Deep Learning Applications in Medical Imaging

2021 ◽  
pp. 178-208
Author(s):  
S. Sasikala ◽  
S. J. Subhashini ◽  
P. Alli ◽  
J. Jane Rubel Angelina
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Deep Learning ◽  
Medical Imaging ◽  
Learning Model ◽  
Learning Models ◽  
Driving System ◽  
Machine Learning Model ◽  
Car Driving ◽  
Machine Learning Models

Machine learning is a technique of parsing data, learning from that data, and then applying what has been learned to make informed decisions. Deep learning is actually a subset of machine learning. It technically is machine learning and functions in the same way, but it has different capabilities. The main difference between deep and machine learning is, machine learning models become well progressively, but the model still needs some guidance. If a machine learning model returns an inaccurate prediction, then the programmer needs to fix that problem explicitly, but in the case of deep learning, the model does it by itself. Automatic car driving system is a good example of deep learning. On other hand, Artificial Intelligence is a different thing from machine learning and deep learning. Deep learning and machine learning both are the subsets of AI.

scholarly journals Can artificial intelligence predict the need for oxygen therapy in early stage COVID-19 pneumonia?

2020 ◽  
Author(s):  
Hirofumi Obinata ◽  
Peiying Ruan ◽  
Hitoshi Mori ◽  
Wentao Zhu ◽  
Hisashi Sasaki ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Deep Learning ◽  
Clinical Features ◽  
Oxygen Therapy ◽  
Learning Model ◽  
Ct Images ◽  
Learning Models ◽  
Deep Learning Model ◽  
Machine Learning Models

Abstract This study investigated the utility of artificial intelligence in predicting disease progression. We analysed 194 patients with COVID-19 confirmed by reverse transcription polymerase chain reaction. Among them, 31 patients had oxygen therapy administered after admission. To assess the utility of artificial intelligence in the prediction of disease progression, we used three machine learning models employing clinical features (patient’s background, laboratory data, and symptoms), one deep learning model employing computed tomography (CT) images, and one multimodal deep learning model employing a combination of clinical features and CT images. We also evaluated the predictive values of these models and analysed the important features required to predict worsening in cases of COVID-19. The multimodal deep learning model had the highest accuracy. The CT image was an important feature of multimodal deep learning model. The area under the curve of all machine learning models employing clinical features and the deep learning model employing CT images exceeded 90%, and sensitivity of these models exceeded 95%. C-reactive protein and lactate dehydrogenase were important features of machine learning models. Our machine learning model, while slightly less accurate than the multimodal model, still provides a valuable medical triage tool for patients in the early stages of COVID-19.

scholarly journals A Physics-Infused Deep Learning Model for the Prediction of Refractive Indices and Its Use for the Large-Scale Screening of Organic Compound Space

2019 ◽  
Author(s):  
Mojtaba Haghighatlari ◽  
Gaurav Vishwakarma ◽  
Mohammad Atif Faiz Afzal ◽  
Johannes Hachmann
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Large Scale ◽  
Organic Molecules ◽  
Learning Model ◽  
Training Data ◽  
Refractive Indices ◽  
Learning Models ◽  
Deep Learning Model ◽  
Machine Learning Models

<div><div><div><p>We present a multitask, physics-infused deep learning model to accurately and efficiently predict refractive indices (RIs) of organic molecules, and we apply it to a library of 1.5 million compounds. We show that it outperforms earlier machine learning models by a significant margin, and that incorporating known physics into data-derived models provides valuable guardrails. Using a transfer learning approach, we augment the model to reproduce results consistent with higher-level computational chemistry training data, but with a considerably reduced number of corresponding calculations. Prediction errors of machine learning models are typically smallest for commonly observed target property values, consistent with the distribution of the training data. However, since our goal is to identify candidates with unusually large RI values, we propose a strategy to boost the performance of our model in the remoter areas of the RI distribution: We bias the model with respect to the under-represented classes of molecules that have values in the high-RI regime. By adopting a metric popular in web search engines, we evaluate our effectiveness in ranking top candidates. We confirm that the models developed in this study can reliably predict the RIs of the top 1,000 compounds, and are thus able to capture their ranking. We believe that this is the first study to develop a data-derived model that ensures the reliability of RI predictions by model augmentation in the extrapolation region on such a large scale. These results underscore the tremendous potential of machine learning in facilitating molecular (hyper)screening approaches on a massive scale and in accelerating the discovery of new compounds and materials, such as organic molecules with high-RI for applications in opto-electronics.</p></div></div></div>

