Deep Learning-based Algorithm Research Articles

Fault transfer diagnosis of rolling bearings across different devices via multi-domain information fusion and multi-kernel maximum mean discrepancy

The current deep learning-based intelligent diagnosis algorithms depend on large amounts of well-labeled data, but they may not perform well in engineering practice where the fault data available from a single device is limited. The significant difference in data distribution between different devices makes it more difficult to transfer diagnosis across devices. Therefore, aiming at the problem that the great difference in the data distribution of source and target domains in cross-device transfer diagnosis leads to the low diagnosis accuracy, this paper investigates a deep transfer network model based on multi-domain information fusion and multi-kernel maximum mean discrepancy (MK-MMD) for fault diagnosis of rolling bearings across different devices. Firstly, to address the problem that single-domain information is difficult to adequately characterize rolling bearing health status information of different devices, a multi-domain information feature extractor based on multi-attention mechanism is constructed to capture the important features of vibration signals in different transform domains. Secondly, the extracted multi-domain features are input into the bidirectional gated recurrent unit (BiGRU) network for feature fusion, and MK-MMD is used to narrow the distribution distance between the source domain and target domain to obtain domain invariant features, and then achieve the transfer diagnosis of rolling bearing faults between different devices. Finally, the method investigated is verified by rolling bearing fault transfer diagnosis tests between different devices, and the results show that the suggested method can adapt to the feature distribution between different domains, and improve the transfer diagnosis accuracy between different devices.

CauDR: A causality-inspired domain generalization framework for fundus-based diabetic retinopathy grading

Diabetic retinopathy (DR) is the most common diabetic complication, which usually leads to retinal damage, vision loss, and even blindness. A computer-aided DR grading system has a significant impact on helping ophthalmologists with rapid screening and diagnosis. Recent advances in fundus photography have precipitated the development of novel retinal imaging cameras and their subsequent implementation in clinical practice. However, most deep learning-based algorithms for DR grading demonstrate limited generalization across domains. This inferior performance stems from variance in imaging protocols and devices inducing domain shifts. We posit that declining model performance between domains arises from learning spurious correlations in the data. Incorporating do-operations from causality analysis into model architectures may mitigate this issue and improve generalizability. Specifically, a novel universal structural causal model (SCM) was proposed to analyze spurious correlations in fundus imaging. Building on this, a causality-inspired diabetic retinopathy grading framework named CauDR was developed to eliminate spurious correlations and achieve more generalizable DR diagnostics. Furthermore, existing datasets were reorganized into 4DR benchmark for DG scenario. Results demonstrate the effectiveness and the state-of-the-art (SOTA) performance of CauDR. Diabetic retinopathy (DR) is the most common diabetic complication, which usually leads to retinal damage, vision loss, and even blindness. A computer-aided DR grading system has a significant impact on helping ophthalmologists with rapid screening and diagnosis. Recent advances in fundus photography have precipitated the development of novel retinal imaging cameras and their subsequent implementation in clinical practice. However, most deep learning-based algorithms for DR grading demonstrate limited generalization across domains. This inferior performance stems from variance in imaging protocols and devices inducing domain shifts. We posit that declining model performance between domains arises from learning spurious correlations in the data. Incorporating do-operations from causality analysis into model architectures may mitigate this issue and improve generalizability. Specifically, a novel universal structural causal model (SCM) was proposed to analyze spurious correlations in fundus imaging. Building on this, a causality-inspired diabetic retinopathy grading framework named CauDR was developed to eliminate spurious correlations and achieve more generalizable DR diagnostics. Furthermore, existing datasets were reorganized into 4DR benchmark for DG scenario. Results demonstrate the effectiveness and the state-of-the-art (SOTA) performance of CauDR.

Cascaded cross-attention transformers and convolutional neural networks for multi-organ segmentation in male pelvic computed tomography.

