Lung Pattern Classification for Interstitial Lung Diseases Using a Deep Convolutional Neural Network

A new type of neurons for machine learning

In machine learning, an artificial neural network is the mainstream approach. Such a network consists of many neurons. These neurons are of the same type characterized by the 2 features: (1) an inner product of an input vector and a matching weighting vector of trainable parameters and (2) a nonlinear excitation function. Here, we investigate the possibility of replacing the inner product with a quadratic function of the input vector, thereby upgrading the first-order neuron to the second-order neuron, empowering individual neurons and facilitating the optimization of neural networks. Also, numerical examples are provided to illustrate the feasibility and merits of the second-order neurons. Finally, further topics are discussed.

A hybrid CNN feature model for pulmonary nodule malignancy risk differentiation

The malignancy risk differentiation of pulmonary nodule is one of the most challenge tasks of computer-aided diagnosis (CADx). Most recently reported CADx methods or schemes based on texture and shape estimation have shown relatively satisfactory on differentiating the risk level of malignancy among the nodules detected in lung cancer screening. However, the existing CADx schemes tend to detect and analyze characteristics of pulmonary nodules from a statistical perspective according to local features only. Enlightened by the currently prevailing learning ability of convolutional neural network (CNN), which simulates human neural network for target recognition and our previously research on texture features, we present a hybrid model that takes into consideration of both global and local features for pulmonary nodule differentiation using the largest public database founded by the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI). By comparing three types of CNN models in which two of them were newly proposed by us, we observed that the multi-channel CNN model yielded the best discrimination in capacity of differentiating malignancy risk of the nodules based on the projection of distributions of extracted features. Moreover, CADx scheme using the new multi-channel CNN model outperformed our previously developed CADx scheme using the 3D texture feature analysis method, which increased the computed area under a receiver operating characteristic curve (AUC) from 0.9441 to 0.9702.

Particle swarm optimization-based automatic parameter selection for deep neural networks and its applications in large-scale and high-dimensional data

In this paper, we propose a new automatic hyperparameter selection approach for determining the optimal network configuration (network structure and hyperparameters) for deep neural networks using particle swarm optimization (PSO) in combination with a steepest gradient descent algorithm. In the proposed approach, network configurations were coded as a set of real-number m-dimensional vectors as the individuals of the PSO algorithm in the search procedure. During the search procedure, the PSO algorithm is employed to search for optimal network configurations via the particles moving in a finite search space, and the steepest gradient descent algorithm is used to train the DNN classifier with a few training epochs (to find a local optimal solution) during the population evaluation of PSO. After the optimization scheme, the steepest gradient descent algorithm is performed with more epochs and the final solutions (pbest and gbest) of the PSO algorithm to train a final ensemble model and individual DNN classifiers, respectively. The local search ability of the steepest gradient descent algorithm and the global search capabilities of the PSO algorithm are exploited to determine an optimal solution that is close to the global optimum. We constructed several experiments on hand-written characters and biological activity prediction datasets to show that the DNN classifiers trained by the network configurations expressed by the final solutions of the PSO algorithm, employed to construct an ensemble model and individual classifier, outperform the random approach in terms of the generalization performance. Therefore, the proposed approach can be regarded an alternative tool for automatic network structure and parameter selection for deep neural networks.

Computational biology: deep learning

Deep learning is the trendiest tool in a computational biologist's toolbox. This exciting class of methods, based on artificial neural networks, quickly became popular due to its competitive performance in prediction problems. In pioneering early work, applying simple network architectures to abundant data already provided gains over traditional counterparts in functional genomics, image analysis, and medical diagnostics. Now, ideas for constructing and training networks and even off-the-shelf models have been adapted from the rapidly developing machine learning subfield to improve performance in a range of computational biology tasks. Here, we review some of these advances in the last 2 years.

