Deep learning model for Pathogen Classification using Feature fusion and Data augmentation

2020 ◽  
Vol 15 ◽  
Author(s):  
Fareed Ahmad ◽  
Amjad Farooq ◽  
Muhammad Usman Ghani Khan
Keyword(s):  
Deep Learning ◽  
Cross Validation ◽  
Data Augmentation ◽  
Feature Vector ◽  
Feature Fusion ◽  
High Capacity ◽  
Automatic Classification ◽  
Fine Tuning ◽  
Correct Identification ◽  
Learning Models

Background: Bacterial pathogens are deadly for animals and humans. The ease of their dissemination, coupled with their high capacity for ailment and death in infected individuals, makes them a threat to society. Objective: Due to high similarity among genera and species of pathogens, it is sometimes difficult for microbiologists to differentiate between them. Their automatic classification using deep-learning models can help in reliable, and accurate outcomes. Method: Deep-learning models, namely; AlexNet, GoogleNet, ResNet101, and InceptionV3 are used with numerous variations including training model from scratch, fine-tuning without pre-trained weights, fine-tuning along with freezing weights of initial layers, fine-tuning along with adjusting weights of all layers and augmenting the dataset by random translation and reflection. Moreover, as the dataset is small, fine-tuning and data augmentation strategies are applied to avoid overfitting and produce a generalized model. A merged feature vector is produced using two best-performing models and accuracy is calculated by xgboost algorithm on the feature vector by applying cross-validation. Results: Fine-tuned models where augmentation is applied produces the best results. Out of these, two-best-performing deep models i.e. (ResNet101, and InceptionV3) selected for feature fusion, produced a similar validation accuracy of 95.83 with a loss of 0.0213 and 0.1066, and a testing accuracy of 97.92 and 93.75, respectively. The proposed model used xgboost to attained a classification accuracy of 98.17% by using 35-folds cross-validation. Conclusion: The automatic classification using these models can help experts in the correct identification of pathogens. Consequently, they can help in controlling epidemics and thereby minimizing the socio-economic impact on the community.

scholarly journals Deep Ensemble Model for Classification of Novel Coronavirus in Chest X-Ray Images

2021 ◽  
Vol 2021 ◽  
pp. 1-17
Author(s):  
Fareed Ahmad ◽  
Amjad Farooq ◽  
Muhammad Usman Ghani
Keyword(s):  
Ensemble Learning ◽  
High Capacity ◽  
Automatic Classification ◽  
Fine Tuning ◽  
Radiographic Examination ◽  
Correct Identification ◽  
Ensemble Model ◽  
X Rays ◽  
Expert Radiologist ◽  
Novel Coronavirus

The novel coronavirus, SARS-CoV-2, can be deadly to people, causing COVID-19. The ease of its propagation, coupled with its high capacity for illness and death in infected individuals, makes it a hazard to the community. Chest X-rays are one of the most common but most difficult to interpret radiographic examination for early diagnosis of coronavirus-related infections. They carry a considerable amount of anatomical and physiological information, but it is sometimes difficult even for the expert radiologist to derive the related information they contain. Automatic classification using deep learning models can help in better assessing these infections swiftly. Deep CNN models, namely, MobileNet, ResNet50, and InceptionV3, were applied with different variations, including training the model from the start, fine-tuning along with adjusting learned weights of all layers, and fine-tuning with learned weights along with augmentation. Fine-tuning with augmentation produced the best results in pretrained models. Out of these, two best-performing models (MobileNet and InceptionV3) selected for ensemble learning produced accuracy and FScore of 95.18% and 90.34%, and 95.75% and 91.47%, respectively. The proposed hybrid ensemble model generated with the merger of these deep models produced a classification accuracy and FScore of 96.49% and 92.97%. For test dataset, which was separately kept, the model generated accuracy and FScore of 94.19% and 88.64%. Automatic classification using deep ensemble learning can help radiologists in the correct identification of coronavirus-related infections in chest X-rays. Consequently, this swift and computer-aided diagnosis can help in saving precious human lives and minimizing the social and economic impact on society.

