Proposed Deep Learning Model Research Articles

Deep Learning-based Segmentation of Cervical Posterior Longitudinal Ligament Ossification in CT Images and Assessment of Spinal Cord Compression: A Two-center Study.

This study aims to develop a fully automated, CT-based deep learning（DL） model to segment ossified lesions of the posterior longitudinal ligament (OPLL) and to measure the thickness of the ossified material and calculate the cervical spinal cord compression factor. A total of 307 patients were enrolled, with 260 patients from Shanghai Changzheng Hospital, And 47 patients from the Traditional Chinese Medicine Hospital of Southwest Medical University. CT images were used to manually segment the OPLL by four experienced radiologists. The DL model employing a 3D U-Net framework was developed to segment the OPLLs. The system also measures the thickness of the ossified material at its thickest point and the diameter of the spinal canal at the corresponding level. Segmentation performance was evaluated using the Dice Similarity Coefficient (DSC), Average Surface Distance (ASD), and Intra-Class Correlation (ICC) between ground truth and segmentation volumes. Concordance between the radiologists' and the DL system's measurements of the ossified material thickness, residual spinal canal diameter at maximum compression, and cervical spinal cord compression coefficient was assessed in a randomly selected subset of 30 cases from the training set using ICCs and Bland-Altman plots. The DL system demonstrated average DSC of 0.81, 0.75, and 0.71 for the training, internal validation, and external test sets, respectively. The mean ASD was 1.30 for the training set, 2.35 for the internal validation set, and 2.63 for the external test set. The intraclass correlation coefficient (ICC) values of 0.958 for the thickness of the ossified material and 0.974 for the residual canal diameter measurement. The proposed DL model effectively detects and separates ossification foci in OPLL on CT images. It exhibits comparable performance to radiologists in quantifying spinal cord compression metrics.

Advanced Deep Learning Fusion Model for Early Multi-Classification of Lung and Colon Cancer Using Histopathological Images.

In recent years, the healthcare field has experienced significant advancements. New diagnostic techniques, treatments, and insights into the causes of various diseases have emerged. Despite these progressions, cancer remains a major concern. It is a widespread illness affecting individuals of all ages and leads to one out of every six deaths. Lung and colon cancer alone account for nearly two million fatalities. Though it is rare for lung and colon cancers to co-occur, the spread of cancer cells between these two areas-known as metastasis-is notably high. Early detection of cancer greatly increases survival rates. Currently, histopathological image (HI) diagnosis and appropriate treatment are key methods for reducing cancer mortality and enhancing survival rates. Digital image processing (DIP) and deep learning (DL) algorithms can be employed to analyze the HIs of five different types of lung and colon tissues. Therefore, this paper proposes a refined DL model that integrates feature fusion for the multi-classification of lung and colon cancers. The proposed model incorporates three DL architectures: ResNet-101V2, NASNetMobile, and EfficientNet-B0. Each model has limitations concerning variations in the shape and texture of input images. To address this, the proposed model utilizes a concatenate layer to merge the pre-trained individual feature vectors from ResNet-101V2, NASNetMobile, and EfficientNet-B0 into a single feature vector, which is then fine-tuned. As a result, the proposed DL model achieves high success in multi-classification by leveraging the strengths of all three models to enhance overall accuracy. This model aims to assist pathologists in the early detection of lung and colon cancer with reduced effort, time, and cost. The proposed DL model was evaluated using the LC25000 dataset, which contains colon and lung HIs. The dataset was pre-processed using resizing and normalization techniques. The model was tested and compared with recent DL models, achieving impressive results: 99.8% for precision, 99.8% for recall, 99.8% for F1-score, 99.96% for specificity, and 99.94% for accuracy. Thus, the proposed DL model demonstrates exceptional performance across all classification categories.

Clinical validation of a deep learning-based approach for preoperative decision-making in implant size for total knee arthroplasty.

