Artificial intelligence for human gunshot wound classification

Certain features are helpful in the identification of gunshot entrance and exit wounds, such as the presence of muzzle imprints, peripheral tears, stippling, bone beveling, and wound border irregularity. Some cases are less straightforward and wounds can thus pose challenges to an emergency room doctor or forensic pathologist. In recent years, deep learning has shown promise in various automated medical image classification tasks. This study explores the feasibility of using a deep learning model to classify entry and exit gunshot wounds in digital color images. A collection of 2418 images of entrance and exit gunshot wounds were procured. Of these, 2028 entrance and 1314 exit wounds were cropped, focusing on the area around each gunshot wound. A ConvNext Tiny deep learning model was trained using the Fastai deep learning library, with a train/validation split ratio of 70/30, until a maximum validation accuracy of 92.6% was achieved. An additional 415 entrance and 293 exit wound images were collected for the test (holdout) set. The model achieved an accuracy of 87.99%, precision of 83.99%, recall of 87.71%, and F1-score 85.81% on the holdout set. Correctly classified were 88.19% of entrance wounds and 87.71% of exit wounds. The results are comparable to what a forensic pathologist can achieve without other morphologic cues. This study represents one of the first applications of artificial intelligence to the field of forensic pathology. This work demonstrates that deep learning models can discern entrance and exit gunshot wounds in digital images with high accuracy.

Hierarchical classification of early microscopic lung nodule based on cascade network

Purpose

Early-stage lung cancer is typically characterized clinically by the presence of isolated lung nodules. Thousands of cases are examined each year, and one case usually contains numerous lung CT slices. Detecting and classifying early microscopic lung nodules is demanding due to their diminutive dimensions and restricted characterization capabilities. Therefore, a lung nodule classification model that performs well and is sensitive to microscopic lung nodules is needed to accurately classify lung nodules.

Methods

This paper uses the Resnet34 network as a basic classification model. A new cascade lung nodule classification method is proposed to classify lung nodules into 6 classes instead of the traditional 2 or 4 classes. It can effectively classify six different nodule types including ground-glass and solid nodules, benign and malignant nodules, and nodules with predominantly ground-glass or solid components.

Results

In this paper, the traditional multi-classification method and the cascade classification method proposed in this paper were tested using real lung nodule data collected in the clinic. The test results demonstrate that the cascade classification method in this study achieves an accuracy of 80.04% , outperforming the conventional multi-classification approach.

Conclusions

Different from the existing methods for categorizing the benign and malignant nature of lung nodules, the approach presented in this paper can classify lung nodules into 6 categories more accurately. At the same time, This paper proposes a rapid, precise, and dependable approach for classifying six distinct categories of lung nodules, which increases the accuracy categorization compared with the traditional multivariate categorization method.

Near-infrared spectroscopy based on colorimetric sensor array coupled with convolutional neural network detecting zearalenone in wheat

Wheat is a vital global cereal crop, but its susceptibility to contamination by mycotoxins can render it unusable. This study explored the integration of two novel non-destructive detection methodologies with convolutional neural network (CNN) for the identification of zearalenone (ZEN) contamination in wheat. Firstly, the colorimetric sensor array composed of six selected porphyrin-based materials was used to capture the olfactory signatures of wheat samples. Subsequently, the colorimetric sensor array, after undergoing a reaction, was characterized by its near-infrared spectral features. Then, the CNN quantitative analysis model was proposed based on the data, alongside the establishment of traditional machine learning models, partial least squares regression (PLSR) and support vector machine regression (SVR), for comparative purposes. The outcomes demonstrated that the CNN model had superior predictive performance, with a root mean square error of prediction (RMSEP) of 40.92 μ g ∙ kg-1 and a coefficient of determination on the prediction (R P 2 ) of 0.91. These results affirmed the potential of integrating colorimetric sensor array with near-infrared spectroscopy in evaluating the safety of wheat and potentially other grains. Moreover, CNN can have the capacity to autonomously learn and distill features from spectral data, enabling further spectral analysis and making it a forward-looking spectroscopic tool.

MEF-UNet: An end-to-end ultrasound image segmentation algorithm based on multi-scale feature extraction and fusion

Ultrasound image segmentation is a challenging task due to the complexity of lesion types, fuzzy boundaries, and low-contrast images along with the presence of noises and artifacts. To address these issues, we propose an end-to-end multi-scale feature extraction and fusion network (MEF-UNet) for the automatic segmentation of ultrasound images. Specifically, we first design a selective feature extraction encoder, including detail extraction stage and structure extraction stage, to precisely capture the edge details and overall shape features of the lesions. In order to enhance the representation capacity of contextual information, we develop a context information storage module in the skip-connection section, responsible for integrating information from adjacent two-layer feature maps. In addition, we design a multi-scale feature fusion module in the decoder section to merge feature maps with different scales. Experimental results indicate that our MEF-UNet can significantly improve the segmentation results in both quantitative analysis and visual effects.

Ensemble fusion model for improved lung abnormality classification: Leveraging pre-trained models

Lung abnormalities pose significant health concerns, underscoring the need for swift and accurate diagnoses to facilitate timely medical intervention. This study introduces a novel methodology for the sub-classification of lung abnormalities within chest X-rays captured via smartphones. An accurate and timely diagnosis of lung abnormalities is essential for the successful implementation of appropriate therapy. In this paper, we propose a novel approach using a Convolutional neural network (CNN) with three maximum pooling layers and early fusion for sub-classifying lung abnormalities from chest Xrays. Based on the kind of abnormality, the CheXpert dataset is divided into 13 sub-classes, each of which is trained using a different sub-model. An early fusion procedure is then used to integrate the outputs of the sub-model.•3M-CNN (Method 1): We employed a Convolutional Neural Network (CNN) with three max pooling layers and an early fusion strategy to train dedicated sub-models for each of the 13 distinct sub-classes of lung abnormalities using the CheXpert dataset.•Ensemble Model (Method 2): Our 'Ensemble model' integrated the outputs of the trained sub-models, providing a powerful approach for the sub-classification of lung abnormalities.•Exceptional Accuracy: Our '3M-CNN' and 'fused model' achieved an accuracy of 98.79%, surpassing established methodologies, which is beneficial in resource-constrained environments embracing smartphone-based imaging.

Quantitative analysis of microplastics in water environments based on Raman spectroscopy and convolutional neural network

With the increasing interest in microplastics (MPs) pollutants, quantitative analysis of MPs in water environment is an important issue. Vibrational spectroscopy, represented by Raman spectroscopy, is widely used in MP detection because they can provide unique fingerprint characteristics of chemical components of MPs, but it is difficult to provide quantitative information. In this paper, an ingenious method for quantitative analysis of MPs in water environment by combining Raman spectroscopy and convolutional neural network (CNN) is proposed. It is innovatively proposed to collect the average mapping spectra (AMS) of the samples to improve the uniformity of Raman spectroscopy detection, and to increase the effective detection range of concentration by filtering different volumes of the same MP solutions. In order to verify the universality and effectiveness of the proposed method, 6 different sizes of Polyethylene (PE) MPs were used as detection objects and mixed into 5 different actual water environments. The R2 and RMSE of CNN for identifying the concentration of PE solutions could reach 0.9972 and 0.033, respectively. Meanwhile, by comparing machine learning models such as Random Forest (RF) and Support Vector Machine (SVM) were compared, and CNN combined with Raman spectroscopy has significant advantages in identifying the concentration of MPs.

Integrated surface-enhanced Raman spectroscopy and convolutional neural network for quantitative and qualitative analysis of pesticide residues on pericarp

Pesticide residue poses a significant global public health concern, necessitating improved detection methods. Here, a novel platform was introduced based on surface-enhanced Raman spectroscopy (SERS) to detect ten distinct types of pesticides. Notably, the sensitivity of this approach is exemplified by detecting trace amounts of 50 pM (10 ppt) thiabendazole. The correlation between the characteristic peak intensity of coexisting pesticides and their concentrations displays an exceptional linear relationship (R2 = 0.9999), underscoring its utility for quantitative mixed pesticide detection. Additionally, qualitative analysis of five mixed pesticides was conducted leveraging distinctive peak labeling. Harnessing machine learning techniques, a model for classifying and predicting pesticides on pericarps was developed. Remarkably, the convolutional neural network achieved classification accuracy of 100 % and prediction accuracy of 99.62 %. This innovative approach accurately identifies and quantifies diverse pesticides, thus offering a feasible scheme for in-situ detection of pesticide residues. Ultimately, this strategy contributes to ensuring food safety and public health.

