Deep learning based classification of facial dermatological disorders

Horse Herd Optimization with Gate Recurrent Unit for an Automatic Classification of Different Facial Skin Disease

The human body's largest organ is the skin which covers the entire body. The facial skin is one area of the body that needs careful handling. It can cause several facial skin diseases like acne, eczema, moles, melanoma, rosacea, and many other fungal infections. Diagnosing these diseases has been difficult due to challenges like the high cost of medical equipment and the lack of medical competence. However, various existing systems are utilized to detect the type of facial skin disease, but those approaches are time-consuming and inaccurate to detect the disease at early stages. To address various issues, a deep learning-based gate recurrent unit (GRU) has been developed. Non-linear diffusion is used to acquire and pre-process raw pictures, adaptive histogram equalization (AHE) and high boost filtering (HBF). The image noise is removed by using non-linear diffusion. The contrast of the image is maximized using AHE. The image's edges are sharpened by using HBF. After pre-processing, textural and colour features are extracted by applying a grey level run-length matrix (GLRM) and chromatic co-occurrence local binary pattern (CCoLBP). Then, appropriate features are selected using horse herd optimization (HOA). Finally, selected features are classified using GRU to identify the types of facial skin disease. The proposed model is investigated using the Kaggle database that consists of different face skin disease images such as rosacea, eczema, basal cell carcinoma, acnitic keratosis, and acne. Further, the acquired dataset is split into training and testing. Considering the investigation's findings, the proposed method yields 98.2% accuracy, 1.8% error, 97.1% precision, and 95.5% f1-score. In comparison to other current techniques, the proposed technique performs better. The created model is, therefore, the best choice for classifying the various facial skin conditions.

Deep Learning Based Classification of Dermatological Disorders

Automated medical diagnosis has become crucial and significantly supports medical doctors. Thus, there is a demand for inventing deep learning (DL) and convolutional networks for analyzing medical images. Dermatology, in particular, is one of the domains that was recently targeted by AI specialists to introduce new DL algorithms or enhance convolutional neural network (CNN) architectures. A significantly high proportion of studies in the field are concerned with skin cancer, whereas other dermatological disorders are still limited. In this work, we examined the performance of 6 CNN architectures named VGG16, EfficientNet, InceptionV3, MobileNet, NasNet, and ResNet50 for the top 3 dermatological disorders that frequently appear in the Middle East. An Image filtering and denoising were imposed in this work to enhance image quality and increase architecture performance. Experimental results revealed that MobileNet achieved the highest performance and accuracy among the CNN architectures and can classify disorder with high performance (95.7% accuracy). Future scope will focus more on proposing a new methodology for deep-based classification. In addition, we will expand the dataset for more images that consider new disorders and variations.

Skin lesion classification based on two-modal images using a multi-scale fully-shared fusion network

BACKGROUND AND OBJECTIVE

Due to the complexity of skin lesion features, computer-aided diagnosis of skin diseases based on multi-modal images is considered a challenging task. Dermoscopic images and clinical images are commonly used to diagnose skin diseases in clinical scenarios, and the complementarity of their features promotes the research of multi-modality classification in the computer-aided diagnosis field. Most current methods focus on the fusion between modalities and ignore the complementary information within each of them, which leads to the loss of the intra-modality relation. Multi-modality models for integrating features both within single modalities and across multiple modalities are limited in the literature. Therefore, a multi-modality model based on dermoscopic and clinical images is proposed to address this issue.

METHODS

We propose a Multi-scale Fully-shared Fusion Network (MFF-Net) that gathers features of dermoscopic images and clinical images for skin lesion classification. In MFF-Net, the multi-scale fusion structure combines deep and shallow features within individual modalities to reduce the loss of spatial information in high-level feature maps. Then Dermo-Clinical Block (DCB) integrates the feature maps from dermoscopic images and clinical images through channel-wise concatenation and using a fully-shared fusion strategy that explores complementary information at different stages.

RESULTS

We validated our model on a four-class two-modal skin diseases dataset, and proved that the proposed multi-scale structure, the fusion module DCBs, and the fully-shared fusion strategy improve the performance of MFF-Net independently. Our method achieved the highest average accuracy of 72.9% on the 7-point checklist dataset, outperforming the state-of-the-art single-modality and multi-modality methods with an accuracy boost of 7.1% and 3.4%, respectively.