Orthopedic surgeons use manual measurements, acetate templating, and dedicated software to determine the appropriate implant size for total knee arthroplasty (TKA). This study aimed to use deep learning (DL) to assist in deciding the femoral and tibial implant sizes without manual manipulation and to evaluate the clinical validity of the DL decision by comparing it with conventional manual procedures. Two types of DL were used to detect the femoral and tibial regions using the You Only Look Once algorithm model and to determine the implant size from the detected regions using convolutional neural network. An experienced surgeon predicted the implant size for 234 patient cases using manual procedures, and the DL model also predicted the implant sizes for the same cases. The exact accuracies of the surgeon's template were 61.54% and 68.38% for predicting femoral and tibial implant sizes, respectively. Meanwhile, the proposed DL model reported exact accuracies of 89.32% and 90.60% for femoral and tibial implant sizes, respectively. The accuracy ± 1 levels of the surgeon and proposed DL model were 97.44% and 97.86%, respectively, for the femoral implant size and 98.72% for both the surgeon and proposed DL model for the tibial implant size. The observed differences and higher agreement levels achieved by the proposed DL model demonstrate its potential as a valuable tool in preoperative decision-making for TKA. By providing accurate predictions of implant size, the proposed DL model has the potential to optimize implant selection, leading to improved surgical outcomes.

A Deep Learning Model for Cervical Optical Coherence Tomography Image Classification.

Objectives: Optical coherence tomography (OCT) has recently been used in gynecology to detect cervical lesions in vivo and proven more effective than colposcopy in clinical trials. However, most gynecologists are unfamiliar with this new imaging technique, requiring intelligent computer-aided diagnosis approaches to help them interpret cervical OCT images efficiently. This study aims to (1) develop a clinically-usable deep learning (DL)-based classification model of 3D OCT volumes from cervical tissue and (2) validate the DL model's effectiveness in detecting high-risk cervical lesions, including high-grade squamous intraepithelial lesions and cervical cancer. Method: The proposed DL model, designed based on the convolutional neural network architecture, combines a feature pyramid network (FPN) with texture encoding and deep supervision. We extracted, represent, and fused four-scale texture features to improve classification performance on high-risk local lesions. We also designed an auxiliary classification mechanism based on deep supervision to adjust the weight of each scale in FPN adaptively, enabling low-cost training of the whole model. Results: In the binary classification task detecting positive subjects with high-risk cervical lesions, our DL model achieved an 81.55% (95% CI, 72.70-88.51%) F1-score with 82.35% (95% CI, 69.13-91.60%) sensitivity and 81.48% (95% CI, 68.57-90.75%) specificity on the Renmin dataset, outperforming five experienced medical experts. It also achieved an 84.34% (95% CI, 74.71-91.39%) F1-score with 87.50% (95% CI, 73.20-95.81%) sensitivity and 90.59% (95% CI, 82.29-95.85%) specificity on the Huaxi dataset, comparable to the overall level of the best investigator. Moreover, our DL model provides visual diagnostic evidence of histomorphological and texture features learned in OCT images to assist gynecologists in making clinical decisions quickly. Conclusions: Our DL model holds great promise to be used in cervical lesion screening with OCT efficiently and effectively.

Deep learning based approach: automated gingival inflammation grading model using gingival removal strategy

Gingival inflammation grade serves as a well-established index in periodontitis. The aim of this study was to develop a deep learning network utilizing a novel feature extraction method for the automatic assessment of gingival inflammation. T-distributed Stochastic Neighbor Embedding (t-SNE) was utilized for dimensionality reduction. A convolutional neural network (CNN) model based on DenseNet was developed for the identification and evaluation of gingival inflammation. To enhance the performance of the deep learning (DL) model, a novel teeth removal algorithm was implemented. Additionally, a Grad-CAM + + encoder was applied to generate heatmaps for computer visual attention analysis. The mean Intersection over Union (MIoU) for the identification of gingivitis was 0.727 ± 0.117. The accuracy rates for the five inflammatory degrees were 77.09%, 77.25%, 74.38%, 73.68% and 79.22%. The Area Under the Receiver Operating Characteristic (AUROC) values were 0.83, 0.80, 0.81, 0.81 and 0.84, respectively. The attention ratio towards gingival tissue increased from 37.73% to 62.20%, and within 8 mm of the gingival margin, it rose from 21.11% to 38.23%. On the gingiva, the overall attention ratio increased from 51.82% to 78.21%. The proposed DL model with novel feature extraction method provides high accuracy and sensitivity for identifying and grading gingival inflammation.

