3D Deep Convolutional Neural Network Research Articles

Persistent Luminescence Lifetime-Based Near-Infrared Nanoplatform via Deep Learning for High-Fidelity Biosensing of Hypochlorite.

In light of deep tissue penetration and ultralow background, near-infrared (NIR) persistent luminescence (PersL) bioprobes have become powerful tools for bioapplications. However, the inhomogeneous signal attenuation may significantly limit its application for precise biosensing owing to tissue absorption and scattering. In this work, a PersL lifetime-based nanoplatform via deep learning was proposed for high-fidelity bioimaging and biosensing in vivo. The persistent luminescence imaging network (PLI-Net), which consisted of a 3D-deep convolutional neural network (3D-CNN) and the PersL imaging system, was logically constructed to accurately extract the lifetime feature from the profile of PersL intensity-based decay images. Significantly, the NIR PersL nanomaterials represented by Zn1+xGa2-2xSnxO4: 0.4 % Cr (ZGSO) were precisely adjusted over their lifetime, enabling the PersL lifetime-based imaging with high-contrast signals. Inspired by the adjustable and reliable PersL lifetime imaging of ZGSO NPs, a proof-of-concept PersL nanoplatform was further developed and showed exceptional analytical performance for hypochlorite detection via a luminescence resonance energy transfer process. Remarkably, on the merits of the dependable and anti-interference PersL lifetimes, this PersL lifetime-based nanoprobe provided highly sensitive and accurate imaging of both endogenous and exogenous hypochlorite. This breakthrough opened up a new way for the development of high-fidelity biosensing in complex matrix systems.

Deep learning-based radiomics of computed tomography angiography to predict adverse events after initial endovascular repair for acute uncomplicated Stanford type B aortic dissection

PurposeThis study aimed to construct a predictive model integrating deep learning-derived radiomic features from computed tomography angiography (CTA) and clinical biomarkers to forecast postoperative adverse events (AEs) in patients with acute uncomplicated Stanford type B aortic dissection (uTBAD) undergoing initial thoracic endovascular aortic repair (TEVAR). MethodsWe retrospectively evaluated 369 patients treated with TEVAR for acute uTBAD from January 2015 to December 2022. A three-dimensional (3D) deep convolutional neural network (CNN) automated radiomic feature extraction from CTA images. Feature selection, using Analysis of Variance (ANOVA) and the Least Absolute Shrinkage and Selection Operator (LASSO) algorithms, refined a radiomic score (Rad-Score). This score, alongside clinical parameters, was modelled via Extreme Gradient Boosting (XGBoost) analysis. Model calibration was assessed by calibration curves. ResultsThe integration of the Rad-Score with clinical factors including albumin and C-reactive protein levels moderately enhanced predictive efficiency, exhibiting an area under the curve (AUC) of 1.000 (95%CI, 1.000–1.000) in the training cohort and 0.990 (95%CI, 0.966–1.000) in the internal validation cohort. In an independent validation cohort from another hospital, the combined model yielded an AUC of 0.985 (95%CI, 0.965–1.000), with an accuracy, precision, sensitivity, and specificity of 0.92, 0.92, 0.94, and 0.91, respectively. ConclusionsThe synergistic application of deep learning-based radiomics from CTA and clinical indicators holds promise for anticipating AEs post-initial thoracic endovascular aortic repair in patients with acute uTBAD. The clinical utility of the constructed combined model, offering prognostic foresight during follow-up, has been substantiated.

