Evaluation of a task specific self-supervised learning framework in digital pathology relative to transfer learning approaches and existing foundation models.

An integral stage in typical digital pathology workflows involves deriving specific features from tiles extracted from a tessellated whole slide image. Notably, various computer vision neural network architectures, particularly the ImageNet pre-trained, have been extensively used in this domain. This study critically analyzes multiple strategies for encoding tiles to understand the extent of transfer learning and identify the most effective approach. The study categorizes neural network performance into three weight initialization methods: random, ImageNet-based, and self-supervised learning. Additionally, we propose a framework based on task-specific self-supervised learning (TS-SSL) which introduces a shallow feature extraction method, employing a spatial-channel attention block to glean distinctive features optimized for histopathology intricacies. Across two different downstream classification tasks (patch classification, and weakly supervised whole slide image classification) with diverse classification datasets, including Colorectal cancer histology, Patch Camelyon, PANDA, TCGA and CIFAR-10, our task specific self-supervised encoding approach consistently outperforms other CNN-based encoders. The better performances highlight the potential of task-specific-attention based self-supervised training in tailoring feature extraction for histopathology, indicating a shift from utilizing pretrained models originating outside the histopathology domain. Our study supports the idea that task-specific self-supervised learning allows domain-specific feature extraction, encouraging a more focused analysis.

Machine learning and discriminant function analysis in the formulation of generic models for sex prediction using patella measurements.

Sex prediction from bone measurements that display sexual dimorphism is one of the most important aspects of forensic anthropology. Some bones like the skull and pelvis display distinct morphological traits that are based on shape. These morphological traits which are sexually dimorphic across different population groups have been shown to provide an acceptably high degree of accuracy in the prediction of sex. A sample of 100 patella of Mixed Ancestry South Africans (MASA) was collected from the Dart collection. Six parameters: maximum height (maxh), maximum breadth (maxw), maximum thickness (maxt), the height of articular facet (haf), lateral articular facet breadth (lafb), and medial articular facet breath (mafb) were used in this study. Stepwise and direct discriminant function analyses were performed for measurements that exhibited significant differences between male and female mean measurements, and the "leave-one-out" approach was used for validation. Moreover, we have used eight classical machine learning techniques along with feature ranking techniques to identify the best feature combinations for sex prediction. A stacking machine learning technique was trained and validated to classify the sex of the subject. Here, we have used the top performing three ML classifiers as base learners and the predictions of these models were used as inputs to different machine learning classifiers as meta learners to make the final decision. The measurements of the patella of South Africans are sexually dimorphic and this observation is consistent with previous studies on the patella of different countries. The range of average accuracies obtained for pooled multivariate discriminant function equations is 81.9-84.2%, while the stacking ML technique provides 90.8% accuracy which compares well with those presented for previous studies in other parts of the world. In conclusion, the models proposed in this study from measurements of the patella of different population groups in South Africa are useful resent with reasonably high average accuracies.

A Shallow U-Net Architecture for Reliably Predicting Blood Pressure (BP) from Photoplethysmogram (PPG) and Electrocardiogram (ECG) Signals.

Cardiovascular diseases are the most common causes of death around the world. To detect and treat heart-related diseases, continuous blood pressure (BP) monitoring along with many other parameters are required. Several invasive and non-invasive methods have been developed for this purpose. Most existing methods used in hospitals for continuous monitoring of BP are invasive. On the contrary, cuff-based BP monitoring methods, which can predict systolic blood pressure (SBP) and diastolic blood pressure (DBP), cannot be used for continuous monitoring. Several studies attempted to predict BP from non-invasively collectible signals such as photoplethysmograms (PPG) and electrocardiograms (ECG), which can be used for continuous monitoring. In this study, we explored the applicability of autoencoders in predicting BP from PPG and ECG signals. The investigation was carried out on 12,000 instances of 942 patients of the MIMIC-II dataset, and it was found that a very shallow, one-dimensional autoencoder can extract the relevant features to predict the SBP and DBP with state-of-the-art performance on a very large dataset. An independent test set from a portion of the MIMIC-II dataset provided a mean absolute error (MAE) of 2.333 and 0.713 for SBP and DBP, respectively. On an external dataset of 40 subjects, the model trained on the MIMIC-II dataset provided an MAE of 2.728 and 1.166 for SBP and DBP, respectively. For both the cases, the results met British Hypertension Society (BHS) Grade A and surpassed the studies from the current literature.