Knowledge-Augmented Deep Learning for Segmenting and Detecting Cerebral Aneurysms With CT Angiography: A Multicenter Study.

Background Deep learning (DL) could improve the labor-intensive, challenging processes of diagnosing cerebral aneurysms but requires large multicenter data sets. Purpose To construct a DL model using a multicenter data set for accurate cerebral aneurysm segmentation and detection on CT angiography (CTA) images and to compare its performance with radiology reports. Materials and Methods Consecutive head or head and neck CTA images of suspected unruptured cerebral aneurysms were gathered retrospectively from eight hospitals between February 2018 and October 2021 for model development. An external test set with reference standard digital subtraction angiography (DSA) scans was obtained retrospectively from one of the eight hospitals between February 2022 and February 2023. Radiologists (reference standard) assessed aneurysm segmentation, while model performance was evaluated using the Dice similarity coefficient (DSC). The model's aneurysm detection performance was assessed by sensitivity and comparing areas under the receiver operating characteristic curves (AUCs) between the model and radiology reports in the DSA data set with use of the DeLong test. Results Images from 6060 patients (mean age, 56 years ± 12 [SD]; 3375 [55.7%] female) were included for model development (training: 4342; validation: 1086; and internal test set: 632). Another 118 patients (mean age, 59 years ± 14; 79 [66.9%] female) were included in an external test set to evaluate performance based on DSA. The model achieved a DSC of 0.87 for aneurysm segmentation performance in the internal test set. Using DSA, the model achieved 85.7% (108 of 126 aneurysms [95% CI: 78.1, 90.1]) sensitivity in detecting aneurysms on per-vessel analysis, with no evidence of a difference versus radiology reports (AUC, 0.93 [95% CI: 0.90, 0.95] vs 0.91 [95% CI: 0.87, 0.94]; P = .67). Model processing time from reconstruction to detection was 1.76 minutes ± 0.32 per scan. Conclusion The proposed DL model could accurately segment and detect cerebral aneurysms at CTA with no evidence of a significant difference in diagnostic performance compared with radiology reports. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Payabvash in this issue.

Automatic classification and segmentation of multiclass jaw lesions in cone-beam CT using deep learning.

To develop and validate a modified deep learning (DL) model based on nnU-Net for classifying and segmenting five-class jaw lesions using cone-beam CT (CBCT). A total of 368 CBCT scans (37 168 slices) were used to train a multi-class segmentation model. The data underwent manual annotation by two oral and maxillofacial surgeons (OMSs) to serve as ground truth. Sensitivity, specificity, precision, F1-score, and accuracy were used to evaluate the classification ability of the model and doctors, with or without artificial intelligence assistance. The dice similarity coefficient (DSC), average symmetric surface distance (ASSD), and segmentation time were used to evaluate the segmentation effect of the model. The model achieved the dual task of classifying and segmenting jaw lesions in CBCT. For classification, the sensitivity, specificity, precision, and accuracy of the model were 0.871, 0.974, 0.874, and 0.891, respectively, surpassing oral and maxillofacial radiologists (OMFRs) and OMSs, approaching the specialist. With the model's assistance, the classification performance of OMFRs and OMSs improved, particularly for odontogenic keratocyst (OKC) and ameloblastoma (AM), with F1-score improvements ranging from 6.2% to 12.7%. For segmentation, the DSC was 87.2% and the ASSD was 1.359 mm. The model's average segmentation time was 40 ± 9.9 s, contrasting with 25 ± 7.2 min for OMSs. The proposed DL model accurately and efficiently classified and segmented five classes of jaw lesions using CBCT. In addition, it could assist doctors in improving classification accuracy and segmentation efficiency, particularly in distinguishing confusing lesions (eg, AM and OKC).