Searching for prostate cancer by fully automated magnetic resonance imaging classification: deep learning versus non-deep learning

Prostate cancer (PCa) is a major cause of death since ancient time documented in Egyptian Ptolemaic mummy imaging. PCa detection is critical to personalized medicine and varies considerably under an MRI scan. 172 patients with 2,602 morphologic images (axial 2D T2-weighted imaging) of the prostate were obtained. A deep learning with deep convolutional neural network (DCNN) and a non-deep learning with SIFT image feature and bag-of-word (BoW), a representative method for image recognition and analysis, were used to distinguish pathologically confirmed PCa patients from prostate benign conditions (BCs) patients with prostatitis or prostate benign hyperplasia (BPH). In fully automated detection of PCa patients, deep learning had a statistically higher area under the receiver operating characteristics curve (AUC) than non-deep learning (P = 0.0007 < 0.001). The AUCs were 0.84 (95% CI 0.78-0.89) for deep learning method and 0.70 (95% CI 0.63-0.77) for non-deep learning method, respectively. Our results suggest that deep learning with DCNN is superior to non-deep learning with SIFT image feature and BoW model for fully automated PCa patients differentiation from prostate BCs patients. Our deep learning method is extensible to image modalities such as MR imaging, CT and PET of other organs.

ELHnet: a convolutional neural network for classifying cochlear endolymphatic hydrops imaged with optical coherence tomography

Detection of endolymphatic hydrops is important for diagnosing Meniere's disease, and can be performed non-invasively using optical coherence tomography (OCT) in animal models as well as potentially in the clinic. Here, we developed ELHnet, a convolutional neural network to classify endolymphatic hydrops in a mouse model using learned features from OCT images of mice cochleae. We trained ELHnet on 2159 training and validation images from 17 mice, using only the image pixels and observer-determined labels of endolymphatic hydrops as the inputs. We tested ELHnet on 37 images from 37 mice that were previously not used, and found that the neural network correctly classified 34 of the 37 mice. This demonstrates an improvement in performance from previous work on computer-aided classification of endolymphatic hydrops. To the best of our knowledge, this is the first deep CNN designed for endolymphatic hydrops classification.

Residual Deep Convolutional Neural Network Predicts MGMT Methylation Status

Predicting methylation of the O6-methylguanine methyltransferase (MGMT) gene status utilizing MRI imaging is of high importance since it is a predictor of response and prognosis in brain tumors. In this study, we compare three different residual deep neural network (ResNet) architectures to evaluate their ability in predicting MGMT methylation status without the need for a distinct tumor segmentation step. We found that the ResNet50 (50 layers) architecture was the best performing model, achieving an accuracy of 94.90% (+/- 3.92%) for the test set (classification of a slice as no tumor, methylated MGMT, or non-methylated). ResNet34 (34 layers) achieved 80.72% (+/- 13.61%) while ResNet18 (18 layers) accuracy was 76.75% (+/- 20.67%). ResNet50 performance was statistically significantly better than both ResNet18 and ResNet34 architectures (p < 0.001). We report a method that alleviates the need of extensive preprocessing and acts as a proof of concept that deep neural architectures can be used to predict molecular biomarkers from routine medical images.

Gastric precancerous diseases classification using CNN with a concise model

Gastric precancerous diseases (GPD) may deteriorate into early gastric cancer if misdiagnosed, so it is important to help doctors recognize GPD accurately and quickly. In this paper, we realize the classification of 3-class GPD, namely, polyp, erosion, and ulcer using convolutional neural networks (CNN) with a concise model called the Gastric Precancerous Disease Network (GPDNet). GPDNet introduces fire modules from SqueezeNet to reduce the model size and parameters about 10 times while improving speed for quick classification. To maintain classification accuracy with fewer parameters, we propose an innovative method called iterative reinforced learning (IRL). After training GPDNet from scratch, we apply IRL to fine-tune the parameters whose values are close to 0, and then we take the modified model as a pretrained model for the next training. The result shows that IRL can improve the accuracy about 9% after 6 iterations. The final classification accuracy of our GPDNet was 88.90%, which is promising for clinical GPD recognition.