Ultrasensitive SERS quantitative detection of antioxidants via diazo derivatization reaction and deep learning for signal fluctuation mitigation

Synthetic antioxidants serve as essential protectors against oxidation and deterioration of edible oils, however, prudent evaluation is necessary regarding potential health risks associated with excessive intake. The direct adsorption of antioxidants onto conventional surface-enhanced Raman scattering (SERS) substrates is challenging due to the presence of phenolic hydroxyl groups in their molecular structures, resulting in weak Raman scattering signals and rendering direct SERS detection difficult. In this study, a diazo derivatization reaction was employed to enhance SERS signals by converting antioxidant molecules into azo derivatives, enabling the amplification of the weak Raman scattering signals through the strong vibrational modes induced by the N = N double bond. The resulting diazo derivatives were characterized using UV-visible absorption and infrared spectroscopy, confirming the occurrence of diazo derivatization of the antioxidants. The proposed method successfully achieved the rapid detection of three commonly used synthetic antioxidants, namely butylated hydroxyanisole (BHA), tert-butylhydroquinone (TBHQ), and propyl gallate (PG) on interfacial self-assembled gold nanoparticles. Furthermore, rapid predictions of BHA, PG, and TBHQ within the concentration range of 1 × 10-6 to 2 × 10-3 mol/L were achieved by integrating a convolutional neural network model. The predictive range of this model surpassed the traditional quantitative method of manually selecting characteristic peaks, with linear coefficients (R2) of 0.9992, 0.9997, and 0.9997, respectively. The recovery of antioxidants in real soybean oil samples ranged from 73.0 % to 126.4 %. Based on diazo derivatization, the proposed SERS method eliminates the need for complex substrates and enables the analysis and determination of synthetic antioxidants in edible oils within 20 min, providing a convenient analytical approach for quality control in the food industry.

Non-destructive detection of defective maize kernels using hyperspectral imaging and convolutional neural network with attention module

Rapid, effective and non-destructive detection of the defective maize kernels is crucial for their high-quality storage in granary. Hyperspectral imaging (HSI) coupled with convolutional neural network (CNN) based on spectral and spatial attention (Spl-Spal-At) module was proposed for identifying the different types of maize kernels. The HSI data within 380-1000 nm of six classes of sprouted, heat-damaged, insect-damaged, moldy, broken and healthy kernels was collected. The CNN-Spl-At, CNN-Spal-At and CNN-Spl-Spal-At models were established based on the spectra, images and their fusion features as inputs for the recognition of different kernels. Further compared the performances of proposed models and conventional models were built by support vector machine (SVM) and extreme learning machine (ELM). The results indicated that the recognition ability of CNN with attention series models was significantly better than that of SVM and ELM models and fused features were more conducive to expressing the appearance of different kernels than single features. And the CNN-Spl-Spal-At model had an optimal recognition result with high average classification accuracy of 98.04 % and 94.56 % for the training and testing sets, respectively. The recognition results were visually presented on the surface image of kernels with different colors. The CNN-Spl-Spal-At model was built in this study could effectively detect defective maize kernels, and it also had great potential to provide the analysis approaches for the development of non-destructive testing equipment based on HSI technique for maize quality.

A diagnostic strategy for pulmonary fat embolism based on routine H&E staining using computational pathology

Pulmonary fat embolism (PFE) as a cause of death often occurs in trauma cases such as fractures and soft tissue contusions. Traditional PFE diagnosis relies on subjective methods and special stains like oil red O. This study utilizes computational pathology, combining digital pathology and deep learning algorithms, to precisely quantify fat emboli in whole slide images using conventional hematoxylin-eosin (H&E) staining. The results demonstrate deep learning's ability to identify fat droplet morphology in lung microvessels, achieving an area under the receiver operating characteristic (ROC) curve (AUC) of 0.98. The AI-quantified fat globules generally matched the Falzi scoring system with oil red O staining. The relative quantity of fat emboli against lung area was calculated by the algorithm, determining a diagnostic threshold of 8.275% for fatal PFE. A diagnostic strategy based on this threshold achieved a high AUC of 0.984, similar to manual identification with special stains but surpassing H&E staining. This demonstrates computational pathology's potential as an affordable, rapid, and precise method for fatal PFE diagnosis in forensic practice.

Elucidating immune cell dynamics in chronic lung allograft dysfunction: A comprehensive single-cell transcriptomic study

Chronic Lung Allograft Dysfunction (CLAD) is a critical post-transplant complication that predominantly determines the long-term survival rates and quality of life of patients undergoing lung transplantation. The limited efficacy of current immunosuppressive strategies underscores our incomplete understanding of the immunological aspects of CLAD. Hence, there is an urgent need for more comprehensive and targeted research to unravel the complex interplay of immune cells in the development and progression of CLAD. This study conducts an in-depth analysis of the immune environment in CLAD. By examining the gene expression profiles of T cells, natural killer cells, B cells, macrophages, and monocytes, we have elucidated a unique immunological landscape in CLAD compared to healthy controls. We highlight the heterogeneity within the immune populations and provide a comprehensive understanding of the immune mechanisms driving CLAD. Enrichment analysis identified specific pathways that are either overactive or suppressed in CLAD, revealing potential molecular targets for therapeutic intervention. Our findings emphasize the crucial role of T cells in the pathophysiology of CLAD, coordinating the immune response and revealing an amplified immune cell network, potentially leading to maladaptive tissue responses. By integrating a comprehensive cellular and molecular portrait of the immune environment, our research not only deepens our understanding of the pathogenesis of CLAD but also lays a foundational approach for the development of targeted therapies.

Transformer based multi-modal MRI fusion for prediction of post-menstrual age and neonatal brain development analysis

The brain development during the perinatal period is characterized by rapid changes in both structure and function, which have significant impact on the cognitive and behavioral abilities later in life. Accurate assessment of brain age is a crucial indicator for brain development maturity and can help predict the risk of neonatal pathology. However, evaluating neonatal brains using magnetic resonance imaging (MRI) is challenging due to its complexity, multi-dimension, and noise with subtle alterations. In this paper, we propose a multi-modal deep learning framework based on transformers for precise post-menstrual age (PMA) estimation and brain development analysis using T2-weighted structural MRI (T2-sMRI) and diffusion MRI (dMRI) data. First, we build a two-stream dense network to learn modality-specific features from T2-sMRI and dMRI of brain individually. Then, a transformer module based on self-attention mechanism integrates these features for PMA prediction and preterm/term classification. Finally, saliency maps on brain templates are used to enhance the interpretability of results. Our method is evaluated on the multi-modal MRI dataset of the developing Human Connectome Project (dHCP), which contains 592 neonates, including 478 term-born and 114 preterm-born subjects. The results demonstrate that our method achieves a 0.5-week mean absolute error (MAE) in PMA estimation for term-born subjects. Notably, preterm-born subjects exhibit delayed brain development, worsening with increasing prematurity. Our method also achieves 95% accuracy in classification of term-born and preterm-born subjects, revealing significant group differences.

Deep learning performance compared to healthcare experts in detecting wrist fractures from radiographs: A systematic review and meta-analysis

OBJECTIVE

To perform a systematic review and meta-analysis of the diagnostic accuracy of deep learning (DL) algorithms in the diagnosis of wrist fractures (WF) on plain wrist radiographs, taking healthcare experts consensus as reference standard.

METHODS

Embase, Medline, PubMed, Scopus and Web of Science were searched in the period from 1 Jan 2012 to 9 March 2023. Eligible studies were patients with wrist radiographs for radial and ulnar fractures as the target condition, studies using DL algorithms based on convolutional neural networks (CNN), and healthcare experts consensus as the minimum reference standard. Studies were assessed with a modified QUADAS-2 tool, and we applied a bivariate random-effects model for meta-analysis of diagnostic test accuracy data.

RESULTS

Our study was registered at PROSPERO with ID: CRD42023431398. We included 6 unique studies for meta-analysis, with a total of 33,026 radiographs. CNN performance compared to reference standards for the included articles found a summary sensitivity of 92% (95% CI: 80%-97%) and a summary specificity of 93% (95% CI: 76%-98%). The generalized bivariate I-squared statistic indicated considerable heterogeneity between the studies (81.90%). Four studies had one or more domains at high risk of bias and two studies had concerns regarding applicability.

CONCLUSION

The diagnostic accuracy of CNNs was comparable to that of healthcare experts in wrist radiographs for investigation of WF. There is a need for studies with a robust reference standard, external data-set validation and investigation of diagnostic performance of healthcare experts aided with CNNs.

CLINICAL RELEVANCE STATEMENT

DL matches healthcare experts in diagnosing WFs, which potentially benefits patient diagnosis.