Levenshtein Augmentation Improves Performance of SMILES Based Deep-Learning Synthesis Prediction

2020 ◽  
Author(s):  
Dean Sumner ◽  
Jiazhen He ◽  
Amol Thakkar ◽  
Ola Engkvist ◽  
Esben Jannik Bjerrum
Keyword(s):  
Neural Networks ◽  
Pattern Recognition ◽  
Deep Learning ◽  
Recurrent Neural Networks ◽  
Data Augmentation ◽  
State Of The Art ◽  
Sequence Similarity ◽  
Learning Models ◽  
Underlying Network

<p>SMILES randomization, a form of data augmentation, has previously been shown to increase the performance of deep learning models compared to non-augmented baselines. Here, we propose a novel data augmentation method we call “Levenshtein augmentation” which considers local SMILES sub-sequence similarity between reactants and their respective products when creating training pairs. The performance of Levenshtein augmentation was tested using two state of the art models - transformer and sequence-to-sequence based recurrent neural networks with attention. Levenshtein augmentation demonstrated an increase performance over non-augmented, and conventionally SMILES randomization augmented data when used for training of baseline models. Furthermore, Levenshtein augmentation seemingly results in what we define as <i>attentional gain </i>– an enhancement in the pattern recognition capabilities of the underlying network to molecular motifs.</p>

scholarly journals Deep Learning-Based Differentiation between Mucinous Cystic Neoplasm and Serous Cystic Neoplasm in the Pancreas Using Endoscopic Ultrasonography

Diagnostics ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 1052
Author(s):  
Leang Sim Nguon ◽  
Kangwon Seo ◽  
Jung-Hyun Lim ◽  
Tae-Jun Song ◽  
Sung-Hyun Cho ◽  
...  
Keyword(s):  
Decision Making ◽  
Deep Learning ◽  
Network Model ◽  
Endoscopic Ultrasonography ◽  
Data Augmentation ◽  
Clinical Information ◽  
Training Data ◽  
Fine Tuning ◽  
Cystic Neoplasm ◽  
Cystic Neoplasms

Mucinous cystic neoplasms (MCN) and serous cystic neoplasms (SCN) account for a large portion of solitary pancreatic cystic neoplasms (PCN). In this study we implemented a convolutional neural network (CNN) model using ResNet50 to differentiate between MCN and SCN. The training data were collected retrospectively from 59 MCN and 49 SCN patients from two different hospitals. Data augmentation was used to enhance the size and quality of training datasets. Fine-tuning training approaches were utilized by adopting the pre-trained model from transfer learning while training selected layers. Testing of the network was conducted by varying the endoscopic ultrasonography (EUS) image sizes and positions to evaluate the network performance for differentiation. The proposed network model achieved up to 82.75% accuracy and a 0.88 (95% CI: 0.817–0.930) area under curve (AUC) score. The performance of the implemented deep learning networks in decision-making using only EUS images is comparable to that of traditional manual decision-making using EUS images along with supporting clinical information. Gradient-weighted class activation mapping (Grad-CAM) confirmed that the network model learned the features from the cyst region accurately. This study proves the feasibility of diagnosing MCN and SCN using a deep learning network model. Further improvement using more datasets is needed.

MUFFIN: multi-scale feature fusion for drug–drug interaction prediction

2021 ◽  
Author(s):  
Yujie Chen ◽  
Tengfei Ma ◽  
Xixi Yang ◽  
Jianmin Wang ◽  
Bosheng Song ◽  
...  
Keyword(s):  
Molecular Structure ◽  
Deep Learning ◽  
Medical Information ◽  
Feature Fusion ◽  
Molecular Graph ◽  
Knowledge Graph ◽  
Sequence Information ◽  
Learning Models ◽  
Scale Feature ◽  
Multi Scale

Abstract Motivation Adverse drug–drug interactions (DDIs) are crucial for drug research and mainly cause morbidity and mortality. Thus, the identification of potential DDIs is essential for doctors, patients and the society. Existing traditional machine learning models rely heavily on handcraft features and lack generalization. Recently, the deep learning approaches that can automatically learn drug features from the molecular graph or drug-related network have improved the ability of computational models to predict unknown DDIs. However, previous works utilized large labeled data and merely considered the structure or sequence information of drugs without considering the relations or topological information between drug and other biomedical objects (e.g. gene, disease and pathway), or considered knowledge graph (KG) without considering the information from the drug molecular structure. Results Accordingly, to effectively explore the joint effect of drug molecular structure and semantic information of drugs in knowledge graph for DDI prediction, we propose a multi-scale feature fusion deep learning model named MUFFIN. MUFFIN can jointly learn the drug representation based on both the drug-self structure information and the KG with rich bio-medical information. In MUFFIN, we designed a bi-level cross strategy that includes cross- and scalar-level components to fuse multi-modal features well. MUFFIN can alleviate the restriction of limited labeled data on deep learning models by crossing the features learned from large-scale KG and drug molecular graph. We evaluated our approach on three datasets and three different tasks including binary-class, multi-class and multi-label DDI prediction tasks. The results showed that MUFFIN outperformed other state-of-the-art baselines. Availability and implementation The source code and data are available at https://github.com/xzenglab/MUFFIN.