CONCLUSIONS

The multi-scale fusion structure demonstrates the significance of intra-modality relations between clinical images and dermoscopic images. The proposed network combined with the multi-scale structure, DCBs, and the fully-shared fusion strategy, can effectively integrate the features of the skin lesions across the two modalities and achieved a promising accuracy among different skin diseases.

A Novel Multi-Task Learning Network Based on Melanoma Segmentation and Classification with Skin Lesion Images

Melanoma is known worldwide as a malignant tumor and the fastest-growing skin cancer type. It is a very life-threatening disease with a high mortality rate. Automatic melanoma detection improves the early detection of the disease and the survival rate. In accordance with this purpose, we presented a multi-task learning approach based on melanoma recognition with dermoscopy images. Firstly, an effective pre-processing approach based on max pooling, contrast, and shape filters is used to eliminate hair details and to perform image enhancement operations. Next, the lesion region was segmented with a VGGNet model-based FCN Layer architecture using enhanced images. Later, a cropping process was performed for the detected lesions. Then, the cropped images were converted to the input size of the classifier model using the very deep super-resolution neural network approach, and the decrease in image resolution was minimized. Finally, a deep learning network approach based on pre-trained convolutional neural networks was developed for melanoma classification. We used the International Skin Imaging Collaboration, a publicly available dermoscopic skin lesion dataset in experimental studies. While the performance measures of accuracy, specificity, precision, and sensitivity, obtained for segmentation of the lesion region, were produced at rates of 96.99%, 92.53%, 97.65%, and 98.41%, respectively, the performance measures achieved rates for classification of 97.73%, 99.83%, 99.83%, and 95.67%, respectively.

FACES: A Deep-Learning-Based Parametric Model to Improve Rosacea Diagnoses

Rosacea is a chronic inflammatory skin disorder that causes visible blood vessels and redness on the nose, chin, cheeks, and forehead. However, visual assessment, the current standard method used to identify rosacea, is often subjective among clinicians and results in high variation. Recent advances in artificial intelligence have allowed for the effective detection of various skin diseases with high accuracy and consistency. In this study, we develop a new methodology, coined "five accurate CNNs-based evaluation system (FACES)", to identify and classify rosacea more efficiently. First, 19 CNN-based models that have been widely used for image classification were trained and tested via training and validation data sets. Next, the five best performing models were selected based on accuracy, which served as a weight value for FACES. At the same time, we also applied a majority rule to five selected models to detect rosacea. The results exhibited that the performance of FACES was superior to that of the five individual CNN-based models and the majority rule in terms of accuracy, sensitivity, specificity, and precision. In particular, the accuracy and sensitivity of FACES were the highest, and the specificity and precision were higher than most of the individual models. To improve the performance of our system, future studies must consider patient details, such as age, gender, and race, and perform comparison tests between our model system and clinicians.

Research on melanoma image segmentation by incorporating medical prior knowledge

BACKGROUND

Melanoma image segmentation has important clinical value in the diagnosis and treatment of skin diseases. However, due to the difficulty of obtaining data sets, and the sample imbalance, the quality of melanoma image data sets is low, which reduces the accuracy and the effectiveness of computer aided diagnosis of melanoma image.

OBJECTIVE

In this work, a method of melanoma image segmentation by incorporating medical prior knowledge is proposed to improve the fidelity of melanoma image segmentation.

METHODS

Anatomical analysis of the melanoma image reveal the star shape of the melanoma image, which can be encoded into the loss function of the UNet model as a prior knowledge.

RESULTS

Our experimental results on the ISIC-2017 data set demonstrate that the model by incorporating medical prior knowledge obtain a mIoU (Mean Intersection over Union) of 87.41%, a Dice Similarity Coefficient of 93.49%.

CONCLUSION

Therefore, the model by incorporating medical prior knowledge achieve the first rank in the segmentation task comparing to other models and has high clinical value.