A sleep staging model on wavelet-based adaptive spectrogram reconstruction and light weight CNN

Effective methods for automatic sleep staging are important for diagnosis and treatment of sleep disorders. EEG has weak signal properties and complex frequency components during the transition of sleep stages. Wavelet-based adaptive spectrogram reconstruction (WASR) by seed growth is utilized to capture dominant time-frequency patterns of sleep EEG. We introduced variant energy from Teager operator in WASR to capture hidden dynamic patterns of EEG, which produced additional spectrograms. These spectrograms enabled a light weight CNN to detect and extract finer details of different sleep stages, which improved the feature representation of EEG. With specially designed depthwise separable convolution, the light weight CNN achieved more robust sleep stage classification. Experimental results on Sleep-EDF 20 dataset showed that our proposed model yielded overall accuracy of 87.6%, F1-score of 82.1%, and Cohen kappa of 0.83, which is competitive compared with baselines with reduced computation cost.

CECT: Controllable ensemble CNN and transformer for COVID-19 image classification

The COVID-19 pandemic has resulted in hundreds of million cases and numerous deaths worldwide. Here, we develop a novel classification network CECT by controllable ensemble convolutional neural network and transformer to provide a timely and accurate COVID-19 diagnosis. The CECT is composed of a parallel convolutional encoder block, an aggregate transposed-convolutional decoder block, and a windowed attention classification block. Each block captures features at different scales from 28 × 28 to 224 × 224 from the input, composing enriched and comprehensive information. Different from existing methods, our CECT can capture features at both multi-local and global scales without any sophisticated module design. Moreover, the contribution of local features at different scales can be controlled with the proposed ensemble coefficients. We evaluate CECT on two public COVID-19 datasets and it reaches the highest accuracy of 98.1% in the intra-dataset evaluation, outperforming existing state-of-the-art methods. Moreover, the developed CECT achieves an accuracy of 90.9% on the unseen dataset in the inter-dataset evaluation, showing extraordinary generalization ability. With remarkable feature capture ability and generalization ability, we believe CECT can be extended to other medical scenarios as a powerful diagnosis tool. Code is available at https://github.com/NUS-Tim/CECT.

CD-Net: Cascaded 3D Dilated convolutional neural network for pneumonia lesion segmentation

COVID-19 is a global pandemic that has caused significant global, social, and economic disruption. To effectively assist in screening and monitoring diagnosed cases, it is crucial to accurately segment lesions from Computer Tomography (CT) scans. Due to the lack of labeled data and the presence of redundant parameters in 3D CT, there are still significant challenges in diagnosing COVID-19 in related fields. To address the problem, we have developed a new model called the Cascaded 3D Dilated convolutional neural network (CD-Net) for directly processing CT volume data. To reduce memory consumption when cutting volume data into small patches, we initially design a cascade architecture in CD-Net to preserve global information. Then, we construct a Multi-scale Parallel Dilated Convolution (MPDC) block to aggregate features of different sizes and simultaneously reduce the parameters. Moreover, to alleviate the shortage of labeled data, we employ classical transfer learning, which requires only a small amount of data while achieving better performance. Experimental results conducted on the different public-available datasets verify that the proposed CD-Net has reduced the negative-positive ratio and outperformed other existing segmentation methods while requiring less data.

Implementation of near-infrared spectroscopy and convolutional neural networks for predicting particle size distribution in fluidized bed granulation

Monitoring the particle size distribution (PSD) is crucial for controlling product quality during fluidized bed granulation. This paper proposed a rapid analytical method that quantifies the D10, D50, and D90 values using a Convolutional Block Attention Module-Convolutional Neural Network (CBAM-CNN) framework tailored for deep learning with near-infrared (NIR) spectroscopy. This innovative framework, which fuses CBAM with CNN, excels at extracting intricate features while prioritizing crucial ones, thereby facilitating the creation of a robust multi-output regression model. To expand the training dataset, we incorporated the C-Mixup algorithm, ensuring that the deep learning model was trained comprehensively. Additionally, the Bayesian optimization algorithm was introduced to optimize the hyperparameters, improving the prediction performance of the deep learning model. Compared with the commonly used Partial Least Squares (PLS), Support Vector Machine (SVM), and Artificial Neural Network (ANN) models, the CBAM-CNN model yielded higher prediction accuracy. Furthermore, the CBAM-CNN model avoided spectral preprocessing, preserved the spectral information to the maximum extent, and returned multiple predicted values at one time without degrading the prediction accuracy. Therefore, the CBAM-CNN model showed better prediction performance and modeling convenience for analyzing PSD values in fluidized bed granulation.

UroAngel: a single-kidney function prediction system based on computed tomography urography using deep learning

BACKGROUND

Accurate estimation of the glomerular filtration rate (GFR) is clinically crucial for determining the status of obstruction, developing treatment strategies, and predicting prognosis in obstructive nephropathy (ON). We aimed to develop a deep learning-based system, named UroAngel, for non-invasive and convenient prediction of single-kidney function level.

METHODS

We retrospectively collected computed tomography urography (CTU) images and emission computed tomography diagnostic reports of 520 ON patients. A 3D U-Net model was used to segment the renal parenchyma, and a logistic regression multi-classification model was used to predict renal function level. We compared the predictive performance of UroAngel with the Modification of Diet in Renal Disease (MDRD), Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations, and two expert radiologists in an additional 40 ON patients to validate clinical effectiveness.

RESULTS

UroAngel based on 3D U-Net convolutional neural network could segment the renal cortex accurately, with a Dice similarity coefficient of 0.861. Using the segmented renal cortex to predict renal function stage had high performance with an accuracy of 0.918, outperforming MDRD and CKD-EPI and two radiologists.

CONCLUSIONS

We proposed an automated 3D U-Net-based analysis system for direct prediction of single-kidney function stage from CTU images. UroAngel could accurately predict single-kidney function in ON patients, providing a novel, reliable, convenient, and non-invasive method.

Multimodal ECG heartbeat classification method based on a convolutional neural network embedded with FCA

Arrhythmias are irregular heartbeat rhythms caused by various conditions. Automated ECG signal classification aids in diagnosing and predicting arrhythmias. Current studies mostly focus on 1D ECG signals, overlooking the fusion of multiple ECG modalities for enhanced analysis. We converted ECG signals into modal images using RP, GAF, and MTF, inputting them into our classification model. To optimize detail retention, we introduced a CNN-based model with FCA for multimodal ECG tasks. Achieving 99.6% accuracy on the MIT-BIH arrhythmia database for five arrhythmias, our method outperforms prior models. Experimental results confirm its reliability for ECG classification tasks.

Tinnitus classification based on resting-state functional connectivity using a convolutional neural network architecture

OBJECTIVES

Many studies have investigated aberrant functional connectivity (FC) using resting-state functional MRI (rs-fMRI) in subjective tinnitus patients. However, no studies have verified the efficacy of resting-state FC as a diagnostic imaging marker. We established a convolutional neural network (CNN) model based on rs-fMRI FC to distinguish tinnitus patients from healthy controls, providing guidance and fast diagnostic tools for the clinical diagnosis of subjective tinnitus.

METHODS

A CNN architecture was trained on rs-fMRI data from 100 tinnitus patients and 100 healthy controls using an asymmetric convolutional layer. Additionally, a traditional machine learning model and a transfer learning model were included for comparison with the CNN, and each of the three models was tested on three different brain atlases.

RESULTS

Of the three models, the CNN model outperformed the other two models with the highest area under the curve, especially on the Dos_160 atlas (AUC = 0.944). Meanwhile, the model with the best classification performance highlights the crucial role of the default mode network, salience network, and sensorimotor network in distinguishing between normal controls and patients with subjective tinnitus.

CONCLUSION

Our CNN model could appropriately tackle the diagnosis of tinnitus patients using rs-fMRI and confirmed the diagnostic value of FC as measured by rs-fMRI.

Robust QRS detection based on simulated degenerate optical parametric oscillator-assisted neural network

Accurately detecting the depolarization QRS complex in the ventricles is a fundamental requirement for cardiovascular disease detection using electrocardiography (ECG). In contrast to traditional signal enhancement algorithms, emerging neural network approaches have shown promise for QRS detection because of their generalizability on complex data. However, the inevitable noise present during ECG recording leads to a decrease in the performance of neural networks. To enhance the robustness and performance of neural network-based QRS detectors, we propose a simulated degeneration unit (SDU)-assisted convolutional neural network (CNN). An SDU simulates the physical degeneration process of interfering optical pulses, which can effectively suppress in-band noise. Through comprehensive performance evaluations on three open-source databases, the SDU-enhanced CNN-based approach demonstrated better performance in detecting QRS complexes than other recently reported QRS detectors. Furthermore, real-world noise injection tests indicate that the optimal noise robustness boundary for the CNN equipped with SDU is 167-300% higher than that for the CNN without SDU.