Segmentation of the prostate and surrounding organs at risk from computed tomography is required for radiation therapy treatment planning. We propose an automatic two-step deep learning-based segmentation pipeline that consists of an initial multi-organ segmentation network for organ localization followed by organ-specific fine segmentation. Initial segmentation of all target organs is performed using a hybrid convolutional-transformer model, axial cross-attention UNet. The output from this model allows for region of interest computation and is used to crop tightly around individual organs for organ-specific fine segmentation. Information from this network is also propagated to the fine segmentation stage through an image enhancement module, highlighting regions of interest in the original image that might be difficult to segment. Organ-specific fine segmentation is performed on these cropped and enhanced images to produce the final output segmentation. We apply the proposed approach to segment the prostate, bladder, rectum, seminal vesicles, and femoral heads from male pelvic computed tomography (CT). When tested on a held-out test set of 30 images, our two-step pipeline outperformed other deep learning-based multi-organ segmentation algorithms, achieving average dice similarity coefficient (DSC) of (prostate), (bladder), (rectum), (seminal vesicles), and (femoral heads). Our results demonstrate that a two-step segmentation pipeline with initial multi-organ segmentation and additional fine segmentation can delineate male pelvic CT organs well. The utility of this additional layer of fine segmentation is most noticeable in challenging cases, as our two-step pipeline produces noticeably more accurate and less erroneous results compared to other state-of-the-art methods on such images.

A Review of Document Binarization: Main Techniques, New Challenges, and Trends

Document image binarization is a challenging task, especially when it comes to text segmentation in degraded document images. The binarization, as a pre-processing step of Optical Character Recognition (OCR), is one of the most fundamental and commonly used segmentation methods. It separates the foreground text from the background of the document image to facilitate subsequent image processing. In view of the different degradation degrees of document images, researchers have proposed a variety of solutions. In this paper, we have summarized some challenges and difficulties in the field of document image binarization. Approximately 60 methods documenting image binarization techniques are mentioned, including traditional algorithms and deep learning-based algorithms. Here, we evaluated the performance of 25 image binarization techniques on the H-DIBCO2016 dataset to provide some help for future research.

S-Net: an S-shaped network for nodule detection in 3D CT images

Objective. Accurate and automatic detection of pulmonary nodules is critical for early lung cancer diagnosis, and promising progress has been achieved in developing effective deep models for nodule detection. However, most existing nodule detection methods merely focus on integrating elaborately designed feature extraction modules into the backbone of the detection network to extract rich nodule features while ignore disadvantages of the structure of detection network itself. This study aims to address these disadvantages and develop a deep learning-based algorithm for pulmonary nodule detection to improve the accuracy of early lung cancer diagnosis. Approach. In this paper, an S-shaped network called S-Net is developed with the U-shaped network as backbone, where an information fusion branch is used to propagate lower-level details and positional information critical for nodule detection to higher-level feature maps, head shared scale adaptive detection strategy is utilized to capture information from different scales for better detecting nodules with different shapes and sizes and the feature decoupling detection head is used to allow the classification and regression branches to focus on the information required for their respective tasks. A hybrid loss function is utilized to fully exploit the interplay between the classification and regression branches. Main results. The proposed S-Net network with ResSENet and other three U-shaped backbones from SANet, OSAF-YOLOv3 and MSANet (R+SC+ECA) models achieve average CPM scores of 0.914, 0.915, 0.917 and 0.923 on the LUNA16 dataset, which are significantly higher than those achieved with other existing state-of-the-art models. Significance. The experimental results demonstrate that our proposed method effectively improves nodule detection performance, which implies potential applications of the proposed method in clinical practice.