Reducing variations in multi-center Alzheimer's disease classification with convolutional adversarial autoencoder

Based on brain magnetic resonance imaging (MRI), multiple variations ranging from MRI scanners to center-specific parameter settings, imaging protocols, and brain region-of-interest (ROI) definitions pose a big challenge for multi-center Alzheimer's disease characterization and classification. Existing approaches to reduce such variations require intricate multi-step, often manual preprocessing pipelines, including skull stripping, segmentation, registration, cortical reconstruction, and ROI outlining. Such procedures are time-consuming, and more importantly, tend to be user biased. Contrasting costly and biased preprocessing pipelines, the question arises whether we can design a deep learning model to automatically reduce these variations from multiple centers for Alzheimer's disease classification? In this study, we used T1 and T2-weighted structural MRI from Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset based on three groups with 375 subjects, respectively: patients with Alzheimer's disease (AD) dementia, with mild cognitive impairment (MCI), and healthy controls (HC); to test our approach, we defined AD classification as classifying an individual's structural image to one of the three group labels. We first introduced a convolutional adversarial autoencoder (CAAE) to reduce the variations existing in multi-center raw MRI scans by automatically registering them into a common aligned space. Afterward, a convolutional residual soft attention network (CRAT) was further proposed for AD classification. Canonical classification procedures demonstrated that our model achieved classification accuracies of 91.8%, 90.05%, and 88.10% for the 2-way classification tasks using the RAW aligned MRI scans, including AD vs. HC, AD vs. MCI, and MCI vs. HC, respectively. Thus, our automated approach achieves comparable or even better classification performance by comparing it with many baselines with dedicated conventional preprocessing pipelines. Furthermore, the uncovered brain hotpots, i.e., hippocampus, amygdala, and temporal pole, are consistent with previous studies.

Recognition of cervical precancerous lesions based on probability distribution feature guidance

INTRODUCTION:

Cervical cancer is a high incidence of cancer in women and cervical precancerous screening plays an important role in reducing the mortality rate.

METHOD:

- In this study, we proposed a multichannel feature extraction method based on the probability distribution features of the acetowhite (AW) region to identify cervical precancerous lesions, with the overarching goal to improve the accuracy of cervical precancerous screening. A k-means clustering algorithm was first used to extract the cervical region images from the original colposcopy images. We then used a deep learning model called DeepLab V3+ to segment the AW region of the cervical image after the acetic acid experiment, from which the probability distribution map of the AW region after segmentation was obtained. This probability distribution map was fed into a neural network classification model for multichannel feature extraction, which resulted in the final classification performance.

RESULT:

Results of the experimental evaluation showed that the proposed method achieved an average accuracy of 87.7%, an average sensitivity of 89.3%, and an average specificity of 85.6%. Compared with the methods that did not add segmented probability features, the proposed method increased the average accuracy rate, sensitivity, and specificity by 8.3%, 8%, and 8.4%, respectively.

CONCLUSION:

Overall, the proposed method holds great promise for enhancing the screening of cervical precancerous lesions in the clinic by providing the physician with more reliable screening results that might reduce their workload.

Weakly Supervised Segmentation of COVID19 Infection with Scribble Annotation on CT Images

Segmentation of infections from CT scans is important for accurate diagnosis and follow-up in tackling the COVID-19. Although the convolutional neural network has great potential to automate the segmentation task, most existing deep learning-based infection segmentation methods require fully annotated ground-truth labels for training, which is time-consuming and labor-intensive. This paper proposed a novel weakly supervised segmentation method for COVID-19 infections in CT slices, which only requires scribble supervision and is enhanced with the uncertainty-aware self-ensembling and transformation-consistent techniques. Specifically, to deal with the difficulty caused by the shortage of supervision, an uncertainty-aware mean teacher is incorporated into the scribble-based segmentation method, encouraging the segmentation predictions to be consistent under different perturbations for an input image. This mean teacher model can guide the student model to be trained using information in images without requiring manual annotations. On the other hand, considering the output of the mean teacher contains both correct and unreliable predictions, equally treating each prediction in the teacher model may degrade the performance of the student network. To alleviate this problem, the pixel level uncertainty measure on the predictions of the teacher model is calculated, and then the student model is only guided by reliable predictions from the teacher model. To further regularize the network, a transformation-consistent strategy is also incorporated, which requires the prediction to follow the same transformation if a transform is performed on an input image of the network. The proposed method has been evaluated on two public datasets and one local dataset. The experimental results demonstrate that the proposed method is more effective than other weakly supervised methods and achieves similar performance as those fully supervised.