Deep transfer learning with gravitational search algorithm for enhanced plant disease classification

Plant diseases annually cause damage and loss of much of the crop, if not its complete destruction, and this constitutes a significant challenge for farm owners, governments, and consumers alike. Therefore, identifying and classifying diseases at an early stage is very important in order to sustain local and global food security. In this research, we designed a new method to identify plant diseases by combining transfer learning and Gravitational Search Algorithm (GSA). Two state-of-the-art pretrained models have been adopted for extracting features in this study, which are MobileNetV2 and ResNe50V2. Multilayer feature extraction is applied in this study to ensure representations of plant leaves from different levels of abstraction for precise classification. These features are then concatenated and passed to GSA for optimizing them. Finally, optimized features are passed to Multinomial Logistic Regression (MLR) for final classification. This integration is essential for categorizing 18 different types of infected and healthy leaf samples. The performance of our approach is strengthened by a comparative analysis that incorporates features optimized by the Genetic Algorithm (GA). Additionally, the MLR algorithm is contrasted with K-Nearest Neighbors (KNN). The empirical findings indicate that our model, which has been refined using GSA, achieves very high levels of precision. Specifically, the average precision for MLR is 99.2%, while for KNN it is 98.6%. The resulting results significantly exceed those achieved with GA-optimized features, thereby highlighting the superiority of our suggested strategy. One important result of our study is that we were able to decrease the number of features by more than 50%. This reduction greatly reduces the processing requirements without sacrificing the quality of the diagnosis. This work presents a robust and efficient approach to the early detection of plant diseases. The work demonstrates the utilization of sophisticated computational methods in agriculture, enabling the development of novel data-driven strategies for plant health management, therefore enhancing worldwide food security.

Classification of skin blemishes with cell phone images using deep learning techniques

Skin blemishes can be caused by multiple events or diseases and, in some cases, it is difficult to distinguish where they come from. Therefore, there may be cases with a dangerous origin that go unnoticed or the opposite case (which can lead to overcrowding of health services). To avoid this, the use of artificial intelligence-based classifiers using images taken with mobile devices is proposed; this would help in the initial screening process and provide some information to the patient prior to their final diagnosis. To this end, this work proposes an optimization mechanism based on two phases in which a global search for the best classifiers (from among more than 150 combinations) is carried out, and, in the second phase, the best candidates are subjected to a phase of evaluation of the robustness of the system by applying the cross-validation technique. The results obtained reach 99.95% accuracy for the best case and 99.75% AUC. Comparing the developed classifier with previous works, an improvement in terms of classification rate is appreciated, as well as in the reduction of the classifier complexity, which allows our classifier to be integrated in a specific purpose system with few computational resources.

Multiple marine algae identification based on three-dimensional fluorescence spectroscopy and multi-label convolutional neural network

Accurate identification of algal populations plays a pivotal role in monitoring seawater quality. Fluorescence-based techniques are effective tools for quickly identifying different algae. However, multiple coexisting algae and their similar photosynthetic pigments can constrain the efficacy of fluorescence methods. This study introduces a multi-label classification model that combines a specific Excitation-Emission matric convolutional neural network (EEM-CNN) with three-dimensional (3D) fluorescence spectroscopy to detect single and mixed algal samples. Spectral data can be input directly into the model without transforming into images. Rectangular convolutional kernels and double convolutional layers are applied to enhance the extraction of balanced and comprehensive spectral features for accurate classification. A dataset comprising 3D fluorescence spectra from eight distinct algae species representing six different algal classes was obtained, preprocessed, and augmented to create input data for the classification model. The classification model was trained and validated using 4448 sets of test samples and 60 sets of test samples, resulting in an accuracy of 0.883 and an F1 score of 0.925. This model exhibited the highest recognition accuracy in both single and mixed algae samples, outperforming comparative methods such as ML-kNN and N-PLS-DA. Furthermore, the classification results were extended to three different algae species and mixed samples of skeletonema costatum to assess the impact of spectral similarity on multi-label classification performance. The developed classification models demonstrated robust performance across samples with varying concentrations and growth stages, highlighting CNN's potential as a promising tool for the precise identification of marine algae.

Predicting DNA toehold-mediated strand displacement rate constants using a DNA-BERT transformer deep learning model

Dynamic DNA nanotechnology is driving exciting developments in molecular computing, cargo delivery, sensing and detection. Combining this innovative area of research with the progress made in machine learning will aid in the design of sophisticated DNA machinery. Herein, we present a novel framework based on a transformer architecture and a deep learning model which can predict the rate constant of toehold-mediated strand displacement, the underlying process in dynamic DNA nanotechnology. Initially, a dataset of 4450 DNA sequences and corresponding rate constants were generated in-silico using KinDA. Subsequently, a 1D convolution neural network was trained using specific local features and DNA-BERT sequence embedding to produce predicted rate constants. As a result, the newly trained deep learning model predicted toehold-mediated strand displacement rate constants with a root mean square error of 0.76, during testing. These findings demonstrate that DNA-BERT can improve prediction accuracy, negating the need for extensive computational simulations or experimentation. Finally, the impact of various local features during model training is discussed, and a detailed comparison between the One-hot encoder and DNA-BERT sequences representation methods is presented.

Brain tumor recognition by an optimized deep network utilizing ammended grasshopper optimization

Brain tumors are abnormal cell masses that can get originated in the brain spread from other organs. They can be categorized as either malignant (cancerous) or benign (noncancerous), and their growth rates and locations can impact the functioning of the nerve system. The timely detection of brain tumors is crucial for effective treatment and prognosis. In this study, a new approach has been proposed for diagnosing brain tumors using deep learning and a meta-heuristic algorithm. The method involves three main steps: (1) extracting features from brain MRI images using AlexNet, (2) reducing the complexity of AlexNet by employing an Extreme Learning Machine (ELM) network as a classification layer, and (3) fine-tuning the parameters of the ELM network using an Amended Grasshopper Optimization Algorithm (AGOA). The performance of the method has been evaluated on a publicly available dataset consisting of 20 patients with newly diagnosed glioblastoma that is compared with several state-of-the-art techniques. Experimental results demonstrate that the method achieves the highest accuracy, precision, specificity, F1-score, sensitivity, and MCC with values of 0.96, 0.94, 0.96, 0.96, 0.94, and 0.90, respectively. Furthermore, the robustness and stability of the method have been illustrated when subjected to different levels of noise and image resolutions. The proposed approach offers a rapid, accurate, and dependable diagnosis of brain tumors and holds potential for application in other medical image analysis tasks.

A convolutional neural network prediction model for aviation nitrogen oxides emissions throughout all flight phases

In recent years, there has been an increasing amount of research on nitrogen oxides (NOx) emissions, and the environmental impact of aviation NOx emissions at cruising altitudes has received widespread attention. NOx may play a crucial role in altering the composition of the atmosphere, particularly regarding ozone formation in the upper troposphere. At present, the ground emission database based on the landing and takeoff (LTO) cycle is more comprehensive, while high-altitude emission data is scarce due to the prohibitively high cost and the inevitable measurement uncertainty associated with in-flight sampling. Therefore, it is necessary to establish a comprehensive NOx emission database for the entire flight envelope, encompassing both ground and cruise phases. This will enable a thorough assessment of the impact of aviation NOx emissions on climate and air quality. In this study, a prediction model has been developed via convolutional neural network (CNN) technology. This model can predict the ground and cruise NOx emission index for turbofan engines and mixed turbofan engines fueled by either conventional aviation kerosene or sustainable aviation fuels (SAFs). The model utilizes data from the engine emission database (EEDB) released by the International Civil Aviation Organization (ICAO) and results obtained from several in-situ emission measurements conducted during ground and cruise phases. The model has been validated by comparing measured and predicted data, and the results demonstrate its high prediction accuracy for both the ground (R2 > 0.95) and cruise phases (R2 > 0.9). This surpasses traditional prediction models that rely on fuel flow rate, such as the Boeing Fuel Flow Method 2 (BFFM2). Furthermore, the model can predict NOx emissions from aircrafts burning SAFs with satisfactory accuracy, facilitating the development of a more complete and accurate aviation NOx emission inventory, which can serve as a basis for aviation environmental and climatic research. SYNOPSIS: The utilization of the ANOEPM-CNN offers a foundation for establishing more precise emission inventories, thereby reducing inaccuracies in assessing the impact of aviation NOx emissions on climate and air quality.

Convolutional neural network algorithm trained on lumbar spine radiographs to predict outcomes of transforaminal epidural steroid injection for lumbosacral radicular pain from spinal stenosis

Little is known about the therapeutic outcomes of transforaminal epidural steroid injection (TFESI) in patients with lumbosacral radicular pain due to lumbar spinal stenosis (LSS). Using lumbar spine radiographs as input data, we trained a convolutional neural network (CNN) to predict therapeutic outcomes after lumbar TFESI in patients with lumbosacral radicular pain caused by LSS. We retrospectively recruited 193 patients for this study. The lumbar spine radiographs included anteroposterior, lateral, and bilateral (left and right) oblique views. We cut each lumbar spine radiograph image into a square shape that included the vertebra corresponding to the level at which the TFESI was performed and the vertebrae juxta below and above that level. Output data were divided into "favorable outcome" (≥ 50% reduction in the numeric rating scale [NRS] score at 2 months post-TFESI) and "poor outcome" (< 50% reduction in the NRS score at 2 months post-TFESI). Using these input and output data, we developed a CNN model for predicting TFESI outcomes. The area under the curve of our model was 0.920. Its accuracy was 87.2%. Our CNN model has an excellent capacity for predicting therapeutic outcomes after lumbar TFESI in patients with lumbosacral radicular pain induced by LSS.