Train Deep Learning Models using subsurface geological images datasets

2021 ◽  
Author(s):  
Ramy Abdallah ◽  
Clare E. Bond ◽  
Robert W.H. Butler
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Learning Models ◽  
Structural Interpretation ◽  
Performance Accuracy ◽  
Learning Techniques ◽  
Machine Learning Models

&lt;p&gt;Machine learning is being presented as a new solution for a wide range of geoscience problems. Primarily machine learning has been used for 3D seismic data processing, seismic facies analysis and well log data correlation. The rapid development in technology with open-source artificial intelligence libraries and the accessibility of affordable computer graphics processing units (GPU) makes the application of machine learning in geosciences increasingly tractable. However, the application of artificial intelligence in structural interpretation workflows of subsurface datasets is still ambiguous. This study aims to use machine learning techniques to classify images of folds and fold-thrust structures. Here we show that convolutional neural networks (CNNs) as supervised deep learning techniques provide excellent algorithms to discriminate between geological image datasets. Four different datasets of images have been used to train and test the machine learning models. These four datasets are a seismic character dataset with five classes (faults, folds, salt, flat layers and basement), folds types with three classes (buckle, chevron and conjugate), fault types with three classes (normal, reverse and thrust) and fold-thrust geometries with three classes (fault bend fold, fault propagation fold and detachment fold). These image datasets are used to investigate three machine learning models. One Feedforward linear neural network model and two convolutional neural networks models (Convolution 2d layer transforms sequential model and Residual block model (ResNet with 9, 34, and 50 layers)). Validation and testing datasets forms a critical part of testing the model&amp;#8217;s performance accuracy. The ResNet model records the highest performance accuracy score, of the machine learning models tested. Our CNN image classification model analysis provides a framework for applying machine learning to increase structural interpretation efficiency, and shows that CNN classification models can be applied effectively to geoscience problems. The study provides a starting point to apply unsupervised machine learning approaches to sub-surface structural interpretation workflows.&lt;/p&gt;

Improving Logging Prediction on Imbalanced Datasets

2016 ◽  
Vol 7 (2) ◽  
pp. 43-71 ◽  
Author(s):  
Sangeeta Lal ◽  
Neetu Sardana ◽  
Ashish Sureka
Keyword(s):  
Machine Learning ◽  
Open Source ◽  
Class Imbalance ◽  
Learning Model ◽  
Learning Models ◽  
Class Imbalance Problem ◽  
Imbalanced Datasets ◽  
Imbalance Problem ◽  
Machine Learning Model ◽  
Machine Learning Models

Logging is an important yet tough decision for OSS developers. Machine-learning models are useful in improving several steps of OSS development, including logging. Several recent studies propose machine-learning models to predict logged code construct. The prediction performances of these models are limited due to the class-imbalance problem since the number of logged code constructs is small as compared to non-logged code constructs. No previous study analyzes the class-imbalance problem for logged code construct prediction. The authors first analyze the performances of J48, RF, and SVM classifiers for catch-blocks and if-blocks logged code constructs prediction on imbalanced datasets. Second, the authors propose LogIm, an ensemble and threshold-based machine-learning model. Third, the authors evaluate the performance of LogIm on three open-source projects. On average, LogIm model improves the performance of baseline classifiers, J48, RF, and SVM, by 7.38%, 9.24%, and 4.6% for catch-blocks, and 12.11%, 14.95%, and 19.13% for if-blocks logging prediction.

scholarly journals An Introduction on Interpretable Machine Learning

2020 ◽  
Vol 9 (7S) ◽  
pp. 107-111
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Human Life ◽  
Accuracy Score ◽  
Learning Models ◽  
Ethical Practices ◽  
Interpretable Machine Learning ◽  
Machine Learning Model ◽  
Machine Learning Applications ◽  
Machine Learning Models

As Artificial Intelligence penetrates all aspects of human life, more and more questions about ethical practices and fair uses arise, which has motivated the research community to look inside and develop methods to interpret these Artificial Intelligence/Machine Learning models. This concept of interpretability can not only help with the ethical questions but also can provide various insights into the working of these machine learning models, which will become crucial in trust-building and understanding how a model makes decisions. Furthermore, in many machine learning applications, the feature of interpretability is the primary value that they offer. However, in practice, many developers select models based on the accuracy score and disregarding the level of interpretability of that model, which can be chaotic as predictions by many high accuracy models are not easily explainable. In this paper, we introduce the concept of Machine Learning Model Interpretability, Interpretable Machine learning, and the methods used for interpretation and explanations.

scholarly journals Technical note: how to rationally compare the performances of different machine learning models?