DeepDispatch: Deep Reinforcement Learning-Based Vehicle Dispatch Algorithm for Advanced Air Mobility

Near-future air taxi operations with electric vertical takeoff and landing aircraft will be constrained by the need for frequent recharging and limited takeoff and landing pads in vertiports and will be subject to time-varying demand and electricity prices, making the dispatch problem unique and particularly challenging to solve. Previously, the authors have developed optimization models to address this problem. Such optimization models, however, suffer from prohibitively high computational run times when the scale of the problem increases, making them less practical for real-world implementation. To overcome this issue, the authors have developed two deep reinforcement learning-based dispatch algorithms, namely, single-agent and multi-agent double dueling deep Q-network dispatch algorithms, where the objective is to maximize operating profit. A passenger transportation simulation environment was built to assess the performance of these algorithms across 36 numerical cases with varying numbers of vehicles and vertiports and amounts of demand. The results indicate that the multi-agent dispatch algorithm can closely approximate the optimal dispatch policy with significantly less computational expenses compared to the benchmark optimization model. The multi-agent algorithm was found to outperform the single-agent counterpart with respect to both profits generated and training time. Additionally, we implemented a heuristic-based algorithm, faster but less effective in generating profits compared to our two deep reinforcement learning-based algorithms.

Online Monitoring of Iron Ore Pellet Size Distribution Using Lightweight Convolutional Neural Network

An improved U-Net based convolutional neural network (CNN) is proposed in this paper for online size monitoring of iron ore pellets. The proposed CNN model is a lightweight version of U-Net. In the proposed model each convolution block of the encoder subnet consists of a concatenated <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1\times 1$</tex-math> </inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$3\times 3$</tex-math> </inline-formula> filter kernel, a batch normalization layer, and a ReLU nonlinearity. The proposed CNN model efficiently detects the pellets even under variable illumination conditions and images with overlapping pellets. The proposed model is evaluated using images of iron ore pellets acquired at the discharge end of a pelletizer disc. The proposed CNN model performs satisfactorily using <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\sim$</tex-math> </inline-formula> 39% less learnable parameters, and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\sim$</tex-math> </inline-formula> 16% reduced computational cost than the original U-Net model. The proposed model outperforms the conventional U-Net, different variants of U-Nets, FCN, and SegNet in weighted IoU, BF Score, global accuracy, SSIM, and PSNR. Experimental validation also shows a good agreement between the obtained pellet size distribution and the manual sieving result. This makes the proposed algorithm suitable for online pellet size-distribution measuring instruments. This method provides a more accurate and fast size monitoring system for iron ore pellets, ideal for real-time industrial applications. The image dataset and source codes are available at the below given link. https://github.com/aryadeo/Pellet_Size_Distribution. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Due to the advent of digitalization in industries, several attempts are being made to use innovative industrial monitoring systems in steel manufacturing plants. Most of the pellet plants in steel industries use human supervision for pellet growth monitoring. In this paper, a camera-based pellet size distribution monitoring method is proposed. A novel deep learning-based algorithm is used for the detection of pellets from the captured image followed by computation of pellet size analysis. Due to low computational complexity and better accuracy, the proposed method is suitable for online pellet size distribution system development compared to other previously proposed methods. However, protection from the harsh industrial environment is required for the uninterrupted and continuous operation of the system. As a future scope, the obtained pellet size information can be used as one of the control variables in designing a control system for the automated operation of the pelletizer disc.

Artificial intelligence in lung cancer screening: Detection, classification, prediction, and prognosis.

The exceptional capabilities of artificial intelligence (AI) in extracting image information and processing complex models have led to its recognition across various medical fields. With the continuous evolution of AI technologies based on deep learning, particularly the advent of convolutional neural networks (CNNs), AI presents an expanded horizon of applications in lung cancer screening, including lung segmentation, nodule detection, false-positive reduction, nodule classification, and prognosis. This review initially analyzes the current status of AI technologies. It then explores the applications of AI in lung cancer screening, including lung segmentation, nodule detection, and classification, and assesses the potential of AI in enhancing the sensitivity of nodule detection and reducing false-positive rates. Finally, it addresses the challenges and future directions of AI in lung cancer screening. AI holds substantial prospects in lung cancer screening. It demonstrates significant potential in improving nodule detection sensitivity, reducing false-positive rates, and classifying nodules, while also showing value in predicting nodule growth and pathological/genetic typing. AI offers a promising supportive approach to lung cancer screening, presenting considerable potential in enhancing nodule detection sensitivity, reducing false-positive rates, and classifying nodules. However, the universality and interpretability of AI results need further enhancement. Future research should focus on the large-scale validation of new deep learning-based algorithms and multi-center studies to improve the efficacy of AI in lung cancer screening.