Deep Machine Learning for Medical Diagnosis, Application to Lung Cancer Detection: A Review

Deep learning has emerged as a powerful tool for medical image analysis and diagnosis, demonstrating high performance on tasks such as cancer detection. This literature review synthesizes current research on deep learning techniques applied to lung cancer screening and diagnosis. This review summarizes the state-of-the-art in deep learning for lung cancer detection, highlighting key advances, limitations, and future directions. We prioritized studies utilizing major public datasets, such as LIDC, LUNA16, and JSRT, to provide a comprehensive overview of the field. We focus on deep learning architectures, including 2D and 3D convolutional neural networks (CNNs), dual-path networks, Natural Language Processing (NLP) and vision transformers (ViT). Across studies, deep learning models consistently outperformed traditional machine learning techniques in terms of accuracy, sensitivity, and specificity for lung cancer detection in CT scans. This is attributed to the ability of deep learning models to automatically learn discriminative features from medical images and model complex spatial relationships. However, several challenges remain to be addressed before deep learning models can be widely deployed in clinical practice. These include model dependence on training data, generalization across datasets, integration of clinical metadata, and model interpretability. Overall, deep learning demonstrates great potential for lung cancer detection and precision medicine. However, more research is required to rigorously validate models and address risks. This review provides key insights for both computer scientists and clinicians, summarizing progress and future directions for deep learning in medical image analysis.

Automated MRI liver segmentation for anatomical segmentation, liver volumetry, and the extraction of radiomics.

To develop and evaluate a deep convolutional neural network (DCNN) for automated liver segmentation, volumetry, and radiomic feature extraction on contrast-enhanced portal venous phase magnetic resonance imaging (MRI). This retrospective study included hepatocellular carcinoma patients from an institutional database with portal venous MRI. After manual segmentation, the data was randomly split into independent training, validation, and internal testing sets. From a collaborating institution, de-identified scans were used for external testing. The public LiverHccSeg dataset was used for further external validation. A 3D DCNN was trained to automatically segment the liver. Segmentation accuracy was quantified by theDice similarity coefficient (DSC) with respect to manual segmentation. A Mann-Whitney U test was used to compare the internal and external test sets. Agreement of volumetry and radiomic features was assessed using the intraclass correlation coefficient (ICC). In total, 470 patients met the inclusion criteria (63.9±8.2 years; 376 males) and 20 patients were used for external validation (41±12 years; 13 males). DSC segmentation accuracy of the DCNN was similarly high between the internal (0.97±0.01) and external (0.96±0.03) test sets (p=0.28) and demonstrated robust segmentation performance on public testing (0.93±0.03). Agreement of liver volumetry was satisfactory in the internal (ICC, 0.99), external (ICC, 0.97), and public (ICC, 0.85) test sets. Radiomic features demonstrated excellent agreement in the internal (mean ICC, 0.98±0.04), external (mean ICC, 0.94±0.10), and public (mean ICC, 0.91±0.09) datasets. Automated liver segmentation yields robust and generalizable segmentation performance on MRI data and can be used for volumetry and radiomic feature extraction. Liver volumetry, anatomic localization, and extraction of quantitative imaging biomarkers require accurate segmentation, but manual segmentation is time-consuming. A deep convolutional neural network demonstrates fast and accurate segmentation performance on T1-weighted portal venous MRI. • This deep convolutional neural network yields robust and generalizable liver segmentation performance on internal, external, and public testing data. • Automated liver volumetry demonstrated excellent agreement with manual volumetry. • Automated liver segmentations can be used for robust and reproducible radiomic feature extraction.

A Combination Model of Shifting Joint Angle Changes With 3D-Deep Convolutional Neural Network to Recognize Human Activity.

Research in the field of human activity recognition is very interesting due to its potential for various applications such as in the field of medical rehabilitation. The need to advance its development has become increasingly necessary to enable efficient detection and response to a wide range of movements. Current recognition methods rely on calculating changes in joint distance to classify activity patterns. Therefore, a different approach is required to identify the direction of movement to distinguish activities exhibiting similar joint distance changes but differing motion directions, such as sitting and standing. The research conducted in this study focused on determining the direction of movement using an innovative joint angle shift approach. By analyzing the joint angle shift value between specific joints and reference points in the sequence of activity frames, the research enabled the detection of variations in activity direction. The joint angle shift method was combined with a Deep Convolutional Neural Network (DCNN) model to classify 3D datasets encompassing spatial-temporal information from RGB-D video image data. Model performance was evaluated using the confusion matrix. The results show that the model successfully classified nine activities in the Florence 3D Actions dataset, including sitting and standing, obtaining an accuracy of (96.72 ± 0.83)%. In addition, to evaluate its robustness, this model was tested on the UTKinect Action3D dataset, obtaining an accuracy of 97.44%, proving that state-of-the-art performance has been achieved.