Automated Classification of Lung Cancer Types from Cytological Images Using Deep Convolutional Neural Networks

Lung cancer is a leading cause of death worldwide. Currently, in differential diagnosis of lung cancer, accurate classification of cancer types (adenocarcinoma, squamous cell carcinoma, and small cell carcinoma) is required. However, improving the accuracy and stability of diagnosis is challenging. In this study, we developed an automated classification scheme for lung cancers presented in microscopic images using a deep convolutional neural network (DCNN), which is a major deep learning technique. The DCNN used for classification consists of three convolutional layers, three pooling layers, and two fully connected layers. In evaluation experiments conducted, the DCNN was trained using our original database with a graphics processing unit. Microscopic images were first cropped and resampled to obtain images with resolution of 256 × 256 pixels and, to prevent overfitting, collected images were augmented via rotation, flipping, and filtering. The probabilities of three types of cancers were estimated using the developed scheme and its classification accuracy was evaluated using threefold cross validation. In the results obtained, approximately 71% of the images were classified correctly, which is on par with the accuracy of cytotechnologists and pathologists. Thus, the developed scheme is useful for classification of lung cancers from microscopic images.

Fully Automated Deep Learning System for Bone Age Assessment

Skeletal maturity progresses through discrete phases, a fact that is used routinely in pediatrics where bone age assessments (BAAs) are compared to chronological age in the evaluation of endocrine and metabolic disorders. While central to many disease evaluations, little has changed to improve the tedious process since its introduction in 1950. In this study, we propose a fully automated deep learning pipeline to segment a region of interest, standardize and preprocess input radiographs, and perform BAA. Our models use an ImageNet pretrained, fine-tuned convolutional neural network (CNN) to achieve 57.32 and 61.40% accuracies for the female and male cohorts on our held-out test images. Female test radiographs were assigned a BAA within 1 year 90.39% and within 2 years 98.11% of the time. Male test radiographs were assigned 94.18% within 1 year and 99.00% within 2 years. Using the input occlusion method, attention maps were created which reveal what features the trained model uses to perform BAA. These correspond to what human experts look at when manually performing BAA. Finally, the fully automated BAA system was deployed in the clinical environment as a decision supporting system for more accurate and efficient BAAs at much faster interpretation time (<2 s) than the conventional method.

Automatic tissue characterization of air trapping in chest radiographs using deep neural networks

Significant progress has been made in recent years for computer-aided diagnosis of abnormal pulmonary textures from computed tomography (CT) images. Similar initiatives in chest radiographs (CXR), the common modality for pulmonary diagnosis, are much less developed. CXR are fast, cost effective and low-radiation solution to diagnosis over CT. However, the subtlety of textures in CXR makes them hard to discern even by trained eye. We explore the performance of deep learning abnormal tissue characterization from CXR. Prior studies have used CT imaging to characterize air trapping in subjects with pulmonary disease; however, the use of CT in children is not recommended mainly due to concerns pertaining to radiation dosage. In this work, we present a stacked autoencoder (SAE) deep learning architecture for automated tissue characterization of air-trapping from CXR. To our best knowledge this is the first study applying deep learning framework for the specific problem on 51 CXRs, an F-score of ≈ 76.5% and a strong correlation with the expert visual scoring (R=0.93, p =<; 0.01) demonstrate the potential of the proposed method to characterization of air trapping.