Improving Breast Cancer Detection with Convolutional Neural Networks and Modified ResNet Architecture

BACKGROUND

The pathogenesis of breast cancer is characterized by dysregulated cell proliferation, leading to the formation of a neoplastic mass. Conventional methodologies for analyzing carcinomatous distal areas within whole-slide images (WSIs) tissue regions may lack comprehensive insights.

PURPOSE

This study aims to introduce an innovative methodology based on convolutional neural networks (CNN), specifically employing a CNN Modified ResNet architecture for breast cancer detection. The research seeks to address the limitations of existing approaches and provide a robust solution for the comprehensive analysis of tissue regions.

METHODS

The dataset utilized in this study comprises approximately 275,000 RGB image patches, each standardized at 50x50 pixels. The CNN Modified ResNet architecture is implemented, and a comparative evaluation against diverse architectures is conducted. Rigorous validation tests employing established performance metrics are carried out to assess the proposed methodology.

RESULTS

The proposed architecture achieves a notable 89% accuracy in breast cancer detection, surpassing alternative methods by 2%. The results signify the efficacy and superiority of the CNN Modified ResNet model in analyzing carcinomatous distal areas within WSIs tissue regions.

CONCLUSION

In conclusion, this study demonstrates the potential of the CNN Modified ResNet architecture as an effective tool for breast cancer detection. The enhanced accuracy and comprehensive analysis capabilities make it a promising approach for advancing the understanding of neoplastic masses in WSIs tissue regions. Further research and validation could solidify its role in clinical applications and diagnostic procedures.

Intelligent condition monitoring with CNN and signal enhancement for undersampled signals

High-frequency signals like vibration and acoustic emission are crucial for condition monitoring, but their high sampling rates challenge data acquisition, especially for online monitoring. Our research developed a novel method for condition identification in undersampled signals using a modified convolutional neural network integrated with a signal enhancement approach. A frequency-domain filtering is applied to suppress similar sidebands and obtain more discriminative features of different conditions, followed by an interpolation-based upsampling in the time domain to restore the signal length and strengthen the low-frequency harmonic information. Enhanced signals are converted into two-dimensional grayscale images for neural network analysis. Tested on bearing datasets and real-world data from regenerative thermal oxidizer lift valve leakage, our method effectively extracts features from low-frequency signals, achieving over 95% fault identification accuracy.

An efficient and accurate multi-level cascaded recurrent network for stereo matching

With the advent of Transformer-based convolutional neural networks, stereo matching algorithms have achieved state-of-the-art accuracy in disparity estimation. Nevertheless, this method requires much model inference time, which is the main reason limiting its application in many vision tasks and robots. Facing the trade-off problem between accuracy and efficiency, this paper proposes an efficient and accurate multi-level cascaded recurrent network, LMCR-Stereo. To recover the detailed information of stereo images more accurately, we first design a multi-level network to update the difference values in a coarse-to-fine recurrent iterative manner. Then, we propose a new pair of slow-fast multi-stage superposition inference structures to accommodate the differences between different scene data. Besides, to ensure better disparity estimation accuracy with faster model inference speed, we introduce a pair of adaptive and lightweight group correlation layers to reduce the impact of erroneous rectification and significantly improve model inference speed. The experimental results show that the proposed approach achieves a competitive disparity estimation accuracy with a faster model inference speed than the current state-of-the-art methods. Notably, the model inference speed of the proposed approach is improved by 46.0% and 50.4% in the SceneFlow test set and Middlebury benchmark, respectively.

Registration of multimodal bone images based on edge similarity metaheuristic

OBJECTIVE

Blurry medical images affect the accuracy and efficiency of multimodal image registration, whose existing methods require further improvement.

METHODS

We propose an edge-based similarity registration method optimised for multimodal medical images, especially bone images, by a balance optimiser. First, we use a GPU (graphics processing unit) rendering simulation to convert computed tomography data into digitally reconstructed radiographs. Second, we introduce the improved cascaded edge network (ICENet), a convolutional neural network that extracts edge information of blurry medical images. Then, the bilateral Gaussian-weighted similarity of pairs of X-ray images and digitally reconstructed radiographs is measured. The a balanced optimiser is iteratively applied to finally estimate the best pose to perform image registration.

RESULTS

Experimental results show that, on average, the proposed method with ICENet outperforms other edge detection networks by 20%, 12%, 18.83%, and 11.93% in the overall Dice similarity, overall intersection over union, peak signal-to-noise ratio, and structural similarity index, respectively, with a registration success rate up to 90% and average reduction of 220% in registration time.

CONCLUSION

The proposed method with ICENet can achieve a high registration success rate even for blurry medical images, and its efficiency and robustness are higher than those of existing methods.

SIGNIFICANCE

Our proposal may be suitable for supporting medical diagnosis, radiation therapy, image-guided surgery, and other clinical applications.

Development and prospective clinical validation of a convolutional neural network for automated detection and segmentation of focal cortical dysplasias

PURPOSE

Focal cortical dysplasias (FCDs) are a leading cause of drug-resistant epilepsy. Early detection and resection of FCDs have favorable prognostic implications for postoperative seizure freedom. Despite advancements in imaging methods, FCD detection remains challenging. House et al. (2021) introduced a convolutional neural network (CNN) for automated FCD detection and segmentation, achieving a sensitivity of 77.8%. However, its clinical applicability was limited due to a low specificity of 5.5%. The objective of this study was to improve the CNN's performance through data-driven training and algorithm optimization, followed by a prospective validation on daily-routine MRIs.

MATERIAL AND METHODS

A dataset of 300 3 T MRIs from daily clinical practice, including 3D T1 and FLAIR sequences, was prospectively compiled. The MRIs were visually evaluated by two neuroradiologists and underwent morphometric assessment by two epileptologists. The dataset included 30 FCD cases (11 female, mean age: 28.1 ± 10.1 years) and a control group of 150 normal cases (97 female, mean age: 32.8 ± 14.9 years), along with 120 non-FCD pathological cases (64 female, mean age: 38.4 ± 18.4 years). The dataset was divided into three subsets, each analyzed by the CNN. Subsequently, the CNN underwent a two-phase-training process, incorporating subset MRIs and expert-labeled FCD maps. This training employed both classical and continual learning techniques. The CNN's performance was validated by comparing the baseline model with the trained models at two training levels.

RESULTS

In prospective validation, the best model trained using continual learning achieved a sensitivity of 90.0%, specificity of 70.0%, and accuracy of 72.0%, with an average of 0.41 false positive clusters detected per MRI. For FCD segmentation, an average Dice coefficient of 0.56 was attained. The model's performance improved in each training phase while maintaining a high level of sensitivity. Continual learning outperformed classical learning in this regard.

CONCLUSIONS

Our study presents a promising CNN for FCD detection and segmentation, exhibiting both high sensitivity and specificity. Furthermore, the model demonstrates continuous improvement with the inclusion of more clinical MRI data. We consider our CNN a valuable tool for automated, examiner-independent FCD detection in daily clinical practice, potentially addressing the underutilization of epilepsy surgery in drug-resistant focal epilepsy and thereby improving patient outcomes.

Establishing a reference focal plane using convolutional neural networks and beads for brightfield imaging

Repeatability of measurements from image analytics is difficult, due to the heterogeneity and complexity of cell samples, exact microscope stage positioning, and slide thickness. We present a method to define and use a reference focal plane that provides repeatable measurements with very high accuracy, by relying on control beads as reference material and a convolutional neural network focused on the control bead images. Previously we defined a reference effective focal plane (REFP) based on the image gradient of bead edges and three specific bead image features. This paper both generalizes and improves on this previous work. First, we refine the definition of the REFP by fitting a cubic spline to describe the relationship between the distance from a bead's center and pixel intensity and by sharing information across experiments, exposures, and fields of view. Second, we remove our reliance on image features that behave differently from one instrument to another. Instead, we apply a convolutional regression neural network (ResNet 18) trained on cropped bead images that is generalizable to multiple microscopes. Our ResNet 18 network predicts the location of the REFP with only a single inferenced image acquisition that can be taken across a wide range of focal planes and exposure times. We illustrate the different strategies and hyperparameter optimization of the ResNet 18 to achieve a high prediction accuracy with an uncertainty for every image tested coming within the microscope repeatability measure of 7.5 µm from the desired focal plane. We demonstrate the generalizability of this methodology by applying it to two different optical systems and show that this level of accuracy can be achieved using only 6 beads per image.

