorithms were created. First, a regression loss function CNN was trained to estimate the coordinate location of the carina, which was then used to crop the original radiograph to the distal trachea and proximal bronchi. Second, a classifier CNN was trained using the cropped inputs to determine the presence or absence of an ETT. Finally, for radiographs containing an ETT, a third regression CNN was trained to both refine the coordinate location of the carina and identify the location of the distal ETT tip. Model accuracy was assessed by comparing the absolute distance of prediction and ground-truth coordinates as well as CNN predictions relative to measurements documented in original radiologic reports. Upon five-fold cross validation, binary classification for the presence or absence of ETT demonstrated an accuracy, sensitivity, specificity, PPV, NPV, and AUC of 97.14%, 97.37%, 96.89%, 97.12%, 97.15%, and 99.58% respectively. CNN predicted coordinate location of the carina, and distal ETT tip was estimated within a median error of 0.46 cm and 0.60 cm from ground-truth annotations respectively. Overall final CNN assessment of distance between the carina and distal ETT tip was predicted within a median error of 0.60 cm from manual ground-truth annotations, and a median error of 0.66 cm from measurements documented in the original radiology reports. A serial cascaded CNN approach demonstrates high accuracy for both identification and localization of ETT tip and carina on chest radiographs. High performance of the proposed multi-step strategy is in part related to iterative refinement of coordinate localization as well as explicit image cropping which focuses algorithm attention to key anatomic regions of interest.