Chicken tracking for location mapping of lameness chickens using YOLOv8 and deep learning-based tracking algorithm

The chicken farming industry is one of the biggest food industries that supports the achievement of food security internationally. Farmers need an independent tool that can monitor the welfare conditions of chickens in cages. Using their tools, farmers can ideally detect the condition of chickens. Lameness chickens, can be known for activity and dredging of their location in the cage. Occlusion, and background in the cage are interesting challenges. By observing behavior, image handling practices can be used to identify tainted chicks and provide an early warning of sickness in chickens. In this study, you only look once, version 8 (YOLOv8) which is a convolutional neural network (CNN) network model was chosen to perform the detection, tracking, and mapping of chicken locations. YOLOv8 was combined with various algorithm optimizers to improve training performance, such as root mean square (RMS) Prop, stochastic gradient descent (SGD), ADAM, and ADAMW. Multi-object tracking algorithms such as BOT-sort and ByteTrack are also used to improve tracking performance. Based on the results, YOLOv8 with combinations of optimizer algorithms ADAMW has the best mAP, support, precision and F1-score values compared to the others, with 0.936, 0.993, 0.990, 0.991. Meanwhile, for multi object tracking, ByteTrack is faster in inference time(s) values compared to the others, with 0.2.

CSDFormer: A cloud and shadow detection method for landsat images based on transformer

Cloud and shadow (CS) detection is crucial prerequisite for application of remote sensing images. Current deep learning-based detection algorithms mainly employ Convolutional Neural Networks (CNNs). However, the local receptive field in CNNs cannot effectively capture global contextual information, which hinders accurate characterization of the dependency between clouds and shadows. In vision Transformers, self-attention mechanisms can effectively capture the long-distance dependencies between different regions in an image. Inspired by this, this paper proposed a new CS Detection algorithm based on a Transformer, called CSDFormer. Specifically, we exclusively employed a hierarchical Transformer structure in the encoder stage to extract features of CS. Each Transformer layer contains several multi-head self-attention mechanisms for calculating pixel-wise long-distance connectivity. The designed structure enables the Transformer to better extract global context information, which helps to strengthen the comprehension of the semantic relationships between clouds and shadows. Benefiting from the global feature extraction capability of the encoder stage, we employed several simple multilayer perceptron layers for multi-scale feature map fusion and pixel classification in the decoder stage. The proposed CSDFormer was validated using 898 Landsat 8 Biome images with 512 × 512 pixels, producing an overall accuracy of 95.28 % and a mean intersection over union of 84.08 %, outperforming three state-of-the-art CNN-based algorithms. CSDFormer is consistently more accurate in detection of both clouds and shadows. Owing to the parallel computing capability of the self-attention mechanism, CSDFormer is computationally more efficient than the three CNN-based benchmark methods. For the input spectral bands, the performance of CSDFormer produced can be further enhanced with additional thermal infrared bands.