Automated identification of uncertain cases in deep learning-based classification of dopamine transporter SPECT to improve clinical utility and acceptance

PURPOSE

Deep convolutional neural networks (CNN) are promising for automatic classification of dopamine transporter (DAT)-SPECT images. Reporting the certainty of CNN-based decisions is highly desired to flag cases that might be misclassified and, therefore, require particularly careful inspection by the user. The aim of the current study was to design and validate a CNN-based system for the identification of uncertain cases.

METHODS

A network ensemble (NE) combining five CNNs was trained for binary classification of [123I]FP-CIT DAT-SPECT images as "normal" or "neurodegeneration-typical reduction" with high accuracy (NE for classification, NEfC). An uncertainty detection module (UDM) was obtained by combining two additional NE, one trained for detection of "reduced" DAT-SPECT with high sensitivity, the other with high specificity. A case was considered "uncertain" if the "high sensitivity" NE and the "high specificity" NE disagreed. An internal "development" dataset of 1740 clinical DAT-SPECT images was used for training (n = 1250) and testing (n = 490). Two independent datasets with different image characteristics were used for testing only (n = 640, 645). Three established approaches for uncertainty detection were used for comparison (sigmoid, dropout, model averaging).

RESULTS

In the test data from the development dataset, the NEfC achieved 98.0% accuracy. 4.3% of all test cases were flagged as "uncertain" by the UDM: 2.5% of the correctly classified cases and 90% of the misclassified cases. NEfC accuracy among "certain" cases was 99.8%. The three comparison methods were less effective in labelling misclassified cases as "uncertain" (40-80%). These findings were confirmed in both additional test datasets.

CONCLUSION

The UDM allows reliable identification of uncertain [123I]FP-CIT SPECT with high risk of misclassification. We recommend that automatic classification of [123I]FP-CIT SPECT images is combined with an UDM to improve clinical utility and acceptance. The proposed UDM method ("high sensitivity versus high specificity") might be useful also for DAT imaging with other ligands and for other binary classification tasks.

Accuracy of automated segmentation and volumetry of acute intracerebral hemorrhage following minimally invasive surgery using a patch-based convolutional neural network in a small dataset

PURPOSE

In cases of acute intracerebral hemorrhage (ICH) volume estimation is of prognostic and therapeutic value following minimally invasive surgery (MIS). The ABC/2 method is widely used, but suffers from inaccuracies and is time consuming. Supervised machine learning using convolutional neural networks (CNN), trained on large datasets, is suitable for segmentation tasks in medical imaging. Our objective was to develop a CNN based machine learning model for the segmentation of ICH and of the drain and volumetry of ICH following MIS of acute supratentorial ICH on a relatively small dataset.

METHODS

Ninety two scans were assigned to training (n = 29 scans), validation (n = 4 scans) and testing (n = 59 scans) datasets. The mean age (SD) was 70 (± 13.56) years. Male patients were 36. A hierarchical, patch-based CNN for segmentation of ICH and drain was trained. Volume of ICH was calculated from the segmentation mask.

RESULTS

The best performing model achieved a Dice similarity coefficient of 0.86 and 0.91 for the ICH and drain respectively. Automated ICH volumetry yielded high agreement with ground truth (Intraclass correlation coefficient = 0.94 [95% CI: 0.91, 0.97]). Average difference in the ICH volume was 1.33 mL.

CONCLUSION

Using a relatively small dataset, originating from different CT-scanners and with heterogeneous voxel dimensions, we applied a patch-based CNN framework and successfully developed a machine learning model, which accurately segments the intracerebral hemorrhage (ICH) and the drains. This provides automated and accurate volumetry of the bleeding in acute ICH treated with minimally invasive surgery.

CorNet: Autonomous feature learning in raw Corvis ST data for keratoconus diagnosis via residual CNN approach

PURPOSE

To ascertain whether the integration of raw Corvis ST data with an end-to-end CNN can enhance the diagnosis of keratoconus (KC).

METHOD

The Corvis ST is a non-contact device for in vivo measurement of corneal biomechanics. The CorNet was trained and validated on a dataset consisting of 1786 Corvis ST raw data from 1112 normal eyes and 674 KC eyes. Each raw data consists of the anterior and posterior corneal surface elevation during air-puff induced dynamic deformation. The architecture of CorNet utilizes four ResNet-inspired convolutional structures that employ 1 × 1 convolution in identity mapping. Gradient-weighted Class Activation Mapping (Grad-CAM) was adopted to visualize the attention allocation to diagnostic areas. Discriminative performance was assessed using metrics including the AUC of ROC curve, sensitivity, specificity, precision, accuracy, and F1 score.

RESULTS

CorNet demonstrated outstanding performance in distinguishing KC from normal eyes, achieving an AUC of 0.971 (sensitivity: 92.49%, specificity: 91.54%) in the validation set, outperforming the best existing Corvis ST parameters, namely the Corvis Biomechanical Index (CBI) with an AUC of 0.947, and its updated version for Chinese populations (cCBI) with an AUC of 0.963. Though the ROC curve analysis showed no significant difference between CorNet and cCBI (p = 0.295), it indicated a notable difference between CorNet and CBI (p = 0.011). The Grad-CAM visualizations highlighted the significance of corneal deformation data during the loading phase rather than the unloading phase for KC diagnosis.

CONCLUSION

This study proposed an end-to-end CNN approach utilizing raw biomechanical data by Corvis ST for KC detection, showing effectiveness comparable to or surpassing existing parameters provided by Corvis ST. The CorNet, autonomously learning comprehensive temporal and spatial features, demonstrated a promising performance for advancing KC diagnosis in ophthalmology.

A deep learning based holistic diagnosis system for immunohistochemistry interpretation and molecular subtyping

BACKGROUND

Breast cancer in different molecular subtypes, which is determined by the overexpression rates of human epidermal growth factor receptor 2 (HER2), estrogen receptor (ER), progesterone receptor (PR), and Ki67, exhibit distinct symptom characteristics and sensitivity to different treatment. The immunohistochemical method, one of the most common detecting tools for tumour markers, is heavily relied on artificial judgment and in clinical practice, with an inherent limitation in interpreting stability and operating efficiency. Here, a holistic intelligent breast tumour diagnosis system has been developed for tumour-markeromic analysis, combining the automatic interpretation and clinical suggestion.

METHODS

The holistic intelligent breast tumour diagnosis system included two main modules. The interpreting modules were constructed based on convolutional neural network, for comprehensively extracting and analyzing the multi-features of immunostaining. Referring to the clinical classification criteria, the interpreting results were encoded in a low-dimensional feature representation in the subtyping module, to efficiently output a holistic detecting result of the critical tumour-markeromic with diagnosis suggestions on molecular subtypes.

RESULTS

The overexpression rates of HER2, ER, PR, and Ki67, as well as an effective determination of molecular subtypes were successfully obtained by this diagnosis system, with an average sensitivity of 97.6 % and an average specificity of 96.1 %, among those, the sensitivity and specificity for interpreting HER2 were up to 99.8 % and 96.9 %.

CONCLUSION

The holistic intelligent breast tumour diagnosis system shows improved performance in the interpretation of immunohistochemical images over pathologist-level, which can be expected to overcome the limitations of conventional manual interpretation in efficiency, precision, and repeatability.

CSAP-UNet: Convolution and self-attention paralleling network for medical image segmentation with edge enhancement

Convolution operation is performed within a local window of the input image. Therefore, convolutional neural network (CNN) is skilled in obtaining local information. Meanwhile, the self-attention (SA) mechanism extracts features by calculating the correlation between tokens from all positions in the image, which has advantage in obtaining global information. Therefore, the two modules can complement each other to improve feature extraction ability. An effective fusion method is a problem worthy of further study. In this paper, we propose a CNN and SA paralleling network CSAP-UNet with U-Net as backbone. The encoder consists of two parallel branches of CNN and Transformer to extract the feature from the input image, which takes into account both the global dependencies and the local information. Because medical images come from certain frequency bands within the spectrum, their color channels are not as uniform as natural images. Meanwhile, medical segmentation pays more attention to lesion regions in the image. Attention fusion module (AFM) integrates channel attention and spatial attention in series to fuse the output features of the two branches. The medical image segmentation task is essentially to locate the boundary of the object in the image. The boundary enhancement module (BEM) is designed in the shallow layer of the proposed network to focus more specifically on pixel-level edge details. Experimental results on three public datasets validate that CSAP-UNet outperforms state-of-the-art networks, particularly on the ISIC 2017 dataset. The cross-dataset evaluation on Kvasir and CVC-ClinicDB shows that CSAP-UNet has strong generalization ability. Ablation experiments also indicate the effectiveness of the designed modules. The code for training and test is available at https://github.com/zhouzhou1201/CSAP-UNet.git.