Deep learning-aided topology design of metasurfaces for Rayleigh waves

Abstract Metasurfaces can effectively attenuate Rayleigh waves propagating in soil, safeguarding structures from ambient vibrations or earthquakes. However, there remains a lack of efficient approaches for designing metasurfaces that can isolate Rayleigh waves within desired frequency ranges under different site conditions. This study presents a deep learning (DL)-based topology optimization method for designing metasurfaces isolating Rayleigh waves in a target frequency range, which has potential applications for surface wave control. The proposed DL model employs a variational autoencoder to transform high-dimensional and discrete topologies into low-dimensional and continuous latent vectors, reducing the design difficulty. On this basis, the conditional tandem neural network is constructed to optimize latent vectors under different soil conditions, improving the design efficiency and verifying the universality of the method. The design reliability of the DL model is validated through 100 tests with determination coefficients more than 0.99. In addition, the generations of different metasurfaces for the same target are explored, providing designers with more choices. The insulation capabilities of the designed metasurfaces against Metro-induced Rayleigh waves are demonstrated from the time- and frequency-domain responses. The presented DL-aided design method provides a novel insight into the customization of metasurfaces for manipulating Rayleigh waves.

Enhancing Quantitative Precipitation Estimation of NWP Model With Fundamental Meteorological Variables and Transformer Based Deep Learning Model

AbstractQuantitative precipitation forecasting in numerical weather prediction (NWP) models is contingent upon physicals parameterization schemes. However, uncertainties abound due to limited knowledge of the precipitating processes, leading to degraded forecasting skills. In light of this, our study explores the application of a Swin‐Transformer based deep learning (DL) model as a supplementary tool for enhancing the mapping trajectory between the NWP fundamental variables and the most downstream variable precipitation. Constrained by the observational satellite precipitation product from NOAA CPC Morphing Technique (CMORPH), the DL model serves as the post‐processing tool that can better resolve the precipitation patterns compared to solely based on NWP estimation. Compared to the baseline Weather Research and Forecasting simulation, the DL post‐processing effectively extracts features over meteorological variables, leading to improved precipitation skill scores of 21.7%, 60.5%, and 45.5% for light rain, moderate rain, and heavy rain, respectively, on an hourly basis. We also evaluate two case studies under different driven synoptic conditions and show promising results in estimating heavy precipitation during strong convective precipitation events. Overall, the proposed DL model can provide a vital reference for capturing precipitation‐triggering mechanisms and enhancing precipitation forecasting skills. Additionally, we discuss the sensitivities of the fundamental meteorological variables used in this study, training strategies, and performance limitations.

Precise tooth design using deep learning-based templates

ObjectivesIn prosthodontic procedures, traditional computer-aided design (CAD) is often time-consuming and lacks accuracy in shape restoration. In this study, we combined implicit template and deep learning (DL) to construct a precise neural network for personalized tooth defect restoration. MethodsNinety models of right maxillary central incisor (80 for training, 10 for validation) were collected. A DL model named ToothDIT was trained to establish an implicit template and a neural network capable of predicting unique identifications. In the validation stage, teeth in validation set were processed into corner, incisive, and medium defects. The defective teeth were inputted into ToothDIT to predict the unique identification, which actuated the deformation of the implicit template to generate the highly customized template (DIT) for the target tooth. Morphological restorations were executed with templates from template shape library (TSL), average tooth template (ATT), and DIT in Exocad (GmbH, Germany). RMSestimate, width, length, aspect ratio, incisal edge curvature, incisive end retraction, and guiding inclination were introduced to assess the restorative accuracy. Statistical analysis was conducted using two-way ANOVA and paired t-test for overall and detailed differences. ResultsDIT displayed significantly smaller RMSestimate than TSL and ATT. In 2D detailed analysis, DIT exhibited significantly less deviations from the natural teeth compared to TSL and ATT. ConclusionThe proposed DL model successfully reconstructed the morphology of anterior teeth with various degrees of defects and achieved satisfactory accuracy. This approach provides a more reliable reference for prostheses design, resulting in enhanced accuracy in morphological restoration. Clinical significanceThis DL model holds promise in assisting dentists and technicians in obtaining morphology templates that closely resemble the original shape of the defective teeth. These customized templates serve as a foundation for enhancing the efficiency and precision of digital restorative design for defective teeth.