Analyzing Amazon Products Sentiment: A Comparative Study of Machine and Deep Learning, and Transformer-Based Techniques

In recent years, online shopping has surged in popularity, with customer reviews becoming a crucial aspect of the decision-making process. Reviews not only help potential customers make informed choices, but also provide businesses with valuable feedback and build trust. In this study, we conducted a thorough analysis of the Amazon reviews dataset, which includes several product categories. Our primary objective was to accurately classify sentiments using natural language processing, machine learning, ensemble learning, and deep learning techniques. Our research workflow encompassed several crucial steps. We explore data collection procedures; preprocessing steps, including normalization and tokenization; and feature extraction, utilizing the Bag-of-Words and TF–IDF methods. We conducted experiments employing a variety of machine learning algorithms, including Multinomial Naive Bayes, Random Forest, Decision Tree, and Logistic Regression. Additionally, we harnessed Bagging as an ensemble learning technique. Furthermore, we explored deep learning-based algorithms, such as CNNs, Bidirectional LSTM, and transformer-based models, like XLNet and BERT. Our comprehensive evaluations, utilizing metrics such as accuracy, precision, recall, and F1 score, revealed that the BERT algorithm outperformed others, achieving an impressive accuracy rate of 89%. This research provides valuable insights into the sentiment analysis of Amazon reviews, aiding both consumers and businesses in making informed decisions and enhancing product and service quality.

The economics of deep and machine learning-based algorithms for COVID-19 prediction, detection, and diagnosis shaping the organizational management of hospitals

Research background: Deep and machine learning-based algorithms can assist in COVID-19 image-based medical diagnosis and symptom tracing, optimize intensive care unit admission, and use clinical data to determine patient prioritization and mortality risk, being pivotal in qualitative care provision, reducing medical errors, and increasing patient survival rates, thus diminishing the massive healthcare system burden in relation to severe COVID-19 inpatient stay duration, while increasing operational costs throughout the organizational management of hospitals. Data-driven financial and scenario-based contingency planning, predictive modelling tools, and risk pooling mechanisms should be deployed for additional medical equipment and unforeseen healthcare demand expenses. Purpose of the article: We show that deep and machine learning-based and clinical decision making systems can optimize patient survival likelihood and treatment outcomes with regard to susceptible, infected, and recovered individuals, performing accurate analyses by data modeling based on vital and clinical signs, surveillance data, and infection-related biomarkers, and furthering hospital facility optimization in terms of intensive care unit bed allocation. Methods: The review software systems employed for article screening and quality evaluation were: AMSTAR, AXIS, DistillerSR, Eppi-Reviewer, MMAT, PICO Portal, Rayyan, ROBIS, and SRDR. Findings &amp; value added: Deep and machine learning-based clinical decision support tools can forecast COVID-19 spread, confirmed cases, and infection and mortality rates for data-driven appropriate treatment and resource allocations in effective therapeutic and diagnosis protocol development, by determining suitable measures and regulations and by using symptoms and comorbidities, vital signs, clinical and laboratory data and medical records across intensive care units, impacting the healthcare financing infrastructure. As a result of heightened use of personal protective equipment, hospital pharmacy and medication, outpatient treatment, and medical supplies, revenue loss and financial vulnerability occur, also due to expenses related to hiring additional staff and to critical resource expenditures. Hospital costs for COVID-19 medical care, screening, treatment capacity expansion, and personal protective equipment can lead to further financial losses while affecting COVID-19 frontline hospital workers and patients.

Comparison and Application of Deep Learning-Based Anomaly Detection Algorithms for Transparent Lens Defects

Comparison and Application of Deep Learning-Based Anomaly Detection Algorithms for Transparent Lens Defects

Development of an AI-based image/ultrasonic convergence camera system for accurate gas leak detection in petrochemical plants

Outdoor pipeline leaks are difficult to accurately measure using existing concentration measurement systems installed in petrochemical plants owing to external air currents. Besides, leak detection is only possible for a specific gas. The purpose of this study was to develop an image/ultrasonic convergence camera system that incorporates artificial intelligence (AI) to improve pipe leak detection and establish a real-time monitoring system. Our system includes an advanced ultrasonic camera coupled with a deep learning-based object-detection algorithm trained on pipe image data from petrochemical plants. The collected data improved the accuracy of detected gas leak localization through deep learning. Our detection model achieves an mAP50 (Mean average precision calculated at an intersection over union (IoU) threshold of 0.50)score of 0.45 on our data and is able to detect the majority of leak points within a system. The petrochemical plant environment was simulated by visiting petrochemical plants and reviewing drawings, and an outdoor experimental demonstration site was established. Scenarios such as flange connection failure were set under medium-/low-pressure conditions, and the developed product was experimented under gas leak conditions that simulated leakage accidents. These experiments enabled the removal of potentially confounding surrounding noise sources, which led to the false detection of actual gas leaks using the AI piping detection technique.