A Deep Convolutional Neural Network-Based Framework for Automatic Fetal Facial Standard Plane Recognition

Ultrasound imaging has become a prevalent examination method in prenatal diagnosis. Accurate acquisition of fetal facial standard plane (FFSP) is the most important precondition for subsequent diagnosis and measurement. In the past few years, considerable effort has been devoted to FFSP recognition using various hand-crafted features, but the recognition performance is still unsatisfactory due to the high intraclass variation of FFSPs and the high degree of visual similarity between FFSPs and other non-FFSPs. To improve the recognition performance, we propose a method to automatically recognize FFSP via a deep convolutional neural network (DCNN) architecture. The proposed DCNN consists of 16 convolutional layers with small 3 × 3 size kernels and three fully connected layers. A global average pooling is adopted in the last pooling layer to significantly reduce network parameters, which alleviates the overfitting problems and improves the performance under limited training data. Both the transfer learning strategy and a data augmentation technique tailored for FFSP are implemented to further boost the recognition performance. Extensive experiments demonstrate the advantage of our proposed method over traditional approaches and the effectiveness of DCNN to recognize FFSP for clinical diagnosis.

Towards automatic pulmonary nodule management in lung cancer screening with deep learning

The introduction of lung cancer screening programs will produce an unprecedented amount of chest CT scans in the near future, which radiologists will have to read in order to decide on a patient follow-up strategy. According to the current guidelines, the workup of screen-detected nodules strongly relies on nodule size and nodule type. In this paper, we present a deep learning system based on multi-stream multi-scale convolutional networks, which automatically classifies all nodule types relevant for nodule workup. The system processes raw CT data containing a nodule without the need for any additional information such as nodule segmentation or nodule size and learns a representation of 3D data by analyzing an arbitrary number of 2D views of a given nodule. The deep learning system was trained with data from the Italian MILD screening trial and validated on an independent set of data from the Danish DLCST screening trial. We analyze the advantage of processing nodules at multiple scales with a multi-stream convolutional network architecture, and we show that the proposed deep learning system achieves performance at classifying nodule type that surpasses the one of classical machine learning approaches and is within the inter-observer variability among four experienced human observers.

Automated Melanoma Recognition in Dermoscopy Images via Very Deep Residual Networks

Automated melanoma recognition in dermoscopy images is a very challenging task due to the low contrast of skin lesions, the huge intraclass variation of melanomas, the high degree of visual similarity between melanoma and non-melanoma lesions, and the existence of many artifacts in the image. In order to meet these challenges, we propose a novel method for melanoma recognition by leveraging very deep convolutional neural networks (CNNs). Compared with existing methods employing either low-level hand-crafted features or CNNs with shallower architectures, our substantially deeper networks (more than 50 layers) can acquire richer and more discriminative features for more accurate recognition. To take full advantage of very deep networks, we propose a set of schemes to ensure effective training and learning under limited training data. First, we apply the residual learning to cope with the degradation and overfitting problems when a network goes deeper. This technique can ensure that our networks benefit from the performance gains achieved by increasing network depth. Then, we construct a fully convolutional residual network (FCRN) for accurate skin lesion segmentation, and further enhance its capability by incorporating a multi-scale contextual information integration scheme. Finally, we seamlessly integrate the proposed FCRN (for segmentation) and other very deep residual networks (for classification) to form a two-stage framework. This framework enables the classification network to extract more representative and specific features based on segmented results instead of the whole dermoscopy images, further alleviating the insufficiency of training data. The proposed framework is extensively evaluated on ISBI 2016 Skin Lesion Analysis Towards Melanoma Detection Challenge dataset. Experimental results demonstrate the significant performance gains of the proposed framework, ranking the first in classification and the second in segmentation among 25 teams and 28 teams, respectively. This study corroborates that very deep CNNs with effective training mechanisms can be employed to solve complicated medical image analysis tasks, even with limited training data.

Multiscale Rotation-Invariant Convolutional Neural Networks for Lung Texture Classification

We propose a new multiscale rotation-invariant convolutional neural network (MRCNN) model for classifying various lung tissue types on high-resolution computed tomography. MRCNN employs Gabor-local binary pattern that introduces a good property in image analysis-invariance to image scales and rotations. In addition, we offer an approach to deal with the problems caused by imbalanced number of samples between different classes in most of the existing works, accomplished by changing the overlapping size between the adjacent patches. Experimental results on a public interstitial lung disease database show a superior performance of the proposed method to state of the art.