Surface soil moisture retrieval based on transfer learning using SAR data on a local scale

ABSTRACT Retrieving surface soil moisture on a local scale using Synthetic Aperture Radar (SAR) data and Deep Learning (DL) models necessitates a substantial volume of data, which may not be available in all scenarios. In this study, the application of transfer learning was introduced as a novel approach to address the scarcity of training samples for DL models in the context of soil moisture retrieval. The proposed DL model was initially trained using International Soil Moisture Network (ISMN) data, followed by a fine-tuned process on a local scale using field trip data from an agricultural area in Karaj, Iran. The proposed DL model was compared against Random Forest Regressor (RFR), Support Vector Regression (SVR), and Multi-Layer Perceptron (MLP) based on Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and R2 indicators. All models underwent hyperparameter-tunning, and their performance was evaluated using 8-fold cross-validation (CV) and various combinations of inputs. The proposed DL model outperformed other models on a local scale achieving an RMSE, MAE, and R2 of 2.42 vol%, 1.66 vol%, and 0.90, respectively. The MLP model also exhibited good performance with an RMSE, MAE, and R2 of 2.84 vol%, 2.04 vol%, and 0.88 compared to the RFR with 2.83 vol%, 2.20 vol%, and 0.86, respectively. Additionally, the SVR yielded an RMSE, MAE, and R2 of 3.71 vol%, 3.05 vol%, and 0.78. However, the RMSE, MAE, and R2 of the MLP and the proposed DL model without using transfer learning deteriorated by around 18%, 32%, and 34%, respectively.

Prediction of extraction difficulty for impacted maxillary third molars with deep learning approach

ObjectiveThe aim of this study is to determine if a deep learning (DL) model can predict the surgical difficulty for impacted maxillary third molar tooth using panoramic images before surgery. Materials and MethodsThe dataset consists of 708 panoramic radiographs of the patients who applied to the Oral and Maxillofacial Surgery Clinic for various reasons. Each maxillary third molar difficulty was scored based on dept (V), angulation (H), relation with maxillary sinus (S), and relation with ramus (R) on panoramic images. The YoloV5x architecture was used to perform automatic segmentation and classification. To prevent re-testing of images, participate in the training, the data set was subdivided as: 80 % training, 10 % validation, and 10 % test group. ResultsImpacted Upper Third Molar Segmentation model showed best success on sensitivity, precision and F1 score with 0,9705, 0,9428 and 0,9565, respectively. S-model had a lesser sensitivity, precision and F1 score than the other models with 0,8974, 0,6194, 0,7329, respectively. ConclusionThe results showed that the proposed DL model could be effective for predicting the surgical difficulty of an impacted maxillary third molar tooth using panoramic radiographs and this approach might help as a decision support mechanism for the clinicians in peri-surgical period.

A comparison between centralized and asynchronous federated learning approaches for survival outcome prediction using clinical and PET data from non-small cell lung cancer patients