A Novel Machine Learning Approach for Severity Classification of Diabetic Foot Complications Using Thermogram Images.

Diabetes mellitus (DM) is one of the most prevalent diseases in the world, and is correlated to a high index of mortality. One of its major complications is diabetic foot, leading to plantar ulcers, amputation, and death. Several studies report that a thermogram helps to detect changes in the plantar temperature of the foot, which may lead to a higher risk of ulceration. However, in diabetic patients, the distribution of plantar temperature does not follow a standard pattern, thereby making it difficult to quantify the changes. The abnormal temperature distribution in infrared (IR) foot thermogram images can be used for the early detection of diabetic foot before ulceration to avoid complications. There is no machine learning-based technique reported in the literature to classify these thermograms based on the severity of diabetic foot complications. This paper uses an available labeled diabetic thermogram dataset and uses the k-mean clustering technique to cluster the severity risk of diabetic foot ulcers using an unsupervised approach. Using the plantar foot temperature, the new clustered dataset is verified by expert medical doctors in terms of risk for the development of foot ulcers. The newly labeled dataset is then investigated in terms of robustness to be classified by any machine learning network. Classical machine learning algorithms with feature engineering and a convolutional neural network (CNN) with image-enhancement techniques are investigated to provide the best-performing network in classifying thermograms based on severity. It is found that the popular VGG 19 CNN model shows an accuracy, precision, sensitivity, F1-score, and specificity of 95.08%, 95.08%, 95.09%, 95.08%, and 97.2%, respectively, in the stratification of severity. A stacking classifier is proposed using extracted features of the thermogram, which is created using the trained gradient boost classifier, XGBoost classifier, and random forest classifier. This provides a comparable performance of 94.47%, 94.45%, 94.47%, 94.43%, and 93.25% for accuracy, precision, sensitivity, F1-score, and specificity, respectively.

Thermal Change Index-Based Diabetic Foot Thermogram Image Classification Using Machine Learning Techniques.

Diabetes mellitus (DM) can lead to plantar ulcers, amputation and death. Plantar foot thermogram images acquired using an infrared camera have been shown to detect changes in temperature distribution associated with a higher risk of foot ulceration. Machine learning approaches applied to such infrared images may have utility in the early diagnosis of diabetic foot complications. In this work, a publicly available dataset was categorized into different classes, which were corroborated by domain experts, based on a temperature distribution parameter-the thermal change index (TCI). We then explored different machine-learning approaches for classifying thermograms of the TCI-labeled dataset. Classical machine learning algorithms with feature engineering and the convolutional neural network (CNN) with image enhancement techniques were extensively investigated to identify the best performing network for classifying thermograms. The multilayer perceptron (MLP) classifier along with the features extracted from thermogram images showed an accuracy of 90.1% in multi-class classification, which outperformed the literature-reported performance metrics on this dataset.

An Intelligent and Low-Cost Eye-Tracking System for Motorized Wheelchair Control.

In the 34 developed and 156 developing countries, there are ~132 million disabled people who need a wheelchair, constituting 1.86% of the world population. Moreover, there are millions of people suffering from diseases related to motor disabilities, which cause inability to produce controlled movement in any of the limbs or even head. This paper proposes a system to aid people with motor disabilities by restoring their ability to move effectively and effortlessly without having to rely on others utilizing an eye-controlled electric wheelchair. The system input is images of the user's eye that are processed to estimate the gaze direction and the wheelchair was moved accordingly. To accomplish such a feat, four user-specific methods were developed, implemented, and tested; all of which were based on a benchmark database created by the authors. The first three techniques were automatic, employ correlation, and were variants of template matching, whereas the last one uses convolutional neural networks (CNNs). Different metrics to quantitatively evaluate the performance of each algorithm in terms of accuracy and latency were computed and overall comparison is presented. CNN exhibited the best performance (i.e., 99.3% classification accuracy), and thus it was the model of choice for the gaze estimator, which commands the wheelchair motion. The system was evaluated carefully on eight subjects achieving 99% accuracy in changing illumination conditions outdoor and indoor. This required modifying a motorized wheelchair to adapt it to the predictions output by the gaze estimation algorithm. The wheelchair control can bypass any decision made by the gaze estimator and immediately halt its motion with the help of an array of proximity sensors, if the measured distance goes below a well-defined safety margin. This work not only empowers any immobile wheelchair user, but also provides low-cost tools for the organization assisting wheelchair users.