DeepInfer: Open-Source Deep Learning Deployment Toolkit for Image-Guided Therapy

Deep learning models have outperformed some of the previous state-of-the-art approaches in medical image analysis. Instead of using hand-engineered features, deep models attempt to automatically extract hierarchical representations at multiple levels of abstraction from the data. Therefore, deep models are usually considered to be more flexible and robust solutions for image analysis problems compared to conventional computer vision models. They have demonstrated significant improvements in computer-aided diagnosis and automatic medical image analysis applied to such tasks as image segmentation, classification and registration. However, deploying deep learning models often has a steep learning curve and requires detailed knowledge of various software packages. Thus, many deep models have not been integrated into the clinical research workflows causing a gap between the state-of-the-art machine learning in medical applications and evaluation in clinical research procedures. In this paper, we propose "DeepInfer" - an open-source toolkit for developing and deploying deep learning models within the 3D Slicer medical image analysis platform. Utilizing a repository of task-specific models, DeepInfer allows clinical researchers and biomedical engineers to deploy a trained model selected from the public registry, and apply it to new data without the need for software development or configuration. As two practical use cases, we demonstrate the application of DeepInfer in prostate segmentation for targeted MRI-guided biopsy and identification of the target plane in 3D ultrasound for spinal injections.

Marginal Shape Deep Learning: Applications to Pediatric Lung Field Segmentation

Representation learning through deep learning (DL) architecture has shown tremendous potential for identification, localization, and texture classification in various medical imaging modalities. However, DL applications to segmentation of objects especially to deformable objects are rather limited and mostly restricted to pixel classification. In this work, we propose marginal shape deep learning (MaShDL), a framework that extends the application of DL to deformable shape segmentation by using deep classifiers to estimate the shape parameters. MaShDL combines the strength of statistical shape models with the automated feature learning architecture of DL. Unlike the iterative shape parameters estimation approach of classical shape models that often leads to a local minima, the proposed framework is robust to local minima optimization and illumination changes. Furthermore, since the direct application of DL framework to a multi-parameter estimation problem results in a very high complexity, our framework provides an excellent run-time performance solution by independently learning shape parameter classifiers in marginal eigenspaces in the decreasing order of variation. We evaluated MaShDL for segmenting the lung field from 314 normal and abnormal pediatric chest radiographs and obtained a mean Dice similarity coefficient of 0.927 using only the four highest modes of variation (compared to 0.888 with classical ASM1 (p-value=0.01) using same configuration). To the best of our knowledge this is the first demonstration of using DL framework for parametrized shape learning for the delineation of deformable objects.

Segmentation of Fetal Left Ventricle in Echocardiographic Sequences Based on Dynamic Convolutional Neural Networks

Segmentation of fetal left ventricle (LV) in echocardiographic sequences is important for further quantitative analysis of fetal cardiac function. However, image gross inhomogeneities and fetal random movements make the segmentation a challenging problem. In this paper, a dynamic convolutional neural networks (CNN) based on multiscale information and fine-tuning is proposed for fetal LV segmentation. The CNN is pretrained by amount of labeled training data. In the segmentation, the first frame of each echocardiographic sequence is delineated manually. The dynamic CNN is fine-tuned by deep tuning with the first frame and shallow tuning with the rest of frames, respectively, to adapt to the individual fetus. Additionally, to separate the connection region between LV and left atrium (LA), a matching approach, which consists of block matching and line matching, is used for mitral valve (MV) base points tracking. Advantages of our proposed method are compared with an active contour model (ACM), a dynamical appearance model (DAM), and a fixed multiscale CNN method. Experimental results in 51 echocardiographic sequences show that the segmentation results agree well with the ground truth, especially in the cases with leakage, blurry boundaries, and subject-to-subject variations. The CNN architecture can be simple, and the dynamic fine-tuning is efficient.