Background and Objective: Survival analysis plays an essential role in the medical field for optimal treatment decision-making. Recently, survival analysis based on the deep learning (DL) approach has been proposed and is demonstrating promising results. However, developing an ideal prediction model requires integrating large datasets across multiple institutions, which poses challenges concerning medical data privacy.Methods: In this paper, we propose FedSurv, an asynchronous federated learning (FL) framework designed to predict survival time using clinical information and positron emission tomography (PET)-based features. This study used two datasets: a public radiogenic dataset of non-small cell lung cancer (NSCLC) from the Cancer Imaging Archive (RNSCLC), and an in-house dataset from the Chonnam National University Hwasun Hospital (CNUHH) in South Korea, consisting of clinical risk factors and F-18 fluorodeoxyglucose (FDG) PET images in NSCLC patients. Initially, each dataset was divided into multiple clients according to histological attributes, and each client was trained using the proposed DL model to predict individual survival time. The FL framework collected weights and parameters from the clients, which were then incorporated into the global model. Finally, the global model aggregated all weights and parameters and redistributed the updated model weights to each client. We evaluated different frameworks including single-client-based approach, centralized learning and FL.Results: We evaluated our method on two independent datasets. First, on the RNSCLC dataset, the mean absolute error (MAE) was 490.80±22.95 d and the C-Index was 0.69±0.01. Second, on the CNUHH dataset, the MAE was 494.25±40.16 d and the C-Index was 0.71±0.01. The FL approach achieved centralized method performance in PET-based survival time prediction and outperformed single-client-based approaches.Conclusions: Our results demonstrated the feasibility and effectiveness of employing FL for individual survival prediction in NSCLC patients, using clinical information and PET-based features.

ODL-BCI: Optimal deep learning model for brain-computer interface to classify students confusion via hyperparameter tuning

Brain-computer interface (BCI) research has gained increasing attention in educational contexts, offering the potential to monitor and enhance students’ cognitive states. Real-time classification of students’ confusion levels using electroencephalogram (EEG) data presents a significant challenge in this domain. Since real-time EEG data is dynamic and highly dimensional, current approaches have some limitations for predicting mental states based on this data. This paper introduces an optimal deep learning (DL) model for the BCI, ODL-BCI, optimized through hyperparameter tuning techniques to address the limitations of classifying students’ confusion in real time. Leveraging the “confused student EEG brainwave” dataset, we employ Bayesian optimization to fine-tune hyperparameters of the proposed DL model. The model architecture comprises input and output layers, with several hidden layers whose nodes, activation functions, and learning rates are determined utilizing selected hyperparameters. We evaluate and compare the proposed model with some state-of-the-art methods and standard machine learning (ML) classifiers, including Decision Tree, AdaBoost, Bagging, MLP, Naïve Bayes, Random Forest, SVM, and XG Boost, on the EEG confusion dataset. Our experimental results demonstrate the superiority of the optimized DL model, ODL-BCI. It boosts the accuracy between 4% and 9% over the current approaches, outperforming all other classifiers in the process. The ODL-BCI implementation source codes can be accessed by anyone at https://github.com/MdOchiuddinMiah/ODL-BCI.

High-precision retinal blood vessel segmentation based on a multi-stage and dual-channel deep learning network.

Objective.The high-precision segmentation of retinal vessels in fundus images is important for the early diagnosis of ophthalmic diseases. However, the extraction for microvessels is challenging due to their characteristics of low contrast and high structural complexity. Although some works have been developed to improve the segmentation ability in thin vessels, they have only been successful in recognizing small vessels with relatively high contrast.Approach.Therefore, we develop a deep learning (DL) framework with a multi-stage and dual-channel network model (MSDC_NET) to further improve the thin-vessel segmentation with low contrast. Specifically, an adaptive image enhancement strategy combining multiple preprocessing and the DL method is firstly proposed to elevate the contrast of thin vessels; then, a two-channel model with multi-scale perception is developed to implement whole- and thin-vessel segmentation; and finally, a series of post-processing operations are designed to extract more small vessels in the predicted maps from thin-vessel channels.Main results.Experiments on DRIVE, STARE and CHASE_DB1 demonstrate the superiorities of the proposed MSDC_NET in extracting more thin vessels in fundus images, and quantitative evaluations on several parameters based on the advanced ground truth further verify the advantages of our proposed DL model. Compared with the previous multi-branch method, the specificity and F1score are improved by about 2.18%, 0.68%, 1.73% and 2.91%, 0.24%, 8.38% on the three datasets, respectively.Significance.This work may provide richer information to ophthalmologists for the diagnosis and treatment of vascular-related ophthalmic diseases.