Automated entry of paper-based patient-reported outcomes: Applying deep learning to the Japanese orthopaedic association back pain evaluation questionnaire

BACKGROUND

Health-related patient-reported outcomes (HR-PROs) are crucial for assessing the quality of life among individuals experiencing low back pain. However, manual data entry from paper forms, while convenient for patients, imposes a considerable tallying burden on collectors. In this study, we developed a deep learning (DL) model capable of automatically reading these paper forms.

METHODS

We employed the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire, a globally recognized assessment tool for low back pain. The questionnaire comprised 25 low back pain-related multiple-choice questions and three pain-related visual analog scales (VASs). We collected 1305 forms from an academic medical center as the training set, and 483 forms from a community medical center as the test set. The performance of our DL model for multiple-choice questions was evaluated using accuracy as a categorical classification task. The performance for VASs was evaluated using the correlation coefficient and absolute error as regression tasks.

RESULT

In external validation, the mean accuracy of the categorical questions was 0.997. When outputs for categorical questions with low probability (threshold: 0.9996) were excluded, the accuracy reached 1.000 for the remaining 65 % of questions. Regarding the VASs, the average of the correlation coefficients was 0.989, with the mean absolute error being 0.25.

CONCLUSION

Our DL model demonstrated remarkable accuracy and correlation coefficients when automatic reading paper-based HR-PROs during external validation.

Automatic liver segmentation and assessment of liver fibrosis using deep learning with MR T1-weighted images in rats

OBJECTIVES

To validate the performance of nnU-Net in segmentation and CNN in classification for liver fibrosis using T1-weighted images.

MATERIALS AND METHODS

In this prospective study, animal models of liver fibrosis were induced by injecting subcutaneously a mixture of Carbon tetrachloride and olive oil. A total of 99 male Wistar rats were successfully induced and underwent MR scanning with no contrast agent to get T1-weighted images. The regions of interest (ROIs) of the whole liver were delineated layer by layer along the liver edge by 3D Slicer. For segmentation task, all T1-weighted images were randomly divided into training and test cohorts in a ratio of 7:3. For classification, images containing the hepatic maximum diameter of every rat were selected and 80% images of no liver fibrosis (NLF), early liver fibrosis (ELF) and progressive liver fibrosis (PLF) stages were randomly selected for training, while the rest were used for testing. Liver segmentation was performed by the nnU-Net model. The convolutional neural network (CNN) was used for classification task of liver fibrosis stages. The Dice similarity coefficient was used to evaluate the segmentation performance of nnU-Net. Confusion matrix, ROC curve and accuracy were used to show the classification performance of CNN.

RESULTS

A total of 2628 images were obtained from 99 Wistar rats by MR scanning. For liver segmentation by nnU-Net, the Dice similarity coefficient in the test set was 0.8477. The accuracies of CNN in staging NLF, ELF and PLF were 0.73, 0.89 and 0.84, respectively. The AUCs were 0.76, 0.88 and 0.79, respectively.

CONCLUSION

The nnU-Net architecture is of high accuracy for liver segmentation and CNN for assessment of liver fibrosis with T1-weighted images.

Influence of intraoral scanning coverage on the accuracy of digital implant impressions - An in vitro study

OBJECTIVES

To evaluate the influence of intraoral scanning coverage (IOSC) on digital implant impression accuracy in various partially edentulous situations and predict the optimal IOSC.

METHODS

Five types of resin models were fabricated, each simulating single or multiple tooth loss scenarios with inserted implants and scan bodies. IOSC was subgrouped to cover two, four, six, eight, ten, and twelve teeth, as well as full arch. Each group underwent ten scans. A desktop scanner served as the reference. Accuracy was evaluated by measuring the Root mean square error (RMSE) values of scan bodies. A convolutional neural network (CNN) was trained to predict the optimal IOSC with different edentulous situations. Statistical analysis was performed using one-way ANOVA and Tukey's test.

RESULTS

For single-tooth-missing situations, in anterior sites, significantly better accuracy was observed in groups with IOSC ranging from four teeth to full arch (p < 0.05). In premolar sites, IOSC spanning four to six teeth were more accurate (p < 0.05), while in molar sites, groups with IOSC encompassing two to eight teeth exhibited better accuracy (p < 0.05). For multiple-teeth-missing situations, IOSC covering four, six, and eight teeth, as well as full arch showed better accuracy in anterior gaps (p < 0.05). In posterior gaps, IOSC of two, four, six or eight teeth were more accurate (p < 0.05). The CNN predicted distinct optimal IOSC for different edentulous scenarios.

CONCLUSIONS

Implant impression accuracy can be significantly impacted by IOSC in different partially edentulous situations. The selection of IOSC should be customized to the specific dentition defect condition.

CLINICAL SIGNIFICANCE

The number of teeth scanned can significantly affect digital implant impression accuracy. For missing single or four anterior teeth, scan at least four or six neighboring teeth is acceptable. In lateral cases, two neighboring teeth may suffice, but extending over ten teeth, including contralateral side, might deteriorate the scan.

Detection of caries around restorations on bitewings using deep learning

OBJECTIVE

Secondary caries lesions adjacent to restorations, a leading cause of restoration failure, require accurate diagnostic methods to ensure an optimal treatment outcome. Traditional diagnostic strategies rely on visual inspection complemented by radiographs. Recent advancements in artificial intelligence (AI), particularly deep learning, provide potential improvements in caries detection. This study aimed to develop a convolutional neural network (CNN)-based algorithm for detecting primary caries and secondary caries around restorations using bitewings.

METHODS

Clinical data from 7 general dental practices in the Netherlands, comprising 425 bitewings of 383 patients, were utilized. The study used the Mask-RCNN architecture, for instance, segmentation, supported by the Swin Transformer backbone. After data augmentation, model training was performed through a ten-fold cross-validation. The diagnostic accuracy of the algorithm was evaluated by calculating the area under the Free-Response Receiver Operating Characteristics curve, sensitivity, precision, and F1 scores.

RESULTS

The model achieved areas under FROC curves of 0.806 and 0.804, and F1-scores of 0.689 and 0.719 for primary and secondary caries detection, respectively.

CONCLUSION

An accurate CNN-based automated system was developed to detect primary and secondary caries lesions on bitewings, highlighting a significant advancement in automated caries diagnostics.

CLINICAL SIGNIFICANCE

An accurate algorithm that integrates the detection of both primary and secondary caries will permit the development of automated systems to aid clinicians in their daily clinical practice.

LSTM-CNN: An efficient diagnostic network for Parkinson's disease utilizing dynamic handwriting analysis

BACKGROUND AND OBJECTIVES

Dynamic handwriting analysis, due to its noninvasive and readily accessible nature, has emerged as a vital adjunctive method for the early diagnosis of Parkinson's disease (PD). An essential step involves analysing subtle variations in signals to quantify PD dysgraphia. Although previous studies have explored extracting features from the overall signal, they may ignore the potential importance of local signal segments. In this study, we propose a lightweight network architecture to analyse dynamic handwriting signal segments of patients and present visual diagnostic results, providing an efficient diagnostic method.

METHODS

To analyse subtle variations in handwriting, we investigate time-dependent patterns in local representation of handwriting signals. Specifically, we segment the handwriting signal into fixed-length sequential segments and design a compact one-dimensional (1D) hybrid network to extract discriminative temporal features for classifying each local segment. Finally, the category of the handwriting signal is fully diagnosed through a majority voting scheme.

RESULTS

The proposed method achieves impressive diagnostic performance on the new DraWritePD dataset (with an accuracy of 96.2%, sensitivity of 94.5% and specificity of 97.3%) and the well-established PaHaW dataset (with an accuracy of 90.7%, sensitivity of 94.3% and specificity of 87.5%). Moreover, the network architecture stands out for its excellent lightweight design, occupying a mere 0.084M parameters, with only 0.59M floating-point operations. It also exhibits nearly real-time CPU inference performance, with the inference time for a single handwriting signal ranging from 0.106 to 0.220 s.

CONCLUSIONS

We present a series of experiments with extensive analysis, which systematically demonstrate the effectiveness and efficiency of the proposed method in quantifying dysgraphia for a precise diagnosis of PD.

Using neural networks to autonomously assess adequacy in intraoperative cholangiograms

BACKGROUND

Intraoperative cholangiography (IOC) is a contrast-enhanced X-ray acquired during laparoscopic cholecystectomy. IOC images the biliary tree whereby filling defects, anatomical anomalies and duct injuries can be identified. In Australia, IOC are performed in over 81% of cholecystectomies compared with 20 to 30% internationally (Welfare AIoHa in Australian Atlas of Healthcare Variation, 2017). In this study, we aim to train artificial intelligence (AI) algorithms to interpret anatomy and recognise abnormalities in IOC images. This has potential utility in (a) intraoperative safety mechanisms to limit the risk of missed ductal injury or stone, (b) surgical training and coaching, and (c) auditing of cholangiogram quality.