A Complete Strategy to Achieve High Precision Automatic Segmentation of Challenging Experimental X‐Ray Computed Tomography Data Using Low‐Resemblance Synthetic Training Data

It is shown that preconditioning of experimental X‐ray computed tomography (XCT) data is critical to achieve high‐precision segmentation scores. The challenging experimental XCT datasets and deep convolutional neural networks (DCNNs) are used that are trained with low‐resemblance synthetic XCT data. The material used is a 6‐phase Al–Si metal matrix composite‐reinforced with ceramic fibers and particles. To achieve generalization, in our past studies, specific data augmentation techniques were proposed for the synthetic XCT training data. In addition, two toolsets are devised: (1) special 3D DCNN architecture (3D Triple_UNet), slicing the experimental XCT data from multiple views (MultiView Forwarding), the i.S.Sy.Da.T.A. iterative segmentation algorithm, and (2) nonlocal means (NLM) conditioning (filtering) for the experimental XCT data. This results in good segmentation Dice scores across all phases compared to more standard approaches (i.e., standard UNet architecture, single view slicing, standard single training, and NLM conditioning). Herein, the NLM filter is replaced with the deep conditioning framework BAM SynthCOND introduced in a previous publication, which can be trained with synthetic XCT data. This leads to a significant segmentation precision increase for all phases. The proposed methods are potentially applicable to other materials and imaging techniques.

Clinically Oriented CBCT Periapical Lesion Evaluation via 3D CNN Algorithm

Apical periodontitis (AP) is one of the most prevalent disorders in dentistry. However, it can be underdiagnosed in asymptomatic patients. In addition, the perioperative evaluation of 3-dimensional (3D) lesion volume is of great clinical relevance, but the required slice-by-slice manual delineation method is time- and labor-intensive. Here, for quickly and accurately detecting and segmenting periapical lesions (PALs) associated with AP on cone beam computed tomography (CBCT) images, we proposed and geographically validated a novel 3D deep convolutional neural network algorithm, named PAL-Net. On the internal 5-fold cross-validation set, our PAL-Net achieved an area under the receiver operating characteristic curve (AUC) of 0.98. The algorithm also improved the diagnostic performance of dentists with varying levels of experience, as evidenced by their enhanced average AUC values (junior dentists: 0.89–0.94; senior dentists: 0.91–0.93), and significantly reduced the diagnostic time (junior dentists: 69.3 min faster; senior dentists: 32.4 min faster). Moreover, our PAL-Net achieved an average Dice similarity coefficient over 0.87 (0.85–0.88), which is superior or comparable to that of other existing state-of-the-art PAL segmentation algorithms. Furthermore, we validated the generalizability of the PAL-Net system using multiple external data sets from Central, East, and North China, showing that our PAL-Net has strong robustness. Our PAL-Net can help improve the diagnostic performance and speed of dentists working from CBCT images, provide clinically relevant volume information to dentists, and can potentially be applied in dental clinics, especially without expert-level dentists or radiologists.

Data-efficient resting-state functional magnetic resonance imaging brain mapping with deep learning.