Application of Artificial Intelligence in Diagnosis of Craniopharyngioma

Craniopharyngioma is a congenital brain tumor with clinical characteristics of hypothalamic-pituitary dysfunction, increased intracranial pressure, and visual field disorder, among other injuries. Its clinical diagnosis mainly depends on radiological examinations (such as Computed Tomography, Magnetic Resonance Imaging). However, assessing numerous radiological images manually is a challenging task, and the experience of doctors has a great influence on the diagnosis result. The development of artificial intelligence has brought about a great transformation in the clinical diagnosis of craniopharyngioma. This study reviewed the application of artificial intelligence technology in the clinical diagnosis of craniopharyngioma from the aspects of differential classification, prediction of tissue invasion and gene mutation, prognosis prediction, and so on. Based on the reviews, the technical route of intelligent diagnosis based on the traditional machine learning model and deep learning model were further proposed. Additionally, in terms of the limitations and possibilities of the development of artificial intelligence in craniopharyngioma diagnosis, this study discussed the attentions required in future research, including few-shot learning, imbalanced data set, semi-supervised models, and multi-omics fusion.

New Trends in Melanoma Detection Using Neural Networks: A Systematic Review

Due to its increasing incidence, skin cancer, and especially melanoma, is a serious health disease today. The high mortality rate associated with melanoma makes it necessary to detect the early stages to be treated urgently and properly. This is the reason why many researchers in this domain wanted to obtain accurate computer-aided diagnosis systems to assist in the early detection and diagnosis of such diseases. The paper presents a systematic review of recent advances in an area of increased interest for cancer prediction, with a focus on a comparative perspective of melanoma detection using artificial intelligence, especially neural network-based systems. Such structures can be considered intelligent support systems for dermatologists. Theoretical and applied contributions were investigated in the new development trends of multiple neural network architecture, based on decision fusion. The most representative articles covering the area of melanoma detection based on neural networks, published in journals and impact conferences, were investigated between 2015 and 2021, focusing on the interval 2018-2021 as new trends. Additionally presented are the main databases and trends in their use in teaching neural networks to detect melanomas. Finally, a research agenda was highlighted to advance the field towards the new trends.

An improved residual U-Net with morphological-based loss function for automatic liver segmentation in computed tomography

This paper proposes an improved ResU-Net framework for automatic liver CT segmentation. By employing a new loss function and data augmentation strategy, the accuracy of liver segmentation is improved, and the performance is verified on two public datasets LiTS17 and SLiver07. Firstly, to speed up the convergence of the model, the residual module is used to replace the original convolution module of U-Net. Secondly, to suppress the problem of pixel imbalance, the opposite number of Dice is proposed to replace the cross-entropy loss function, and the morphological method is introduced to weigh the pixels. Finally, to improve the generalization ability of the model, random affine transformation and random elastic deformation are employed for data augmentation. From 20 training datasets of Sliver07, 16 sets were selected as the training set, two sets were used for verification, and two sets were used for the test; meanwhile, from 131 training datasets of LiTS2017, eight sets were selected as the test set. In the experiment, four evaluation metrics, including DICE global, DICE per case, VOE, and RVD, were calculated, with the accuracies of 94.28, 94.24 ± 2.07, 10.83 ± 3.70, and -0.25 ± 2.74, respectively. Compared with U-Net and ResU-Net, the performance of the proposed method is significantly improved. The experimental results show that, although the method's complexity is high, it has a faster convergence speed and stronger generalization ability. The segmentation effect on the 2D image is significantly improved, and the scalability on 3D data is also robust. In addition, the proposed method performs well in the case of low-contrast neighboring organs, which proves the robustness of the proposed method.