Evaluation and mitigation of cognitive biases in medical language models.

Increasing interest in applying large language models (LLMs) to medicine is due in part to their impressive performance on medical exam questions. However, these exams do not capture the complexity of real patient-doctor interactions because of factors like patient compliance, experience, and cognitive bias. We hypothesized that LLMs would produce less accurate responses when faced with clinically biased questions as compared to unbiased ones. To test this, we developed the BiasMedQA dataset, which consists of 1273 USMLE questions modified to replicate common clinically relevant cognitive biases. We assessed six LLMs on BiasMedQA and found that GPT-4 stood out for its resilience to bias, in contrast to Llama 2 70B-chat and PMC Llama 13B, which showed large drops in performance. Additionally, we introduced three bias mitigation strategies, which improved but did not fully restore accuracy. Our findings highlight the need to improve LLMs' robustness to cognitive biases, in order to achieve more reliable applications of LLMs in healthcare.

Brain Tumor Segmentation and Classification from Sensor-Based Portable Microwave Brain Imaging System Using Lightweight Deep Learning Models.

Automated brain tumor segmentation from reconstructed microwave (RMW) brain images and image classification is essential for the investigation and monitoring of the progression of brain disease. The manual detection, classification, and segmentation of tumors are extremely time-consuming but crucial tasks due to the tumor's pattern. In this paper, we propose a new lightweight segmentation model called MicrowaveSegNet (MSegNet), which segments the brain tumor, and a new classifier called the BrainImageNet (BINet) model to classify the RMW images. Initially, three hundred (300) RMW brain image samples were obtained from our sensors-based microwave brain imaging (SMBI) system to create an original dataset. Then, image preprocessing and augmentation techniques were applied to make 6000 training images per fold for a 5-fold cross-validation. Later, the MSegNet and BINet were compared to state-of-the-art segmentation and classification models to verify their performance. The MSegNet has achieved an Intersection-over-Union (IoU) and Dice score of 86.92% and 93.10%, respectively, for tumor segmentation. The BINet has achieved an accuracy, precision, recall, F1-score, and specificity of 89.33%, 88.74%, 88.67%, 88.61%, and 94.33%, respectively, for three-class classification using raw RMW images, whereas it achieved 98.33%, 98.35%, 98.33%, 98.33%, and 99.17%, respectively, for segmented RMW images. Therefore, the proposed cascaded model can be used in the SMBI system.

A Lightweight Deep Learning Based Microwave Brain Image Network Model for Brain Tumor Classification Using Reconstructed Microwave Brain (RMB) Images.

Computerized brain tumor classification from the reconstructed microwave brain (RMB) images is important for the examination and observation of the development of brain disease. In this paper, an eight-layered lightweight classifier model called microwave brain image network (MBINet) using a self-organized operational neural network (Self-ONN) is proposed to classify the reconstructed microwave brain (RMB) images into six classes. Initially, an experimental antenna sensor-based microwave brain imaging (SMBI) system was implemented, and RMB images were collected to create an image dataset. It consists of a total of 1320 images: 300 images for the non-tumor, 215 images for each single malignant and benign tumor, 200 images for each double benign tumor and double malignant tumor, and 190 images for the single benign and single malignant tumor classes. Then, image resizing and normalization techniques were used for image preprocessing. Thereafter, augmentation techniques were applied to the dataset to make 13,200 training images per fold for 5-fold cross-validation. The MBINet model was trained and achieved accuracy, precision, recall, F1-score, and specificity of 96.97%, 96.93%, 96.85%, 96.83%, and 97.95%, respectively, for six-class classification using original RMB images. The MBINet model was compared with four Self-ONNs, two vanilla CNNs, ResNet50, ResNet101, and DenseNet201 pre-trained models, and showed better classification outcomes (almost 98%). Therefore, the MBINet model can be used for reliably classifying the tumor(s) using RMB images in the SMBI system.