scholarly journals "LeukNet" - A Model of Convolutional Neural Network for the Diagnosis of Leukemia

2020 ◽  
Author(s):  
Luis H. S. Vogado ◽  
Rodrigo M. S. Veras ◽  
Kelson R. T. Aires
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Cross Validation ◽  
Data Augmentation ◽  
Fine Tuning ◽  
Current State ◽  
Leukemia Diagnosis ◽  
New Guidelines ◽  
Texture Contrast ◽  
Tuning Methods

Leukemia is a disorder that affects the bone marrow, causing uncontrolled production of leukocytes, impairing the transport of oxygen and causing blood coagulation problems. In this article, we propose a new computational tool, named LeukNet, a Convolutional Neural Network (CNN) architecture based on the VGG-16 convolutional blocks, to facilitate the leukemia diagnosis from blood smear images. We evaluated different architectures and fine-tuning methods using 18 datasets containing 3536 images with distinct characteristics of color, texture, contrast, and resolution. Additionally, data augmentation operations were applied to increase the training set by up to 20 times. The k-fold cross-validation (k = 5) results achieved 98.28% of accuracy. A cross-dataset validation technique, named LeaveOne-Dataset-Out Cross-Validation (LODOCV), is also proposed to evaluate the developed model’s generalization capability. The accuracy of using LODOCV on the ALL-IDB 1, ALL-IDB 2, and UFG datasets was 97.04%, 82.46%, and 70.24%, respectively, overcoming the current state-of-the-art results and offering new guidelines for image-based computer-aided diagnosis (CAD) systems in this area.

scholarly journals Covid-19 detection via deep neural network and occlusion sensitivity maps

2021 ◽  
Author(s):  
Noor Ahmad ◽  
Muhammad Aminu ◽  
Mohd Halim Mohd Noor
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Deep Neural Network ◽  
State Of The Art ◽  
Color Images ◽  
Fine Tuning ◽  
Training Dataset ◽  
Learning Approaches ◽  
Learning Models ◽  
Sensitivity Maps

Deep learning approaches have attracted a lot of attention in the automatic detection of Covid-19 and transfer learning is the most common approach. However, majority of the pre-trained models are trained on color images, which can cause inefficiencies when fine-tuning the models on Covid-19 images which are often grayscale. To address this issue, we propose a deep learning architecture called CovidNet which requires a relatively smaller number of parameters. CovidNet accepts grayscale images as inputs and is suitable for training with limited training dataset. Experimental results show that CovidNet outperforms other state-of-the-art deep learning models for Covid-19 detection.

scholarly journals Deep Learning of COVID-19 Chest X-Rays: New Models or Fine Tuning?

2020 ◽  
Author(s):  
Tuan Pham
Keyword(s):  
Deep Learning ◽  
High Performance ◽  
Data Augmentation ◽  
Dominant Role ◽  
Care Center ◽  
Characteristic Curve ◽  
Fine Tuning ◽  
Urgent Care ◽  
X Rays ◽  
Training Time

Chest X-rays have been found to be very promising for assessing COVID-19 patients, especially for resolving emergency-department and urgent-care-center overcapacity. Deep-learning (DL) methods in artificial intelligence (AI) play a dominant role as high-performance classifiers in the detection of the disease using chest X-rays. While many new DL models have been being developed for this purpose, this study aimed to investigate the fine tuning of pretrained convolutional neural networks (CNNs) for the classification of COVID-19 using chest X-rays. Three pretrained CNNs, which are AlexNet, GoogleNet, and SqueezeNet, were selected and fine-tuned without data augmentation to carry out 2-class and 3-class classification tasks using 3 public chest X-ray databases. In comparison with other recently developed DL models, the 3 pretrained CNNs achieved very high classification results in terms of accuracy, sensitivity, specificity, precision, F1 score, and area under the receiver-operating-characteristic curve. AlexNet, GoogleNet, and SqueezeNet require the least training time among pretrained DL models, but with suitable selection of training parameters, excellent classification results can be achieved without data augmentation by these networks. The findings contribute to the urgent need for harnessing the pandemic by facilitating the deployment of AI tools that are fully automated and readily available in the public domain for rapid implementation.

scholarly journals Multi skin lesions classification using fine-tuning and data-augmentation applying NASNet