Ranking convolutional neural network for Alzheimer's disease mini-mental state examination prediction at multiple time-points

BACKGROUND AND OBJECTIVE

Alzheimer's disease (AD) is a fatal neurodegenerative disease. Predicting Mini-mental state examination (MMSE) based on magnetic resonance imaging (MRI) plays an important role in monitoring the progress of AD. Existing machine learning based methods cast MMSE prediction as a single metric regression problem simply and ignore the relationship between subjects with various scores.

METHODS

In this study, we proposed a ranking convolutional neural network (rankCNN) to address the prediction of MMSE through muti-classification. Specifically, we use a 3D convolutional neural network with sharing weights to extract the feature from MRI, followed by multiple sub-networks which transform the cognitive regression into a series of simpler binary classification. In addition, we further use a ranking layer to measure the ranking information between samples to strengthen the ability of the classification by extracting more discriminative features.

RESULTS

We evaluated the proposed model on ADNI-1 and ADNI-2 datasets with a total of 1,569 subjects. The Root Mean Squared Error (RMSE) of our proposed model at baseline is 2.238 and 2.434 on ADNI-1 and ADNI-2, respectively. Extensive experimental results on ADNI-1 and ADNI-2 datasets demonstrate that our proposed model is superior to several state-of-the-art methods at both baseline and future MMSE prediction of subjects.

CONCLUSION

This paper provides a new method that can effectively predict the MMSE at baseline and future time points using baseline MRI, making it possible to use MRI for accurate early diagnosis of AD. The source code is freely available at https://github.com/fengduqianhe/ADrankCNN-master.

A novel deep learning model DDU-net using edge features to enhance brain tumor segmentation on MR images

Glioma is a relatively common brain tumor disease with high mortality rate. Humans have been seeking a more effective therapy. In the course of treatment, the specific location of the tumor needs to be determined first in any case. Therefore, how to segment tumors from brain tissue accurately and quickly is a persistent problem. In this paper, a new dual-stream decoding CNN architecture combined with U-net for automatic segmentation of brain tumor on MR images namely DDU-net is proposed. Two edge-based optimization strategies are used to enhance the performance of brain tumor segmentation. First, we design a separate branch to process edge stream information. Here, high level edge features are reduced in dimension of channel and integrated into the conventional semantic stream in the way of residual. Second, a regularization loss function is used to encourage the predicted segmentation mask to align with ground truth around the edge mainly by penalizing pixels where the predicted segmentation masks and labels do not match around the edge. In training, we employ a novel edge extraction algorithm for providing edge labels with higher quality. Moreover, we add a self-adaptive balancing class weight coefficient into the cross entropy loss function for solving the serious class imbalance problem in the backpropagation of edge extraction. Our experiments show that this leads to a very efficient architecture which can produce clearer prediction at the edge of the tumor. Our method achieves ideal performance on BraTS2017 and BraTS2018 in terms of Dice coefficient.

Deep learning-based facial image analysis in medical research: a systematic review protocol

INTRODUCTION

Deep learning techniques are gaining momentum in medical research. Evidence shows that deep learning has advantages over humans in image identification and classification, such as facial image analysis in detecting people's medical conditions. While positive findings are available, little is known about the state-of-the-art of deep learning-based facial image analysis in the medical context. For the consideration of patients' welfare and the development of the practice, a timely understanding of the challenges and opportunities faced by research on deep-learning-based facial image analysis is needed. To address this gap, we aim to conduct a systematic review to identify the characteristics and effects of deep learning-based facial image analysis in medical research. Insights gained from this systematic review will provide a much-needed understanding of the characteristics, challenges, as well as opportunities in deep learning-based facial image analysis applied in the contexts of disease detection, diagnosis and prognosis.

METHODS

Databases including PubMed, PsycINFO, CINAHL, IEEEXplore and Scopus will be searched for relevant studies published in English in September, 2021. Titles, abstracts and full-text articles will be screened to identify eligible articles. A manual search of the reference lists of the included articles will also be conducted. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses framework was adopted to guide the systematic review process. Two reviewers will independently examine the citations and select studies for inclusion. Discrepancies will be resolved by group discussions till a consensus is reached. Data will be extracted based on the research objective and selection criteria adopted in this study.