Resting-state functional MRI (RS-fMRI) enables the mapping of function within the brain and is emerging as an efficient tool for the presurgical evaluation of eloquent cortex. Models capable of reliable and precise mapping of resting-state networks (RSNs) with a reduced scanning time would lead to improved patient comfort while reducing the cost per scan. The aims of the present study were to develop a deep 3D convolutional neural network (3DCNN) capable of voxel-wise mapping of language (LAN) and motor (MOT) RSNs with minimal quantities of RS-fMRI data. Imaging data were gathered from multiple ongoing studies at Washington University School of Medicine and other thoroughly characterized, publicly available data sets. All study participants (n = 2252 healthy adults) were cognitively screened and completed structural neuroimaging and RS-fMRI. Random permutations of RS-fMRI regions of interest were used to train a 3DCNN. After training, model inferences were compared using varying amounts of RS-fMRI data from the control data set as well as 5 patients with glioblastoma multiforme. The trained model achieved 96% out-of-sample validation accuracy on data encompassing a large age range collected on multiple scanner types and varying sequence parameters. Testing on out-of-sample control data showed 97.9% similarity between results generated using either 50 or 200 RS-fMRI time points, corresponding to approximately 2.5 and 10 minutes, respectively (96.9% LAN, 96.3% MOT true-positive rate). In evaluating data from patients with brain tumors, the 3DCNN was able to accurately map LAN and MOT networks despite structural and functional alterations. Functional maps produced by the 3DCNN can inform surgical planning in patients with brain tumors in a time-efficient manner. The authors present a highly efficient method for presurgical functional mapping and thus improved functional preservation in patients with brain tumors.

An Integrated Multi-Channel Deep Neural Network for Mesial Temporal Lobe Epilepsy Identification Using Multi-Modal Medical Data.

Epilepsy is a chronic brain disease with recurrent seizures. Mesial temporal lobe epilepsy (MTLE) is the most common pathological cause of epilepsy. With the development of computer-aided diagnosis technology, there are many auxiliary diagnostic approaches based on deep learning algorithms. However, the causes of epilepsy are complex, and distinguishing different types of epilepsy accurately is challenging with a single mode of examination. In this study, our aim is to assess the combination of multi-modal epilepsy medical information from structural MRI, PET image, typical clinical symptoms and personal demographic and cognitive data (PDC) by adopting a multi-channel 3D deep convolutional neural network and pre-training PET images. The results show better diagnosis accuracy than using one single type of medical data alone. These findings reveal the potential of a deep neural network in multi-modal medical data fusion.

Anomalous sound detection for machine condition monitoring using 3D tensor representation of sound and 3D deep convolutional neural network

Anomalous sound detection for machine condition monitoring using 3D tensor representation of sound and 3D deep convolutional neural network

서울시 대기 중 미세먼지 농도 예측을 위한 3D 심층 합성곱 신경망 앙상블 예측

서울시 대기 중 미세먼지 농도 예측을 위한 3D 심층 합성곱 신경망 앙상블 예측

Automated intracranial hemorrhage detection in traumatic brain injury using 3D CNN.

Intracranial hemorrhage (ICH) is a prevalent and potentially fatal consequence of traumatic brain injury (TBI). Timely identification of ICH is crucial to ensure timely intervention and to optimize better patient outcomes. However, the current methods for diagnosing ICH from head computed tomography (CT) scans require skilled personnel (Radiologists and/or Neurosurgeons) who may be unavailable in all centers, especially in rural areas. The aim of this study is to develop a neurotrauma screening tool for identifying ICH from head CT scans of TBI patients. We prospectively collected head CT scans from the Department of Neurosurgery, All India Institute of Medical Sciences, New Delhi. Approximately 738 consecutive head CT scans from patients enrolled in the department were collected for this study spanning a duration of 9 months, that is, January 2020 to September 2020. The metadata collected along with the head CT scans consisted of demographic and clinical details and the radiologist's report which was used as the gold standard. A deep learning-based 3D convolutional neural network (CNN) model was trained on the dataset. The pre-processing, hyperparameters, and augmentation were common for training the 3D CNN model whereas the training modules were set differently. The model was trained along with the save best model option and was monitored by validation metrics. The Institute Ethics Committee permission was taken before starting the study. We developed a 3D CNN model for automatically detecting the ICH from head CT scans. The screening tool was tested in 20 cases and trained on 200 head CT scans, with 99 normal head CT and 101 CT scans with some type of ICH. The final model performed with 90% sensitivity, 70% specificity, and 80% accuracy. Our study reveals that the automated screening tool exhibits a commendable level of accuracy and sensitivity in detecting ICH from the head CT scans. The results indicate that the 3D CNN approach has a potential for further exploring the TBI-related pathologies.