Deep Learning for Health Informatics

With a massive influx of multimodality data, the role of data analytics in health informatics has grown rapidly in the last decade. This has also prompted increasing interests in the generation of analytical, data driven models based on machine learning in health informatics. Deep learning, a technique with its foundation in artificial neural networks, is emerging in recent years as a powerful tool for machine learning, promising to reshape the future of artificial intelligence. Rapid improvements in computational power, fast data storage, and parallelization have also contributed to the rapid uptake of the technology in addition to its predictive power and ability to generate automatically optimized high-level features and semantic interpretation from the input data. This article presents a comprehensive up-to-date review of research employing deep learning in health informatics, providing a critical analysis of the relative merit, and potential pitfalls of the technique as well as its future outlook. The paper mainly focuses on key applications of deep learning in the fields of translational bioinformatics, medical imaging, pervasive sensing, medical informatics, and public health.

Integrating Online and Offline Three-Dimensional Deep Learning for Automated Polyp Detection in Colonoscopy Videos

Automated polyp detection in colonoscopy videos has been demonstrated to be a promising way for colorectal cancer prevention and diagnosis. Traditional manual screening is time consuming, operator dependent, and error prone; hence, automated detection approach is highly demanded in clinical practice. However, automated polyp detection is very challenging due to high intraclass variations in polyp size, color, shape, and texture, and low interclass variations between polyps and hard mimics. In this paper, we propose a novel offline and online three-dimensional (3-D) deep learning integration framework by leveraging the 3-D fully convolutional network (3D-FCN) to tackle this challenging problem. Compared with the previous methods employing hand-crafted features or 2-D convolutional neural network, the 3D-FCN is capable of learning more representative spatio-temporal features from colonoscopy videos, and hence has more powerful discrimination capability. More importantly, we propose a novel online learning scheme to deal with the problem of limited training data by harnessing the specific information of an input video in the learning process. We integrate offline and online learning to effectively reduce the number of false positives generated by the offline network and further improve the detection performance. Extensive experiments on the dataset of MICCAI 2015 Challenge on Polyp Detection demonstrated the better performance of our method when compared with other competitors.

Multisource Transfer Learning With Convolutional Neural Networks for Lung Pattern Analysis

Early diagnosis of interstitial lung diseases is crucial for their treatment, but even experienced physicians find it difficult, as their clinical manifestations are similar. In order to assist with the diagnosis, computer-aided diagnosis systems have been developed. These commonly rely on a fixed scale classifier that scans CT images, recognizes textural lung patterns, and generates a map of pathologies. In a previous study, we proposed a method for classifying lung tissue patterns using a deep convolutional neural network (CNN), with an architecture designed for the specific problem. In this study, we present an improved method for training the proposed network by transferring knowledge from the similar domain of general texture classification. Six publicly available texture databases are used to pretrain networks with the proposed architecture, which are then fine-tuned on the lung tissue data. The resulting CNNs are combined in an ensemble and their fused knowledge is compressed back to a network with the original architecture. The proposed approach resulted in an absolute increase of about 2% in the performance of the proposed CNN. The results demonstrate the potential of transfer learning in the field of medical image analysis, indicate the textural nature of the problem and show that the method used for training a network can be as important as designing its architecture.