ETHICS AND DISSEMINATION

As the study is a protocol for a systematic review, ethical approval is not required. The study findings will be disseminated via peer-reviewed publications and conference presentations.

PROSPERO REGISTRATION NUMBER

CRD42020196473.

Deep attention branch networks for skin lesion classification

BACKGROUND AND OBJECTIVE

The skin lesion usually covers a small region of the dermoscopy image, and the lesions of different categories might own high similarities. Therefore, it is essential to design an elaborate network for accurate skin lesion classification, which can focus on semantically meaningful lesion parts. Although the Class Activation Mapping (CAM) shows good localization capability of highlighting the discriminative parts, it cannot be obtained in the forward propagation process.

METHODS

We propose a Deep Attention Branch Network (DABN) model, which introduces the attention branches to expand the conventional Deep Convolutional Neural Networks (DCNN). The attention branch is designed to obtain the CAM in the training stage, which is then utilized as an attention map to make the network focus on discriminative parts of skin lesions. DABN is applicable to multiple DCNN structures and can be trained in an end-to-end manner. Moreover, a novel Entropy-guided Loss Weighting (ELW) strategy is designed to counter class imbalance influence in the skin lesion datasets.

RESULTS

The proposed method achieves an Average Precision (AP) of 0.719 on the ISIC-2016 dataset and an average area under the ROC curve (AUC) of 0.922 on the ISIC-2017 dataset. Compared with other state-of-the-art methods, our method obtains better performance without external data and ensemble learning. Moreover, extensive experiments demonstrate that it can be applied to multi-class classification tasks and improves mean sensitivity by more than 2.6% in different DCNN structures.

CONCLUSIONS

The proposed method can adaptively focus on the discriminative regions of dermoscopy images and allows for effective training when facing class imbalance, leading to the performance improvement of skin lesion classification, which could also be applied to other clinical applications.

Lack of Transparency and Potential Bias in Artificial Intelligence Data Sets and Algorithms: A Scoping Review

Importance

Clinical artificial intelligence (AI) algorithms have the potential to improve clinical care, but fair, generalizable algorithms depend on the clinical data on which they are trained and tested.

Objective

To assess whether data sets used for training diagnostic AI algorithms addressing skin disease are adequately described and to identify potential sources of bias in these data sets.

Data Sources

In this scoping review, PubMed was used to search for peer-reviewed research articles published between January 1, 2015, and November 1, 2020, with the following paired search terms: deep learning and dermatology, artificial intelligence and dermatology, deep learning and dermatologist, and artificial intelligence and dermatologist.

Study Selection

Studies that developed or tested an existing deep learning algorithm for triage, diagnosis, or monitoring using clinical or dermoscopic images of skin disease were selected, and the articles were independently reviewed by 2 investigators to verify that they met selection criteria.

Consensus Process

Data set audit criteria were determined by consensus of all authors after reviewing existing literature to highlight data set transparency and sources of bias.

Results

A total of 70 unique studies were included. Among these studies, 1 065 291 images were used to develop or test AI algorithms, of which only 257 372 (24.2%) were publicly available. Only 14 studies (20.0%) included descriptions of patient ethnicity or race in at least 1 data set used. Only 7 studies (10.0%) included any information about skin tone in at least 1 data set used. Thirty-six of the 56 studies developing new AI algorithms for cutaneous malignant neoplasms (64.3%) met the gold standard criteria for disease labeling. Public data sets were cited more often than private data sets, suggesting that public data sets contribute more to new development and benchmarks.

Conclusions and Relevance

This scoping review identified 3 issues in data sets that are used to develop and test clinical AI algorithms for skin disease that should be addressed before clinical translation: (1) sparsity of data set characterization and lack of transparency, (2) nonstandard and unverified disease labels, and (3) inability to fully assess patient diversity used for algorithm development and testing.

GCA-Net: global context attention network for intestinal wall vascular segmentation

PURPOSE

Precise segmentation of intestinal wall vessels is vital to colonic perforation prevention. However, there are interferences such as gastric juice in the vessel image of the intestinal wall, especially vessels and the mucosal folds are difficult to distinguish, which easily lead to mis-segmentation. In addition, the insufficient feature extraction of intricate vessel structures may leave out information of tiny vessels that result in rupture. To overcome these challenges, an effective network is proposed for segmentation of intestinal wall vessels.