BIO-CXRNET: a robust multimodal stacking machine learning technique for mortality risk prediction of COVID-19 patients using chest X-ray images and clinical data.

Nowadays, quick, and accurate diagnosis of COVID-19 is a pressing need. This study presents a multimodal system to meet this need. The presented system employs a machine learning module that learns the required knowledge from the datasets collected from 930 COVID-19 patients hospitalized in Italy during the first wave of COVID-19 (March-June 2020). The dataset consists of twenty-five biomarkers from electronic health record and Chest X-ray (CXR) images. It is found that the system can diagnose low- or high-risk patients with an accuracy, sensitivity, and F1-score of 89.03%, 90.44%, and 89.03%, respectively. The system exhibits 6% higher accuracy than the systems that employ either CXR images or biomarker data. In addition, the system can calculate the mortality risk of high-risk patients using multivariate logistic regression-based nomogram scoring technique. Interested physicians can use the presented system to predict the early mortality risks of COVID-19 patients using the web-link: Covid-severity-grading-AI. In this case, a physician needs to input the following information: CXR image file, Lactate Dehydrogenase (LDH), Oxygen Saturation (O2%), White Blood Cells Count, C-reactive protein, and Age. This way, this study contributes to the management of COVID-19 patients by predicting early mortality risk. Supplementary Information: The online version contains supplementary material available at 10.1007/s00521-023-08606-w.

COV-ECGNET: COVID-19 detection using ECG trace images with deep convolutional neural network.

The reliable and rapid identification of the COVID-19 has become crucial to prevent the rapid spread of the disease, ease lockdown restrictions and reduce pressure on public health infrastructures. Recently, several methods and techniques have been proposed to detect the SARS-CoV-2 virus using different images and data. However, this is the first study that will explore the possibility of using deep convolutional neural network (CNN) models to detect COVID-19 from electrocardiogram (ECG) trace images. In this work, COVID-19 and other cardiovascular diseases (CVDs) were detected using deep-learning techniques. A public dataset of ECG images consisting of 1937 images from five distinct categories, such as normal, COVID-19, myocardial infarction (MI), abnormal heartbeat (AHB), and recovered myocardial infarction (RMI) were used in this study. Six different deep CNN models (ResNet18, ResNet50, ResNet101, InceptionV3, DenseNet201, and MobileNetv2) were used to investigate three different classification schemes: (i) two-class classification (normal vs COVID-19); (ii) three-class classification (normal, COVID-19, and other CVDs), and finally, (iii) five-class classification (normal, COVID-19, MI, AHB, and RMI). For two-class and three-class classification, Densenet201 outperforms other networks with an accuracy of 99.1%, and 97.36%, respectively; while for the five-class classification, InceptionV3 outperforms others with an accuracy of 97.83%. ScoreCAM visualization confirms that the networks are learning from the relevant area of the trace images. Since the proposed method uses ECG trace images which can be captured by smartphones and are readily available facilities in low-resources countries, this study will help in faster computer-aided diagnosis of COVID-19 and other cardiac abnormalities.

Estimating the Dissolution of Anticancer Drugs in Supercritical Carbon Dioxide with a Stacked Machine Learning Model.