Inter-hospital moderate and advanced Alzheimer's disease detection through convolutional neural networks

Electroencephalography (EEG) has been a fundamental technique in the identification of health conditions since its discovery. This analysis specifically centers on machine learning (ML) and deep learning (DL) methodologies designed for the analysis of electroencephalogram (EEG) data to categorize individuals with Alzheimer's Disease (AD) into two groups: Moderate or Advanced Alzheimer's dementia.Our study is based on a comprehensive database comprising 668 volunteers from 5 different hospitals, collected over a decade. This diverse dataset enables better training and validation of our results. Among the methods evaluated, the CNN (deep learning) approach outperformed others, achieving a remarkable classification accuracy of 97.45% for patients with Moderate Alzheimer's Dementia (ADM) and 97.03% for patients with Advanced Alzheimer's Dementia (ADA). Importantly, all the compared methods were rigorously assessed under identical conditions. The proposed DL model, specifically CNN, effectively extracts time domain features from EEG data in time, resulting in a significant reduction in learnable parameters and data redundancy.

Hybrid Architecture Strategies for the Prediction of Acute Pulmonary Embolism from Computed Tomography Images

The timely identification of pulmonary embolism is of utmost importance, as the condition has the potential to be life-threatening if not promptly addressed. The assessment of the severity of a pulmonary embolism (PE) frequently necessitates a time-consuming and potentially life-threatening estimation by a medical practitioner. The primary objective of the study was to investigate the potential utility of an artificial neural network in assisting physicians with the identification and prediction of pulmonary embolism risk in patients. Deep learning algorithms are frequently employed in medical imaging to enhance image analysis due to their ability to automatically learn representations from large datasets, as opposed to relying on pre-programmed instructions . The implementation of automated systems has the potential to decrease the level of physical effort needed and enhance the efficiency of diagnostic procedures for medical professionals. Efficient training and calculation processes are crucial for the proper execution of the implementation. The Tensor Processing Units (TPUs) developed by Google are employed to expedite the process of training., with the computational tasks being executed through Google Colab, a platform offered by Google Cloud TPUs. In order to achieve outcomes comparable to human judgment, deep learning algorithms engage in reasonable assessments of data based on a predetermined logical framework. Diagnosing pulmonary embolism (PE), a potentially lethal yet curable condition, poses challenges in early detection. A distinctive convolutional neural network (CNN) model was developed and examined for the purpose of distinguishing between pulmonary embolism (PE) and computed tomography (CT) pictures. The proposed study yields a precision rate of 91.2%, showcasing an enhancement compared to current convolutional neural network (CNN) architectures that include limited trainable parameters. Furthermore, our model provides interpretability by the utilization of computed tomography (CT) images, specifically in the inferno and bone models. Our proposed deep learning model has the potential to predict the presence of PE and other associated features in current cases.

Exploring Deep Learning for Estimating the Isoeffective Dose of FLASH Irradiation From Mouse Intestinal Histological Images

PurposeUltra-high dose rate (FLASH) irradiation has been reported to reduce normal tissue damage compared with conventional dose rate (CONV) irradiation without compromising tumor control. This proof-of-concept study aims to develop a deep learning (DL) approach to quantify the FLASH isoeffective dose (dose of CONV that would be required to produce the same effect as the given physical FLASH dose) with post-irradiation mouse intestinal histological images. Methods and materials84 healthy C57BL/6J female mice underwent 16 MeV electron CONV (0.12Gy/s; n=41) or FLASH (200Gy/s; n=43) single fraction whole abdominal irradiation. Physical dose ranged from 12 to 16Gy for FLASH and 11 to 15Gy for CONV in 1Gy increments. 4 days after irradiation, 9 jejunum cross-sections from each mouse were H&E stained and digitized for histological analysis. CONV dataset was randomly split into training (n=33) and testing (n=8) datasets. ResNet101-based DL models were retrained using the CONV training dataset to estimate the dose based on histological features. The classical manual crypt counting (CC) approach was implemented for model comparison. Cross-section-wise mean squared error (CS-MSE) was computed to evaluate the dose estimation accuracy of both approaches. The validated DL model was applied to the FLASH dataset to map the physical FLASH dose into the isoeffective dose. ResultsThe DL model achieved a CS-MSE of 0.20Gy2 on the CONV testing dataset compared with 0.40Gy2 of the CC approach. Isoeffective doses estimated by the DL model for FLASH doses of 12, 13, 14, 15, and 16 Gy were 12.19±0.46, 12.54±0.37, 12.69±0.26, 12.84±0.26, and 13.03±0.28 Gy, respectively. ConclusionsOur proposed DL model achieved accurate CONV dose estimation. The DL model results indicate that in the physical dose range of 13 to 16 Gy, the biological dose response of small intestinal tissue to FLASH irradiation is represented by a lower isoeffective dose compared to the physical dose. Our DL approach can be a tool for studying isoeffective doses of other radiation dose modifying interventions.