Combining 3D skeleton data and deep convolutional neural network for balance assessment during walking.

Introduction: Balance impairment is an important indicator to a variety of diseases. Early detection of balance impairment enables doctors to provide timely treatments to patients, thus reduce their fall risk and prevent related disease progression. Currently, balance abilities are usually assessed by balance scales, which depend heavily on the subjective judgement of assessors. Methods: To address this issue, we specifically designed a method combining 3D skeleton data and deep convolutional neural network (DCNN) for automated balance abilities assessment during walking. A 3D skeleton dataset with three standardized balance ability levels were collected and used to establish the proposed method. To obtain better performance, different skeleton-node selections and different DCNN hyperparameters setting were compared. Leave-one-subject-out-cross-validation was used in training and validation of the networks. Results and Discussion: Results showed that the proposed deep learning method was able to achieve 93.33% accuracy, 94.44% precision and 94.46% F1 score, which outperformed four other commonly used machine learning methods and CNN-based methods. We also found that data from body trunk and lower limbs are the most important while data from upper limbs may reduce model accuracy. To further validate the performance of the proposed method, we migrated and applied a state-of-the-art posture classification method to the walking balance ability assessment task. Results showed that the proposed DCNN model improved the accuracy of walking balance ability assessment. Layer-wise Relevance Propagation (LRP) was used to interpret the output of the proposed DCNN model. Our results suggest that DCNN classifier is a fast and accurate method for balance assessment during walking.

OptiMD-3D DCNN: A Framework for Restoring the Haze-Free Images by Image Dehazing Techniques Using Heuristic Approach of Adaptive Lifting Wavelet Transform

The recently implemented illustrative techniques are focused to restore haze-free images by using learning approaches and also physical models. Managing finer details of the images when completely removing the haze is a complicated task especially in performing the hazing on single images. Generally, hazy images comprise distorted colors, poor contrast, and lower visibility, which lead to lower efficiency in the recognition, detection of objects, and classification of images. The main aspect of the proposed work is to restore haze-free images without any degradation in image quality. Initially, the required raw hazy images are acquired from the benchmark data sources. Further, these images undergo the decomposing stage, which is accomplished by adopting the Adaptive Lifting Wavelet Transform (ALWT), where the wavelet levels are optimized by using the improved hybrid algorithm of the Modified Tunicate Galactic Swarm Optimization (MTGSO). Once the decomposed images are obtained, Optimal Multiscale Dilation assisted 3D Deep Convolutional Neural Network (OptiMD-3D DCNN) is developed for the image dehazing task, in which the hyperparameters are optimally tuned using a hybrid MTGSO algorithm. At last, the effectiveness is examined and guaranteed that the recommended scheme shows higher performance in estimating with the classical methods.

Resting state network mapping in individuals using deep learning.

Resting state functional MRI (RS-fMRI) is currently used in numerous clinical and research settings. The localization of resting state networks (RSNs) has been utilized in applications ranging from group analysis of neurodegenerative diseases to individual network mapping for pre-surgical planning of tumor resections. Reproducibility of these results has been shown to require a substantial amount of high-quality data, which is not often available in clinical or research settings. In this work, we report voxelwise mapping of a standard set of RSNs using a novel deep 3D convolutional neural network (3DCNN). The 3DCNN was trained on publicly available functional MRI data acquired in n = 2010 healthy participants. After training, maps that represent the probability of a voxel belonging to a particular RSN were generated for each participant, and then used to calculate mean and standard deviation (STD) probability maps, which are made publicly available. Further, we compared our results to previously published resting state and task-based functional mappings. Our results indicate this method can be applied in individual subjects and is highly resistant to both noisy data and fewer RS-fMRI time points than are typically acquired. Further, our results show core regions within each network that exhibit high average probability and low STD. The 3DCNN algorithm can generate individual RSN localization maps, which are necessary for clinical applications. The similarity between 3DCNN mapping results and task-based fMRI responses supports the association of specific functional tasks with RSNs.