METHODS

A global context attention network (GCA-Net) that employs a multi-scale fusion attention (MFA) module is proposed to adaptively integrate local and global context information to improve the distinguishability of mucosal folds and vessels, more importantly, the ability to capture tiny vessels. Also, a parallel decoder is used to introduce a contour loss function to solve the blurry and noisy blood vessel boundaries.

RESULTS

Extensive experimental results demonstrate the superiority of the GCA-Net, with accuracy of 94.84%, specificity of 97.89%, F1-score of 73.80%, AUC of 96.30%, and MeanIOU of 76.46% in fivefold cross-validation, exceeding the comparison methods. In addition, the public dataset DRIVE is used to verify the potential of GCA-Net in retinal vessel image segmentation.

CONCLUSION

A novel network for segmentation of intestinal wall vessels is developed, which can suppress interferences in intestinal wall vessel images, improve the discernibility of blood vessels and mucosal folds, enhance vessel boundaries, and capture tiny vessels. Comprehensive experiments prove that the proposed GCA-Net can accurately segment the intestinal wall vessels.

Semi-AI and Full-AI digitizer: The ways to digitalize visual field big data

BACKGROUND AND OBJECTIVE

Glaucoma is one of the major diseases that cause blindness, which is incurable and irreversible, and it is essential to detect glaucoma vision deficits in treatment and check the progression of vision disorders in advance. In order to minimize the risk of glaucoma, it is necessary not only to diagnose and observe glaucoma but also to predict prognosis via indicators from Visual Field (VF) tests. However, information from the VF test cannot be directly used in clinical studies because most medical institutions store VF test sheets in Portable Document Format (PDF) or image files in different standards.

METHODS

We developed AI-based real-time VF big data digitizing systems that digitalize VF test images in real-time in two ways; Semi-AI and Full-AI digitizer. The Semi-AI digitizer detects the VF text area with actual coordinates derived from mouse handler system. Full-AI digitizer detects the VF text area with Faster Region Based Convolutional Neural Networks (RCNN). After detecting the text area, both systems extract texts with Recurrent Neural Network based Optical Character Recognition. Semi-AI and Full-AI digitizer post-processes the extracted text results with in-system algorithm and out-of-system algorithm, respectively.

RESULTS

Both systems used 325,310 VF test sheets from a tertiary hospital and extracted a total of 5,530,270 texts. From the 100 randomly selected VF sheets, 3,400 texts were used for the validation. Semi-AI and Full-AI digitizer showed 0.993 and 0.983 of accuracy, respectively.

CONCLUSION

This study demonstrates the effectiveness of AI applications in detecting text areas and the different implementation methodologies of the post-processing process. In detecting text area, Semi-AI may be better than Full-AI digitizer in terms of system speed and human labor labeling if the number of types to be classified is small. However, Full-AI digitizer is recommended because it allows detecting text area regardless of resolution and size of the VF sheets, as the types of real-world VF test sheets cannot be predicted, and the types become more unpredictable when extended to multi-hospital studies. For Post-preprocessing, Semi-AI methodology is recommended because Semi-AI produced higher results with less effort and considered the convenience of researchers by implementing them as in-system.

A smart LED therapy device with an automatic facial acne vulgaris diagnosis based on deep learning and internet of things application

In low-level laser therapy, providing an optimal dosage and proposing a proper diagnosis before dermatological treatment are essential to reduce the side effects and potential dangers. In this article, a smart LED therapy system for automatic facial acne vulgaris diagnosis based on deep learning and Internet of Things application is proposed. The main goals of this study were to (1) develop an LED therapy device with different power densities and LED grid control; (2) propose a deep learning model based on modified ResNet50 and YOLOv2 for an automatic acne diagnosis; and (3) develop a smartphone application for facial photography image capture and LED therapy parameter configuration. Furthermore, a healthcare Internet of Things (H-IoT) platform for the connectivity between smartphone apps, the cloud server, and the LED therapy device is proposed to improve the efficiency of the treatment process. Experiments were conducted on test data sets divided by a cross-validation method to verify the feasibility of the proposed LED therapy system with automatic facial acne detection. The obtained results evidenced the practical application of the proposed LED therapy system for automatic acne diagnosis and H-IoT-based solutions.