Early prediction of distant metastasis in patients with uterine cervical cancer treated with definitive chemoradiotherapy by deep learning using pretreatment [ 18 F]fluorodeoxyglucose positron emission tomography/computed tomography.

A deep learning (DL) model using image data from pretreatment [ 18 F]fluorodeoxyglucose ([ 18 F] FDG)-PET or computed tomography (CT) augmented with a novel imaging augmentation approach was developed for the early prediction of distant metastases in patients with locally advanced uterine cervical cancer. This study used baseline [18F]FDG-PET/CT images of newly diagnosed uterine cervical cancer patients. Data from 186 to 25 patients were analyzed for training and validation cohort, respectively. All patients received chemoradiotherapy (CRT) and follow-up. PET and CT images were augmented by using three-dimensional techniques. The proposed model employed DL to predict distant metastases. Receiver operating characteristic (ROC) curve analysis was performed to measure the model's predictive performance. The area under the ROC curves of the training and validation cohorts were 0.818 and 0.830 for predicting distant metastasis, respectively. In the training cohort, the sensitivity, specificity, and accuracy were 80.0%, 78.0%, and 78.5%, whereas, the sensitivity, specificity, and accuracy for distant failure were 73.3%, 75.5%, and 75.2% in the validation cohort, respectively. Through the use of baseline [ 18 F]FDG-PET/CT images, the proposed DL model can predict the development of distant metastases for patients with locally advanced uterine cervical cancer treatment by CRT. External validation must be conducted to determine the model's predictive performance.

Deep Learning Based on Enhanced MRI T1 Imaging to Differentiate Small-cell and Non-small-cell Primary Lung Cancers in Patients with Brain Metastases.

To differentiate the primary small-cell lung cancer (SCLC) and non-small-cell lung cancer (NSCLC) for patients with brain metastases (BMs) based on a deep learning (DL) model using contrast-enhanced magnetic resonance imaging (MRI) T1 weighted (T1CE) images. Out of 711 patients with BMs of lung cancer origin (SCLC 232, NSCLC 479), the MRI datasets of 192 patients (lesions' widths and heights > 30 pixels) with BMs from lung cancer (73 SCLC and 119 NSCLC) confirmed pathologically were enrolled, retrospectively. A typical convolutional neural network ResNet18 was applied for the automatic classification of BMs lesions from lung cancer based on T1CE images, with training and testing groups randomized per patient to eliminate learning bias. A 5-fold cross-validation was performed to evaluate the classification of the model. The receiver operating characteristic (ROC) curve, accuracy, precision, recall and f1 score were calculated. For a 5-fold cross-validation test, the DL model achieved AUCs of 0.8019 and 0.8024 for SCLC and NSCLC patients with BMs, respectively, and a mean overall accuracy of 0.7515±0.04. The DL model performed well in differentiating the primary SCLC and NSCLC with BMs. The proposed DL model is feasible and effective in differentiating the pathological subtypes of SCLC and NSCLC causing BMs, which may be used as a new tool for oncologists to diagnose noninvasively BMs and guide therapy based on the imaging structure of tumors.