Synthesizing micro-/nano-sized pharmaceutical compounds with an appropriate size distribution is a method often followed to enhance drug delivery and reduce side effects. Supercritical CO2 (carbon dioxide) is a well-known solvent utilized in the pharmaceutical synthesis process. Reliable knowledge of a drug's solubility in supercritical CO2 is necessary for feasible study, modeling, design, optimization, and control of such a process. Therefore, the current study constructs a stacked/ensemble model by combining three up-to-date machine learning tools (i.e., extra tree, gradient boosting, and random forest) to predict the solubility of twelve anticancer drugs in supercritical CO2. An experimental databank comprising 311 phase equilibrium samples was gathered from the literature and applied to design the proposed stacked model. This model estimates the solubility of anticancer drugs in supercritical CO2 as a function of solute and solvent properties and operating conditions. Several statistical indices, including average absolute relative deviation (AARD = 8.62%), mean absolute error (MAE = 2.86 × 10-6), relative absolute error (RAE = 2.42%), mean squared error (MSE = 1.26 × 10-10), and regression coefficient (R2 = 0.99809) were used to validate the performance of the constructed model. The statistical, sensitivity, and trend analyses confirmed that the suggested stacked model demonstrates excellent performance for correlating and predicting the solubility of anticancer drugs in supercritical CO2.

Deep Learning Assisted Automated Assessment of Thalassaemia from Haemoglobin Electrophoresis Images.

Haemoglobin (Hb) electrophoresis is a method of blood testing used to detect thalassaemia. However, the interpretation of the result of the electrophoresis test itself is a complex task. Expert haematologists, specifically in developing countries, are relatively few in number and are usually overburdened. To assist them with their workload, in this paper we present a novel method for the automated assessment of thalassaemia using Hb electrophoresis images. Moreover, in this study we compile a large Hb electrophoresis image dataset, consisting of 103 strips containing 524 electrophoresis images with a clear consensus on the quality of electrophoresis obtained from 824 subjects. The proposed methodology is split into two parts: (1) single-patient electrophoresis image segmentation by means of the lane extraction technique, and (2) binary classification (normal or abnormal) of the electrophoresis images using state-of-the-art deep convolutional neural networks (CNNs) and using the concept of transfer learning. Image processing techniques including filtering and morphological operations are applied for object detection and lane extraction to automatically separate the lanes and classify them using CNN models. Seven different CNN models (ResNet18, ResNet50, ResNet101, InceptionV3, DenseNet201, SqueezeNet and MobileNetV2) were investigated in this study. InceptionV3 outperformed the other CNNs in detecting thalassaemia using Hb electrophoresis images. The accuracy, precision, recall, f1-score, and specificity in the detection of thalassaemia obtained with the InceptionV3 model were 95.8%, 95.84%, 95.8%, 95.8% and 95.8%, respectively. MobileNetV2 demonstrated an accuracy, precision, recall, f1-score, and specificity of 95.72%, 95.73%, 95.72%, 95.7% and 95.72% respectively. Its performance was comparable with the best performing model, InceptionV3. Since it is a very shallow network, MobileNetV2 also provides the least latency in processing a single-patient image and it can be suitably used for mobile applications. The proposed approach, which has shown very high classification accuracy, will assist in the rapid and robust detection of thalassaemia using Hb electrophoresis images.

Blind ECG Restoration by Operational Cycle-GANs.

OBJECTIVE: ECG recordings often suffer from a set of artifacts with varying types, severities, and durations, and this makes an accurate diagnosis by machines or medical doctors difficult and unreliable. Numerous studies have proposed ECG denoising; however, they naturally fail to restore the actual ECG signal corrupted with such artifacts due to their simple and naive noise model. In this pilot study, we propose a novel approach for blind ECG restoration using cycle-consistent generative adversarial networks (Cycle-GANs) where the quality of the signal can be improved to a clinical level ECG regardless of the type and severity of the artifacts corrupting the signal. METHODS: To further boost the restoration performance, we propose 1D operational Cycle-GANs with the generative neuron model. RESULTS: The proposed approach has been evaluated extensively using one of the largest benchmark ECG datasets from the China Physiological Signal Challenge (CPSC-2020) with more than one million beats. Besides the quantitative and qualitative evaluations, a group of cardiologists performed medical evaluations to validate the quality and usability of the restored ECG, especially for an accurate arrhythmia diagnosis. SIGNIFICANCE: As a pioneer study in ECG restoration, the corrupted ECG signals can be restored to clinical level quality. CONCLUSION: By means of the proposed ECG restoration, the ECG diagnosis accuracy and performance can significantly improve.

Estimating the Relative Crystallinity of Biodegradable Polylactic Acid and Polyglycolide Polymer Composites by Machine Learning Methodologies.