3D bi-directional transformer U-Net for medical image segmentation.

As one of the popular deep learning methods, deep convolutional neural networks (DCNNs) have been widely adopted in segmentation tasks and have received positive feedback. However, in segmentation tasks, DCNN-based frameworks are known for their incompetence in dealing with global relations within imaging features. Although several techniques have been proposed to enhance the global reasoning of DCNN, these models are either not able to gain satisfying performances compared with traditional fully-convolutional structures or not capable of utilizing the basic advantages of CNN-based networks (namely the ability of local reasoning). In this study, compared with current attempts to combine FCNs and global reasoning methods, we fully extracted the ability of self-attention by designing a novel attention mechanism for 3D computation and proposed a new segmentation framework (named 3DTU) for three-dimensional medical image segmentation tasks. This new framework processes images in an end-to-end manner and executes 3D computation on both the encoder side (which contains a 3D transformer) and the decoder side (which is based on a 3D DCNN). We tested our framework on two independent datasets that consist of 3D MRI and CT images. Experimental results clearly demonstrate that our method outperforms several state-of-the-art segmentation methods in various metrics.

Face ShapeNets for 3D Face Recognition

In this paper, we present a deep learning-based method for 3D face recognition. Unlike some previous works, our process does not rely on face representation methods as a proxy step to be accepted by Convolutional Neural Networks (CNNs). Applying 2D CNNs to irregular domains such as 3D meshes is challenging. Therefore, we employed 3D ShapeNets to recognize faces covering the full 3D shape since 3D face datasets are available and 3D data augmentation techniques to enlarge 3D datasets are widespread. The reduced size of 3D datasets is overcome by an appropriate 3D data augmentation to train our model. 3D ShapeNets are commonly used to recognize and analyse objects. To the best of the authors’ knowledge, this is the first time they are used for face recognition. This research work focuses on the preprocessing step. Whatever the nature of the face image is (either 2D or 3D representation), and whatever the acquisition conditions are, a 3D regular mesh of each input face image is first generated. Furthermore, meshes are converted to voxels to get the occupancy grid across all possible views. Finally, 3D ShapeNets are trained and recognition tests are performed. Indeed, 3D ShapeNets prove the efficiency and efficacy of 3D shape analysis task in 3D face recognition. The experimental results show that 3D face recognition using deep 3D CNNs such as 3D ShapeNets leads to significant improvement over the state-of-the-art performance on LFPW, BU3DFE, and FRAV3D datasets, with competitive recognition rates of 94.25%, 97.9% and 98.31% respectively.

Deep3DCANN: A Deep 3DCNN-ANN framework for spontaneous micro-expression recognition

Facial micro-expressions play a significant role in revealing concealed emotions. However, the recognition of micro-expressions is challenging due to their fleeting nature. Moreover, the visual features of the face and the visual relationships between the facial sub-regions have a strong influence on the presence of micro-expressions. In this work, a novel end-to-end facial micro-expression detection framework, called Deep3DCANN, is proposed to integrate these components for effective micro-expression detection. The first component of our framework is a deep 3D convolutional neural network that learns useful spatiotemporal features from a sequence of facial images. In the second component, a deep artificial neural network is utilized to trace the useful visual associations between different sub-regions of the face. Furthermore, a carefully crafted fusion mechanism is built to combine the learned facial features and the semantic relationships between the regions to predict the micro-expressions. We also construct a new loss function to jointly optimize both modules of our proposed architecture. Our proposed method performs favourably on five benchmark spontaneous micro-expression databases compared to existing micro-expression recognition baselines on videos. In addition, through an extended experiment, we show that our proposed approach can effectively recognize the frame-wise micro-expression changes in a sequence of video frames.