Deep Learning-Based Key Frame Recognition Algorithm for Adrenal Vascular in X-Ray Imaging

Adrenal vein sampling is required for the staging diagnosis of primary aldosteronism, and the frames in which the adrenal veins are presented are called key frames. Currently, the selection of key frames relies on the doctor's visual judgement which is time-consuming and laborious. This study proposes a key frame recognition algorithm based on deep learning. Firstly, wavelet denoising and multi-scale vessel-enhanced filtering are used to preserve the morphological features of the adrenal veins. Furthermore, by incorporating the self-attention mechanism, an improved recognition model called ResNet50-SA is obtained. Compared with commonly used transfer learning, the new model achieves 97.11% in accuracy, precision, recall, F1, and AUC, which is superior to other models and can help clinicians quickly identify key frames in adrenal veins.

A building change detection framework with patch-pairing single-temporal supervised learning and metric guided attention mechanism

Building change detection (CD) aims to detect changes in buildings from bi-temporal pairwise images obtained at different times. Typically, a deep learning-based building CD algorithm requires bi-temporal samples with significant building changes for training. However, obtaining such bi-temporal samples is challenging because building changes have a low probability of occurrence. Fortunately, it is relatively simple to obtain single-temporal samples that include a substantial number of buildings. By using these single-temporal building samples, pseudo bi-temporal building change samples can be generated, which can effectively address the problem of limited available bi-temporal building change samples. In view of that, this study proposes a metric guided single-temporal supervised learning framework that uses single-temporal building samples for building CD. In the proposed framework, patch-pairing single-temporal supervised learning (PPSL) adopts a patch-pairing method to construct pseudo bi-temporal building change samples, while equipping the network to effectively suppresses the negative impact of geometric offset and radiation difference in real samples. To further suppress the impact of radiation difference and enhance the effectiveness of our framework, a metric-guided spatial attention module (MGSAM) is designed to minimize the intra-class feature differences between temporal samples and augment the spatial context modeling ability. The proposed method is verified by experiments on different datasets, and the results demonstrate that the proposed method can outperform the existing methods and achieve superior performance.

The detailed investigation paper on network invasion awareness

Intrusion detection is critical for network security, with deep learning-based algorithms gaining traction. This project introduces NIDS - CNNLSTM, a model designed for the Industrial Internet of Things wireless sensing environment. It effectively identifies network traffic data, ensuring Industrial Internet of Things equipment and operations remain secure. Trained on the NSL_KDD data set, it exhibits strong convergence and performance across three data set, accurately classifying traffic types. Comparative analysis underscores NIDS - CNNLSTM efficacy enhancements. Experimental results validate increased detection rates, classification accuracy, and reduced false alarms, making it suitable for Industrial Internet of Things varied network data scenarios.

Research on collaborative edge network service migration strategy based on crowd clustering