2021 ◽  
Vol 7 ◽  
pp. e371
Author(s):  
Elia Cano ◽  
José Mendoza-Avilés ◽  
Mariana Areiza ◽  
Noemi Guerra ◽  
José Longino Mendoza-Valdés ◽  
...  
Keyword(s):  
Cross Validation ◽  
Data Augmentation ◽  
Skin Diseases ◽  
Medical Image Analysis ◽  
Skin Lesions ◽  
Disease Classification ◽  
Fine Tuning ◽  
Skin Cancer Prevention ◽  
Skin Imaging ◽  
Multi Class Classification

Skin lesions are one of the typical symptoms of many diseases in humans and indicative of many types of cancer worldwide. Increased risks caused by the effects of climate change and a high cost of treatment, highlight the importance of skin cancer prevention efforts like this. The methods used to detect these diseases vary from a visual inspection performed by dermatologists to computational methods, and the latter has widely used automatic image classification applying Convolutional Neural Networks (CNNs) in medical image analysis in the last few years. This article presents an approach that uses CNNs with a NASNet architecture to recognize in a more accurate way, without segmentation, eight skin diseases. The model was trained end-to-end on Keras with augmented skin diseases images from the International Skin Imaging Collaboration (ISIC). The CNN architectures were initialized with weight from ImageNet, fine-tuned in order to discriminate well among the different types of skin lesions, and then 10-fold cross-validation was applied. Finally, some evaluation metrics are calculated as accuracy, sensitivity, and specificity and compare with other CNN trained architectures. This comparison shows that the proposed system offers higher accuracy results, with a significant reduction on the training paraments. To the best of our knowledge and based in the state-of-art recompiling in this work, the application of the NASNet architecture training with skin image lesion from ISIC archive for multi-class classification and evaluated by cross-validation, represents a novel skin disease classification system.

Data Augmentation for Improving Deep Learning Models in Building Inspections or Postdisaster Evaluation

2021 ◽  
Vol 35 (4) ◽  
Author(s):  
Samuel Leach ◽  
Yunhe Xue ◽  
Rahul Sridhar ◽  
Stephanie Paal ◽  
Zhangyang Wang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Data Augmentation ◽  
Learning Models

Abstract P319: Can Deep Learning Find the Ischemic Core on CT? Transfer Learning From Pre-Trained MRI-Based Networks

Stroke ◽  
2021 ◽  
Vol 52 (Suppl_1) ◽  
Author(s):  
Yannan Yu ◽  
Soren Christensen ◽  
Yuan Xie ◽  
Enhao Gong ◽  
Maarten G Lansberg ◽  
...  
Keyword(s):  
Deep Learning ◽  
Ground Truth ◽  
Learning Model ◽  
Fine Tuning ◽  
Learning Models ◽  
Starting Point ◽  
Stroke Lesion ◽  
Ischemic Core ◽  
Deep Learning Model

Objective: Ischemic core prediction from CT perfusion (CTP) remains inaccurate compared with gold standard diffusion-weighted imaging (DWI). We evaluated if a deep learning model to predict the DWI lesion from MR perfusion (MRP) could facilitate ischemic core prediction on CTP. Method: Using the multi-center CRISP cohort of acute ischemic stroke patient with CTP before thrombectomy, we included patients with major reperfusion (TICI score≥2b), adequate image quality, and follow-up MRI at 3-7 days. Perfusion parameters including Tmax, mean transient time, cerebral blood flow (CBF), and cerebral blood volume were reconstructed by RAPID software. Core lab experts outlined the stroke lesion on the follow-up MRI. A previously trained MRI model in a separate group of patients was used as a starting point, which used MRP parameters as input and RAPID ischemic core on DWI as ground truth. We fine-tuned this model, using CTP parameters as input, and follow-up MRI as ground truth. Another model was also trained from scratch with only CTP data. 5-fold cross validation was used. Performance of the models was compared with ischemic core (rCBF≤30%) from RAPID software to identify the presence of a large infarct (volume>70 or >100ml). Results: 94 patients in the CRISP trial met the inclusion criteria (mean age 67±15 years, 52% male, median baseline NIHSS 18, median 90-day mRS 2). Without fine-tuning, the MRI model had an agreement of 73% in infarct >70ml, and 69% in >100ml; the MRI model fine-tuned on CT improved the agreement to 77% and 73%; The CT model trained from scratch had agreements of 73% and 71%; All of the deep learning models outperformed the rCBF segmentation from RAPID, which had agreements of 51% and 64%. See Table and figure. Conclusions: It is feasible to apply MRP-based deep learning model to CT. Fine-tuning with CTP data further improves the predictions. All deep learning models predict the stroke lesion after major recanalization better than thresholding approaches based on rCBF.

Sign in / Sign up

Export Citation Format

Share Document