Biodegradable polymers have recently found significant applications in pharmaceutics processing and drug release/delivery. Composites based on poly (L-lactic acid) (PLLA) have been suggested to enhance the crystallization rate and relative crystallinity of pure PLLA polymers. Despite the large amount of experimental research that has taken place to date, the theoretical aspects of relative crystallinity have not been comprehensively investigated. Therefore, this research uses machine learning methods to estimate the relative crystallinity of biodegradable PLLA/PGA (polyglycolide) composites. Six different artificial intelligent classes were employed to estimate the relative crystallinity of PLLA/PGA polymer composites as a function of crystallization time, temperature, and PGA content. Cumulatively, 1510 machine learning topologies, including 200 multilayer perceptron neural networks, 200 cascade feedforward neural networks (CFFNN), 160 recurrent neural networks, 800 adaptive neuro-fuzzy inference systems, and 150 least-squares support vector regressions, were developed, and their prediction accuracy compared. The modeling results show that a single hidden layer CFFNN with 9 neurons is the most accurate method for estimating 431 experimentally measured datasets. This model predicts an experimental database with an average absolute percentage difference of 8.84%, root mean squared errors of 4.67%, and correlation coefficient (R2) of 0.999008. The modeling results and relevancy studies show that relative crystallinity increases based on the PGA content and crystallization time. Furthermore, the effect of temperature on relative crystallinity is too complex to be easily explained.

Deep Learning for Reliable Classification of COVID-19, MERS, and SARS from Chest X-ray Images.

Novel coronavirus disease (COVID-19) is an extremely contagious and quickly spreading coronavirus infestation. Severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS), which outbreak in 2002 and 2011, and the current COVID-19 pandemic are all from the same family of coronavirus. This work aims to classify COVID-19, SARS, and MERS chest X-ray (CXR) images using deep convolutional neural networks (CNNs). To the best of our knowledge, this classification scheme has never been investigated in the literature. A unique database was created, so-called QU-COVID-family, consisting of 423 COVID-19, 144 MERS, and 134 SARS CXR images. Besides, a robust COVID-19 recognition system was proposed to identify lung regions using a CNN segmentation model (U-Net), and then classify the segmented lung images as COVID-19, MERS, or SARS using a pre-trained CNN classifier. Furthermore, the Score-CAM visualization method was utilized to visualize classification output and understand the reasoning behind the decision of deep CNNs. Several deep learning classifiers were trained and tested; four outperforming algorithms were reported: SqueezeNet, ResNet18, InceptionV3, and DenseNet201. Original and preprocessed images were used individually and all together as the input(s) to the networks. Two recognition schemes were considered: plain CXR classification and segmented CXR classification. For plain CXRs, it was observed that InceptionV3 outperforms other networks with a 3-channel scheme and achieves sensitivities of 99.5%, 93.1%, and 97% for classifying COVID-19, MERS, and SARS images, respectively. In contrast, for segmented CXRs, InceptionV3 outperformed using the original CXR dataset and achieved sensitivities of 96.94%, 79.68%, and 90.26% for classifying COVID-19, MERS, and SARS images, respectively. The classification performance degrades with segmented CXRs compared to plain CXRs. However, the results are more reliable as the network learns from the main region of interest, avoiding irrelevant non-lung areas (heart, bones, or text), which was confirmed by the Score-CAM visualization. All networks showed high COVID-19 detection sensitivity (> 96%) with the segmented lung images. This indicates the unique radiographic signature of COVID-19 cases in the eyes of AI, which is often a challenging task for medical doctors.

Prognostic Model of ICU Admission Risk in Patients with COVID-19 Infection Using Machine Learning.