METHODOLOGY

Semantic segmentation masks were applied to a dataset of 1000 cholangiograms with 10 classes. Classes corresponded to anatomy, filling defects and the cholangiogram catheter instrument. Segmentation masks were applied by a surgical trainee and reviewed by a radiologist. Two convolutional neural networks (CNNs), DeeplabV3+ and U-Net, were trained and validated using 900 (90%) labelled frames. Testing was conducted on 100 (10%) hold-out frames. CNN generated segmentation class masks were compared with ground truth segmentation masks to evaluate performance according to a pixel-wise comparison.

RESULTS

The trained CNNs recognised all classes.. U-Net and DeeplabV3+ achieved a mean F1 of 0.64 and 0.70 respectively in class segmentation, excluding the background class. The presence of individual classes was correctly recognised in over 80% of cases. Given the limited local dataset, these results provide proof of concept in the development of an accurate and clinically useful tool to aid in the interpretation and quality control of intraoperative cholangiograms.

CONCLUSION

Our results demonstrate that a CNN can be trained to identify anatomical structures in IOC images. Future performance can be improved with the use of larger, more diverse training datasets. Implementation of this technology may provide cholangiogram quality control and improve intraoperative detection of ductal injuries or ductal injuries.

V2IED: Dual-view learning framework for detecting events of interictal epileptiform discharges

Interictal epileptiform discharges (IED) as large intermittent electrophysiological events are associated with various severe brain disorders. Automated IED detection has long been a challenging task, and mainstream methods largely focus on singling out IEDs from backgrounds from the perspective of waveform, leaving normal sharp transients/artifacts with similar waveforms almost unattended. An open issue still remains to accurately detect IED events that directly reflect the abnormalities in brain electrophysiological activities, minimizing the interference from irrelevant sharp transients with similar waveforms only. This study then proposes a dual-view learning framework (namely V2IED) to detect IED events from multi-channel EEG via aggregating features from the two phases: (1) Morphological Feature Learning: directly treating the EEG as a sequence with multiple channels, a 1D-CNN (Convolutional Neural Network) is applied to explicitly learning the deep morphological features; and (2) Spatial Feature Learning: viewing the EEG as a 3D tensor embedding channel topology, a CNN captures the spatial features at each sampling point followed by an LSTM (Long Short-Term Memories) to learn the evolution of these features. Experimental results from a public EEG dataset against the state-of-the-art counterparts indicate that: (1) compared with the existing optimal models, V2IED achieves a larger area under the receiver operating characteristic (ROC) curve in detecting IEDs from normal sharp transients with a 5.25% improvement in accuracy; (2) the introduction of spatial features improves performance by 2.4% in accuracy; and (3) V2IED also performs excellently in distinguishing IEDs from background signals especially benign variants.

A deep learning pipeline for automated classification of vocal fold polyps in flexible laryngoscopy

PURPOSE

To develop and validate a deep learning model for distinguishing healthy vocal folds (HVF) and vocal fold polyps (VFP) on laryngoscopy videos, while demonstrating the ability of a previously developed informative frame classifier in facilitating deep learning development.

METHODS

Following retrospective extraction of image frames from 52 HVF and 77 unilateral VFP videos, two researchers manually labeled each frame as informative or uninformative. A previously developed informative frame classifier was used to extract informative frames from the same video set. Both sets of videos were independently divided into training (60%), validation (20%), and test (20%) by patient. Machine-labeled frames were independently verified by two researchers to assess the precision of the informative frame classifier. Two models, pre-trained on ResNet18, were trained to classify frames as containing HVF or VFP. The accuracy of the polyp classifier trained on machine-labeled frames was compared to that of the classifier trained on human-labeled frames. The performance was measured by accuracy and area under the receiver operating characteristic curve (AUROC).

RESULTS

When evaluated on a hold-out test set, the polyp classifier trained on machine-labeled frames achieved an accuracy of 85% and AUROC of 0.84, whereas the classifier trained on human-labeled frames achieved an accuracy of 69% and AUROC of 0.66.

CONCLUSION

An accurate deep learning classifier for vocal fold polyp identification was developed and validated with the assistance of a peer-reviewed informative frame classifier for dataset assembly. The classifier trained on machine-labeled frames demonstrates improved performance compared to the classifier trained on human-labeled frames.

LEVEL OF EVIDENCE: 4

One-shot screening: Utilization of a two-dimensional convolutional neural network for automatic detection of left ventricular hypertrophy using electrocardiograms

BACKGROUND AND OBJECTIVE

Left ventricular hypertrophy (LVH) can impair ejection function and elevate the risk of heart failure. Therefore, early detection through screening is crucial. This study aimed to propose a novel method to enhance LVH detection using 12-lead electrocardiogram (ECG) waveforms with a two-dimensional (2D) convolutional neural network (CNN).

METHODS

Utilizing 42,127 pairs of ECG-transthoracic echocardiogram data, we pre-processed raw data into single-shot images derived from each ECG lead and conducted lead selection to optimize LVH diagnosis. Our proposed one-shot screening method, implemented during pre-processing, enables the superimposition of waveform source data of any length onto a single-frame image, thereby addressing the limitations of the one-dimensional (1D) approach. We developed a deep learning model with a 2D-CNN structure and machine learning models for LVH detection. To assess our method, we also compared our results with conventional ECG criteria and those of a prior study that used a 1D-CNN approach, utilizing the same dataset from the University of Tokyo Hospital for LVH diagnosis.

RESULTS

For LVH detection, the average area under the receiver operating characteristic curve (AUROC) was 0.916 for the 2D-CNN model, which was significantly higher than that obtained using logistic regression and random forest methods, as well as the two conventional ECG criteria (AUROC of 0.766, 0.790, 0.599, and 0.622, respectively). Incorporating additional metadata, such as ECG measurement data, further improved the average AUROC to 0.921. The model's performance remained stable across two different annotation criteria and demonstrated significant superiority over the performance of the 1D-CNN model used in a previous study (AUROC of 0.807).

CONCLUSIONS

This study introduces a robust and computationally efficient method that outperforms 1D-CNN models utilized in previous studies for LVH detection. Our method can transform waveforms of any length into fixed-size images and leverage the selected lead of the ECG, ensuring adaptability in environments with limited computational resources. The proposed method holds promise for integration into clinical practice as a tool for early diagnosis, potentially enhancing patient outcomes by facilitating earlier treatment and management.

Convolutional neural network deep learning model accurately detects rectal cancer in endoanal ultrasounds

BACKGROUND

Imaging is vital for assessing rectal cancer, with endoanal ultrasound (EAUS) being highly accurate in large tertiary medical centers. However, EAUS accuracy drops outside such settings, possibly due to varied examiner experience and fewer examinations. This underscores the need for an AI-based system to enhance accuracy in non-specialized centers. This study aimed to develop and validate deep learning (DL) models to differentiate rectal cancer in standard EAUS images.

METHODS

A transfer learning approach with fine-tuned DL architectures was employed, utilizing a dataset of 294 images. The performance of DL models was assessed through a tenfold cross-validation.

RESULTS

The DL diagnostics model exhibited a sensitivity and accuracy of 0.78 each. In the identification phase, the automatic diagnostic platform achieved an area under the curve performance of 0.85 for diagnosing rectal cancer.

CONCLUSIONS

This research demonstrates the potential of DL models in enhancing rectal cancer detection during EAUS, especially in settings with lower examiner experience. The achieved sensitivity and accuracy suggest the viability of incorporating AI support for improved diagnostic outcomes in non-specialized medical centers.

Boxing behavior recognition based on artificial intelligence convolutional neural network with sports psychology assistant

The purpose of this study is to deeply understand the psychological state of boxers before the competition, and explore an efficient boxing action classification and recognition model supported by artificial intelligence (AI) technology through these psychological characteristics. Firstly, this study systematically measures the key psychological dimensions of boxers, such as anxiety level, self-confidence, team identity, and opponent attitude, through psychological scale survey to obtain detailed psychological data. Then, based on these data, this study innovatively constructs a boxing action classification and recognition model based on BERT fusion 3D-ResNet, which not only comprehensively considers psychological information, but also carefully considers action characteristics to improve the classification accuracy of boxing actions. The performance evaluation shows that the model proposed in this study is significantly superior to the traditional model in terms of loss value, accuracy and F1 value, and the accuracy reaches 96.86%. Therefore, through the comprehensive application of psychology and deep learning, this study successfully constructs a boxing action classification and recognition model that can fully understand the psychological state of boxers, which provides strong support for the psychological training and action classification of boxers.