Does non-COVID-19 lung lesion help? investigating transferability in COVID-19 CT image segmentation

BACKGROUND AND OBJECTIVE

Coronavirus disease 2019 (COVID-19) is a highly contagious virus spreading all around the world. Deep learning has been adopted as an effective technique to aid COVID-19 detection and segmentation from computed tomography (CT) images. The major challenge lies in the inadequate public COVID-19 datasets. Recently, transfer learning has become a widely used technique that leverages the knowledge gained while solving one problem and applying it to a different but related problem. However, it remains unclear whether various non-COVID19 lung lesions could contribute to segmenting COVID-19 infection areas and how to better conduct this transfer procedure. This paper provides a way to understand the transferability of non-COVID19 lung lesions and a better strategy to train a robust deep learning model for COVID-19 infection segmentation.

METHODS

Based on a publicly available COVID-19 CT dataset and three public non-COVID19 datasets, we evaluate four transfer learning methods using 3D U-Net as a standard encoder-decoder method. i) We introduce the multi-task learning method to get a multi-lesion pre-trained model for COVID-19 infection. ii) We propose and compare four transfer learning strategies with various performance gains and training time costs. Our proposed Hybrid-encoder Learning strategy introduces a Dedicated-encoder and an Adapted-encoder to extract COVID-19 infection features and general lung lesion features, respectively. An attention-based Selective Fusion unit is designed for dynamic feature selection and aggregation.

RESULTS

Experiments show that trained with limited data, proposed Hybrid-encoder strategy based on multi-lesion pre-trained model achieves a mean DSC, NSD, Sensitivity, F1-score, Accuracy and MCC of 0.704, 0.735, 0.682, 0.707, 0.994 and 0.716, respectively, with better genetalization and lower over-fitting risks for segmenting COVID-19 infection.

CONCLUSIONS

The results reveal the benefits of transferring knowledge from non-COVID19 lung lesions, and learning from multiple lung lesion datasets can extract more general features, leading to accurate and robust pre-trained models. We further show the capability of the encoder to learn feature representations of lung lesions, which improves segmentation accuracy and facilitates training convergence. In addition, our proposed Hybrid-encoder learning method incorporates transferred lung lesion features from non-COVID19 datasets effectively and achieves significant improvement. These findings promote new insights into transfer learning for COVID-19 CT image segmentation, which can also be further generalized to other medical tasks.

Application of Two New Feature Fusion Networks to Improve Real-time Prostate Capsula Detection

BACKGROUND

Excess prostate tissue is trimmed near the prostate capsula boundary during transurethral plasma kinetic enucleation of prostate (PKEP) and transurethral bipolar plasmakinetic resection of prostate (PKRP) surgeries. If too much tissue is removed, a prostate capsula perforation can potentially occur. As such, real-time accurate prostate capsula (PC) detection is critical for the prevention of these perforations.

OBJECTIVE

This study investigated the potential for using image denoising, image dimension reduction and feature fusion to improve real-time prostate capsula detection with two objective. First, this paper mainly studied feature selection and input dimension reduction. Second, image denoising were evaluated, as they are of paramount importance to transient stability assessment based on neural networks.

METHOD

Two new feature fusion techniques, maxpooling bilinear interpolation single-shot multibox detector (PBSSD) and bilinear interpolation single shot multibox detector (BSSD) were proposed. Before original images were sent to the neural network, they were processed by principal component analysis (PCA) and adaptive median filter (AMF) for dimension reduction and image denoising.

RESULTS

The results showed that application of PCA and AMF with PBSSD increased the mean average precision (mAP) for prostate capsula images by 8.55% and reached 80.15%, compared with single shot multibox detector (SSD) alone. Application of PCA with BSSD increased the mAP for prostate capsula images by 4.6% compared with SSD alone.

CONCLUSION

Compared with other methods, ours were proven to be more accurate for real-time prostate capsula detection. The improved mAP results suggest that the proposed approaches are powerful tools for improving SSD networks.