2018 ◽  
Author(s):  
Terazima Maeda
Keyword(s):  
Machine Learning ◽  
Predictive Accuracy ◽  
Learning Model ◽  
Technical Note ◽  
Learning Models ◽  
Specific Data ◽  
Machine Learning Model ◽  
Specific Prediction ◽  
Data Group ◽  
Machine Learning Models

Nowadays, there is a large number of machine learning models that could be used for various areas. However, different research targets are usually sensitive to the type of models. For a specific prediction target, the predictive accuracy of a machine learning model is always dependent to the data feature, data size and the intrinsic relationship between inputs and outputs. Therefore, for a specific data group and a fixed prediction mission, how to rationally compare the predictive accuracy of different machine learning model is a big question. In this brief note, we show how should we compare the performances of different machine models by raising some typical examples.

scholarly journals Development and Validation of a Quick Sepsis-Related Organ Failure Assessment-Based Machine-Learning Model for Mortality Prediction in Patients with Suspected Infection in the Emergency Department

2020 ◽  
Vol 9 (3) ◽  
pp. 875
Author(s):  
Young Suk Kwon ◽  
Moon Seong Baek
Keyword(s):  
Machine Learning ◽  
Emergency Department ◽  
Learning Model ◽  
Gradient Boosting ◽  
Learning Models ◽  
Suspected Infection ◽  
Machine Learning Model ◽  
Failure Assessment ◽  
Qsofa Score ◽  
Machine Learning Models

The quick sepsis-related organ failure assessment (qSOFA) score has been introduced to predict the likelihood of organ dysfunction in patients with suspected infection. We hypothesized that machine-learning models using qSOFA variables for predicting three-day mortality would provide better accuracy than the qSOFA score in the emergency department (ED). Between January 2016 and December 2018, the medical records of patients aged over 18 years with suspected infection were retrospectively obtained from four EDs in Korea. Data from three hospitals (n = 19,353) were used as training-validation datasets and data from one (n = 4234) as the test dataset. Machine-learning algorithms including extreme gradient boosting, light gradient boosting machine, and random forest were used. We assessed the prediction ability of machine-learning models using the area under the receiver operating characteristic (AUROC) curve, and DeLong’s test was used to compare AUROCs between the qSOFA scores and qSOFA-based machine-learning models. A total of 447,926 patients visited EDs during the study period. We analyzed 23,587 patients with suspected infection who were admitted to the EDs. The median age of the patients was 63 years (interquartile range: 43–78 years) and in-hospital mortality was 4.0% (n = 941). For predicting three-day mortality among patients with suspected infection in the ED, the AUROC of the qSOFA-based machine-learning model (0.86 [95% CI 0.85–0.87]) for three -day mortality was higher than that of the qSOFA scores (0.78 [95% CI 0.77–0.79], p < 0.001). For predicting three-day mortality in patients with suspected infection in the ED, the qSOFA-based machine-learning model was found to be superior to the conventional qSOFA scores.

Machine learning models for Hg prospecting in the Almadén mining district

2021 ◽  
Author(s):  
Julio Alberto López-Gómez ◽  
Daniel Carrasco Pardo ◽  
Pablo Higueras ◽  
Jose María Esbrí ◽  
Saturnino Lorenzo
Keyword(s):  
Machine Learning ◽  
Learning Model ◽  
Mining District ◽  
Learning Models ◽  
Geological Features ◽  
Machine Learning Model ◽  
Mineral Prospectivity ◽  
Data Point ◽  
The One ◽  
Machine Learning Models