Automated meniscus segmentation and tear detection of knee MRI with a 3D mask-RCNN

BackgroundThe diagnostic results of magnetic resonance imaging (MRI) are essential references for arthroscopy as an invasive procedure. A deviation between medical imaging diagnosis and arthroscopy results may cause irreversible damage to patients and lead to excessive medical treatment. To improve the accurate diagnosis of meniscus injury, it is urgent to develop auxiliary diagnosis algorithms to improve the accuracy of radiological diagnosis.PurposeThis study aims to present a fully automatic 3D deep convolutional neural network (DCNN) for meniscus segmentation and detects arthroscopically proven meniscus tears.Materials and methodsOur institution retrospectively included 533 patients with 546 knees who underwent knee magnetic resonance imaging (MRI) and knee arthroscopy. Sagittal proton density-weighted (PDW) images in MRI of 382 knees were regarded as a training set to train our 3D-Mask RCNN. The remaining data from 164 knees were used to validate the trained network as a test set. The masks were hand-drawn by an experienced radiologist, and the reference standard is arthroscopic surgical reports. The performance statistics included Dice accuracy, sensitivity, specificity, FROC, receiver operating characteristic (ROC) curve analysis, and bootstrap test statistics. The segmentation performance was compared with a 3D-Unet, and the detection performance was compared with radiological evaluation by two experienced musculoskeletal radiologists without knowledge of the arthroscopic surgical diagnosis.ResultsOur model produced strong Dice coefficients for sagittal PDW of 0.924, 0.95 sensitivity with 0.823 FPs/knee. 3D-Unet produced a Dice coefficient for sagittal PDW of 0.891, 0.95 sensitivity with 1.355 FPs/knee. The difference in the areas under 3D-Mask-RCNN FROC and 3D-Unet FROC was statistically significant (p = 0.0011) by bootstrap test. Our model detection performance achieved an area under the curve (AUC) value, accuracy, and sensitivity of 0.907, 0.924, 0.941, and 0.785, respectively. Based on the radiological evaluations, the AUC value, accuracy, sensitivity, and specificity were 0.834, 0.835, 0.889, and 0.754, respectively. The difference in the areas between 3D-Mask-RCNN ROC and radiological evaluation ROC was statistically significant (p = 0.0009) by bootstrap test. 3D Mask RCNN significantly outperformed the 3D-Unet and radiological evaluation demonstrated by these results.Conclusions3D-Mask RCNN has demonstrated efficacy and precision for meniscus segmentation and tear detection in knee MRI, which can assist radiologists in improving the accuracy and efficiency of diagnosis. It can also provide effective diagnostic indicators for orthopedic surgeons before arthroscopic surgery and further promote precise treatment.

LCSCNet: A multi-level approach for lung cancer stage classification using 3D dense convolutional neural networks with concurrent squeeze-and-excitation module

Lung cancer, the deadliest disease worldwide, poses a massive threat to humankind. Various researchers have designed Computer-Aided-Diagnosis systems for the early-stage detection of lung cancer. However, patients are primarily diagnosed in advanced stages when treatment becomes complicated and dependent on multiple factors like size, nature, location of the tumor, and proper cancer staging. TNM (Tumor, Node, and Metastasis) staging provides all this information. This study aims to develop a novel and efficient approach to classify lung cancer stages based on TNM standards. We propose a multi-level 3D deep convolutional neural network, LCSCNet (Lung Cancer Stage Classification Network). The proposed network architecture consists of three similar classifier networks to classify three labels, T, N, and M-labels. First, we pre-process the data, in which the CT images are augmented, and the label files are processed to get the corresponding TNM labels. For the classification network, we implement a dense convolutional neural network with a concurrent squeeze & excitation module and asymmetric convolutions for classifying each label separately. The overall stage is determined by combining all three labels. The concurrent squeeze & excitation module helps the network focus on the essential information of the image, due to which the classification performance is enhanced. The asymmetric convolutions are introduced to reduce the computation complexity of the network. Two publicly available datasets are used for this study. We achieved average accuracies of 96.23% for T-Stage, 97.63% for N-Stage, and 96.92% for M-Stage classification. Furthermore, an overall stage classification accuracy of 97% is achieved.