The innovative application of Crowd Intelligent Devices (CIDS) in edge networks has garnered attention due to the rapid development of artificial intelligence and computer technology. This application offers users more reliable and low-latency computing services through computation offloading technology. However, the dynamic nature of network terminals and the limited coverage of edge servers pose challenges, such as data loss and service interruption. Furthermore, the high-speed mobility of intelligent terminals in the dynamic edge network environment further complicates the design of computation offloading and service migration strategies. To address these challenges, this paper explores the computation offloading model of cluster intelligence collaboration in a heterogeneous network environment. This model involves multiple intelligences collaborating to provide computation offloading services for terminals. To accommodate various roles, a switching strategy of split-cluster group collaboration is introduced, assigning the cluster head, the alternate cluster head, and the ordinary user are assigned to a group with different functions. Additionally, the paper formulates the optimal offloading strategy for group smart terminals as a Markov decision process, taking into account factors such as user mobility, service delay, service accuracy, and migration cost. To implement this strategy, the paper utilizes the deep reinforcement learning-based CCSMS algorithm. Simulation results demonstrate that the proposed edge network service migration strategy, rooted in groupwise cluster collaboration, effectively mitigates interruption delay and enhances service migration efficiency.

Number of intraepithelial lymphocytes and presence of a subepithelial band in normal colonic mucosa differs according to stainings and evaluation method

Chronic watery diarrhea is a frequent symptom. In approximately 10% of the patients, a diagnosis of microscopic colitis (MC) is established. The diagnosis relies on specific, but sometimes subtle, histopathological findings. As the histology of normal intestinal mucosa vary, discriminating subtle features of MC from normal tissue can be challenging and therefore auxiliary stainings are increasingly used. The aim of this study was to determine the variance in number of intraepithelial lymphocytes (IELs) and presence of a subepithelial band in normal ileum and colonic mucosa, according to different stains and digital assessment. Sixty-one patients without diarrhea referred to screening colonoscopy due to a positive feacal blood test and presenting with endoscopically normal mucosa were included. Basic histological features, number of IELs, and thickness of a subepithelial band was manually evaluated and a deep learning-based algorithm was developed to digitally determine the number of IELs in each of the two compartments; surface epithelium and cryptal epithelium, and the density of lymphocytes in the lamina propria compartment. The number of IELs was significantly higher on CD3-stained slides compared with slides stained with hematoxylin-and-eosin (HE) (p<0.001), and even higher numbers were reached using digital analysis. No significant difference between right and left colon in IELs or density of CD3-positive lymphocytes in lamina propria was found. No subepithelial band was present in HE-stained slides while a thin band was visualized on special stains. Conclusively, in this cohort of prospectively collected ileum and colonic biopsies from asymptomatic patients, the range of IELs and detection of a subepithelial collagenous band varied depending on the stain and method used for assessment. As assessment of biopsies from patients with diarrhea constitute a considerable workload in the pathology departments digital image analysis is highly desired. Knowledge provided by the present study highlight important differences that should be considered before introducing this method in the clinic.

Development and Validation of a Deep-Learning-Based Algorithm for Detecting and Classifying Metallic Implants in Abdominal and Spinal CT Topograms.

To develop and validate a deep-learning-based algorithm (DLA) that is designed to segment and classify metallic objects in topograms of abdominal and spinal CT. DLA training for implant segmentation and classification was based on a U-net-like architecture with 263 annotated hip implant topograms and 2127 annotated spine implant topograms. The trained DLA was validated with internal and external datasets. Two radiologists independently reviewed the external dataset consisting of 2178 abdomen anteroposterior (AP) topograms and 515 spine AP and lateral topograms, all collected in a consecutive manner. Sensitivity and specificity were calculated per pixel row and per patient. Pairwise intersection over union (IoU) was also calculated between the DLA and the two radiologists. The performance parameters of the DLA were consistently >95% in internal validation per pixel row and per patient. DLA can save 27.4% of reconstruction time on average in patients with metallic implants compared to the existing iMAR. The sensitivity and specificity of the DLA during external validation were greater than 90% for the detection of spine implants on three different topograms and for the detection of hip implants on abdominal AP and spinal AP topograms. The IoU was greater than 0.9 between the DLA and the radiologists. However, the DLA training could not be performed for hip implants on spine lateral topograms. A prototype DLA to detect metallic implants of the spine and hip on abdominal and spinal CT topograms improves the scan workflow with good performance for both spine and hip implants.