&lt;p&gt;Traditionally, prospectivity models were designed using approaches mainly based on expert judgement. These models have been widely applied and they are also known as knowledge-driven prospectivity models (see Harris et al. (2015)). Currently, artificial intelligence approaches, especially machine learning models, are being applied to build prospectivity models since they have been proven to be successful in many other domains (see Sun et al., 2019 and Guerra Prado et al., 2020). They are also known as data-driven prospectivity models. Machine learning models allow to learn from data repositories in order to extract and detect relationships from the data to predict new instances.&lt;/p&gt;&lt;p&gt;In this work, a geological dataset was collected by a team of expert geologists. The data collected includes the geographical coordinates as well as several geological features of points belonged to seventy-seven different mercury deposits in the Almad&amp;#233;n mining district. The resulting dataset is composed by a total of 24798 points and 24 attributes for each point. In particular, we have collected geological and mining-related data regarding the Almad&amp;#233;n mercury (Hg) mining district; these data include the location of the several Hg mineralizations, including their typology, size, mineralogy, and stratigraphic position, as well as other information associated to the metallogenetic model set up by Hern&amp;#225;ndez et al. (1999).&lt;/p&gt;&lt;p&gt;Later, few machine learning models are built to select the one which offers the best results. The aim of this work is twofold: on the one hand, it is intended to build a machine learning model capable of, given the geological features of a data point, to determine the mercury deposit to which it belongs. On the other hand, the aim is to build a machine learning model capable of, given the geological features of a data point, to identify the kind of deposit to which it belongs. The experiments conducted in this work have been properly designed, validating the results obtained using statistical techniques.&lt;/p&gt;&lt;p&gt;Finally, the models built in this work will allow to generate mercury prospectivity maps. The final aim of this process is to get and train a system able to perform antimony prospection in the nearby Guadalmez syncline.&lt;/p&gt;&lt;p&gt;This work was funded by the ANR (ANR-19-MIN2-0002-01), the AEI (MICIU/AEI/REF.: PCI2019-103779) and author&amp;#8217;s institutions in the framework of the ERA-MIN2 AUREOLE project.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;References&lt;/strong&gt;&lt;/p&gt;&lt;p&gt;Guerra Prado E.M.; de Souza Filho C.R.; Carranza E.M.; Motta J.G. (2020). Modeling of Cu-Au prospectivity in the Caraj&amp;#225;s mineral province (Brasil) through machine learning: Dealing with embalanced training data.&lt;/p&gt;&lt;p&gt;Harris, J.R.; Grunsky, E.; Corrigan, D. (2015). Data- and knowledge-driven mineral prospectivity maps for Canda&amp;#8217;s North.&lt;/p&gt;&lt;p&gt;Hern&amp;#225;ndez, A.; J&amp;#233;brak, M.; Higueras, P.; Oyarzun, R.; Morata, D.; Munh&amp;#225;, J. (1999). The Almad&amp;#233;n mercury mining district, Spain. Mineralium Deposita, 34: 539-548.&lt;/p&gt;&lt;p&gt;Sun, T.; Chen, F.; Zhong, L.; Liu, W.; Wang, Y. (2019). GIS-based mineral prospectivity mapping using machine learning methods: A case study from Tongling ore district, eastern China.&lt;/p&gt;

Hybrid Modeling In Unconventional Reservoirs To Forecast Estimated Ultimate Recovery

2021 ◽  
Author(s):  
Cenk Temizel ◽  
Celal Hakan Canbaz ◽  
Karthik Balaji ◽  
Ahsen Ozesen ◽  
Kirill Yanidis ◽  
...  
Keyword(s):  
Machine Learning ◽  
Integrated Approach ◽  
Learning Model ◽  
Hybrid Modeling ◽  
Unconventional Reservoirs ◽  
Learning Models ◽  
Production Forecasting ◽  
Machine Learning Model ◽  
Machine Learning Models ◽  
Estimated Ultimate Recovery

Abstract Machine learning models have worked as a robust tool in forecasting and optimization processes for wells in conventional, data-rich reservoirs. In unconventional reservoirs however, given the large ranges of uncertainty, purely data-driven, machine learning models have not yet proven to be repeatable and scalable. In such cases, integrating physics-based reservoir simulation methods along with machine learning techniques can be used as a solution to alleviate these limitations. The objective of this study is to provide an overview along with examples of implementing this integrated approach for the purpose of forecasting Estimated Ultimate Recovery (EUR) in shale reservoirs. This study is solely based on synthetic data. To generate data for one section of a reservoir, a full-physics reservoir simulator has been used. Simulated data from this section is used to train a machine learning model, which provides EUR as the output. Production from another section of the field with a different range of reservoir properties is then forecasted using a physics-based model. Using the earlier trained model, production forecasting for this section of the reservoir is then carried out to illustrate the integrated approach to EUR forecasting for a section of the reservoir that is not data rich. The integrated approach, or hybrid modeling, production forecasting for different sections of the reservoir that were data-starved, are illustrated. Using the physics-based model, the uncertainty in EUR predictions made by the machine learning model has been reduced and a more accurate forecasting has been attained. This method is primarily applicable in reservoirs, such as unconventionals, where one section of the field that has been developed has a substantial amount of data, whereas, the other section of the field will be data starved. The hybrid model was consistently able to forecast EUR at an acceptable level of accuracy, thereby, highlighting the benefits of this type of an integrated approach. This study advances the application of repeatable and scalable hybrid models in unconventional reservoirs and highlights its benefits as compared to using either physics-based or machine-learning based models separately.

Sign in / Sign up

Export Citation Format

Share Document