An Ensemble of Fine-Tuned Convolutional Neural Networks for Medical Image Classification

The availability of medical imaging data from clinical archives, research literature, and clinical manuals, coupled with recent advances in computer vision offer the opportunity for image-based diagnosis, teaching, and biomedical research. However, the content and semantics of an image can vary depending on its modality and as such the identification of image modality is an important preliminary step. The key challenge for automatically classifying the modality of a medical image is due to the visual characteristics of different modalities: some are visually distinct while others may have only subtle differences. This challenge is compounded by variations in the appearance of images based on the diseases depicted and a lack of sufficient training data for some modalities. In this paper, we introduce a new method for classifying medical images that uses an ensemble of different convolutional neural network (CNN) architectures. CNNs are a state-of-the-art image classification technique that learns the optimal image features for a given classification task. We hypothesise that different CNN architectures learn different levels of semantic image representation and thus an ensemble of CNNs will enable higher quality features to be extracted. Our method develops a new feature extractor by fine-tuning CNNs that have been initialized on a large dataset of natural images. The fine-tuning process leverages the generic image features from natural images that are fundamental for all images and optimizes them for the variety of medical imaging modalities. These features are used to train numerous multiclass classifiers whose posterior probabilities are fused to predict the modalities of unseen images. Our experiments on the ImageCLEF 2016 medical image public dataset (30 modalities; 6776 training images, and 4166 test images) show that our ensemble of fine-tuned CNNs achieves a higher accuracy than established CNNs. Our ensemble also achieves a higher accuracy than methods in the literature evaluated on the same benchmark dataset and is only overtaken by those methods that source additional training data.

A Convolutional Neural Network for Automatic Characterization of Plaque Composition in Carotid Ultrasound

Characterization of carotid plaque composition, more specifically the amount of lipid core, fibrous tissue, and calcified tissue, is an important task for the identification of plaques that are prone to rupture, and thus for early risk estimation of cardiovascular and cerebrovascular events. Due to its low costs and wide availability, carotid ultrasound has the potential to become the modality of choice for plaque characterization in clinical practice. However, its significant image noise, coupled with the small size of the plaques and their complex appearance, makes it difficult for automated techniques to discriminate between the different plaque constituents. In this paper, we propose to address this challenging problem by exploiting the unique capabilities of the emerging deep learning framework. More specifically, and unlike existing works which require a priori definition of specific imaging features or thresholding values, we propose to build a convolutional neural network (CNN) that will automatically extract from the images the information that is optimal for the identification of the different plaque constituents. We used approximately 90 000 patches extracted from a database of images and corresponding expert plaque characterizations to train and to validate the proposed CNN. The results of cross-validation experiments show a correlation of about 0.90 with the clinical assessment for the estimation of lipid core, fibrous cap, and calcified tissue areas, indicating the potential of deep learning for the challenging task of automatic characterization of plaque composition in carotid ultrasound.

Classification of teeth in cone-beam CT using deep convolutional neural network

Dental records play an important role in forensic identification. To this end, postmortem dental findings and teeth conditions are recorded in a dental chart and compared with those of antemortem records. However, most dentists are inexperienced at recording the dental chart for corpses, and it is a physically and mentally laborious task, especially in large scale disasters. Our goal is to automate the dental filing process by using dental x-ray images. In this study, we investigated the application of a deep convolutional neural network (DCNN) for classifying tooth types on dental cone-beam computed tomography (CT) images. Regions of interest (ROIs) including single teeth were extracted from CT slices. Fifty two CT volumes were randomly divided into 42 training and 10 test cases, and the ROIs obtained from the training cases were used for training the DCNN. For examining the sampling effect, random sampling was performed 3 times, and training and testing were repeated. We used the AlexNet network architecture provided in the Caffe framework, which consists of 5 convolution layers, 3 pooling layers, and 2 full connection layers. For reducing the overtraining effect, we augmented the data by image rotation and intensity transformation. The test ROIs were classified into 7 tooth types by the trained network. The average classification accuracy using the augmented training data by image rotation and intensity transformation was 88.8%. Compared with the result without data augmentation, data augmentation resulted in an approximately 5% improvement in classification accuracy. This indicates that the further improvement can be expected by expanding the CT dataset. Unlike the conventional methods, the proposed method is advantageous in obtaining high classification accuracy without the need for precise tooth segmentation. The proposed tooth classification method can be useful in automatic filing of dental charts for forensic identification.