With the onset of the COVID-19 pandemic, the number of critically sick patients in intensive care units (ICUs) has increased worldwide, putting a burden on ICUs. Early prediction of ICU requirement is crucial for efficient resource management and distribution. Early-prediction scoring systems for critically ill patients using mathematical models are available, but are not generalized for COVID-19 and Non-COVID patients. This study aims to develop a generalized and reliable prognostic model for ICU admission for both COVID-19 and non-COVID-19 patients using best feature combination from the patient data at admission. A retrospective cohort study was conducted on a dataset collected from the pulmonology department of Moscow City State Hospital between 20 April 2020 and 5 June 2020. The dataset contains ten clinical features for 231 patients, of whom 100 patients were transferred to ICU and 131 were stable (non-ICU) patients. There were 156 COVID positive patients and 75 non-COVID patients. Different feature selection techniques were investigated, and a stacking machine learning model was proposed and compared with eight different classification algorithms to detect risk of need for ICU admission for both COVID-19 and non-COVID patients combined and COVID patients alone. C-reactive protein (CRP), chest computed tomography (CT), lung tissue affected (%), age, admission to hospital, and fibrinogen parameters at hospital admission were found to be important features for ICU-requirement risk prediction. The best performance was produced by the stacking approach, with weighted precision, sensitivity, F1-score, specificity, and overall accuracy of 84.45%, 84.48%, 83.64%, 84.47%, and 84.48%, respectively, for both types of patients, and 85.34%, 85.35%, 85.11%, 85.34%, and 85.35%, respectively, for COVID-19 patients only. The proposed work can help doctors to improve management through early prediction of the risk of need for ICU admission of patients during the COVID-19 pandemic, as the model can be used for both types of patients.

QCovSML: A reliable COVID-19 detection system using CBC biomarkers by a stacking machine learning model.

The reverse transcription-polymerase chain reaction (RT-PCR) test is considered the current gold standard for the detection of coronavirus disease (COVID-19), although it suffers from some shortcomings, namely comparatively longer turnaround time, higher false-negative rates around 20-25%, and higher cost equipment. Therefore, finding an efficient, robust, accurate, and widely available, and accessible alternative to RT-PCR for COVID-19 diagnosis is a matter of utmost importance. This study proposes a complete blood count (CBC) biomarkers-based COVID-19 detection system using a stacking machine learning (SML) model, which could be a fast and less expensive alternative. This study used seven different publicly available datasets, where the largest one consisting of fifteen CBC biomarkers collected from 1624 patients (52% COVID-19 positive) admitted at San Raphael Hospital, Italy from February to May 2020 was used to train and validate the proposed model. White blood cell count, monocytes (%), lymphocyte (%), and age parameters collected from the patients during hospital admission were found to be important biomarkers for COVID-19 disease prediction using five different feature selection techniques. Our stacking model produced the best performance with weighted precision, sensitivity, specificity, overall accuracy, and F1-score of 91.44%, 91.44%, 91.44%, 91.45%, and 91.45%, respectively. The stacking machine learning model improved the performance in comparison to other state-of-the-art machine learning classifiers. Finally, a nomogram-based scoring system (QCovSML) was constructed using this stacking approach to predict the COVID-19 patients. The cut-off value of the QCovSML system for classifying COVID-19 and Non-COVID patients was 4.8. Six datasets from three different countries were used to externally validate the proposed model to evaluate its generalizability and robustness. The nomogram demonstrated good calibration and discrimination with the area under the curve (AUC) of 0.961 for the internal cohort and average AUC of 0.967 for all external validation cohort, respectively. The external validation shows an average weighted precision, sensitivity, F1-score, specificity, and overall accuracy of 92.02%, 95.59%, 93.73%, 90.54%, and 93.34%, respectively.

QUCoughScope: An Intelligent Application to Detect COVID-19 Patients Using Cough and Breath Sounds.

Problem-Since the outbreak of the COVID-19 pandemic, mass testing has become essential to reduce the spread of the virus. Several recent studies suggest that a significant number of COVID-19 patients display no physical symptoms whatsoever. Therefore, it is unlikely that these patients will undergo COVID-19 testing, which increases their chances of unintentionally spreading the virus. Currently, the primary diagnostic tool to detect COVID-19 is a reverse-transcription polymerase chain reaction (RT-PCR) test from the respiratory specimens of the suspected patient, which is invasive and a resource-dependent technique. It is evident from recent researches that asymptomatic COVID-19 patients cough and breathe in a different way than healthy people. Aim-This paper aims to use a novel machine learning approach to detect COVID-19 (symptomatic and asymptomatic) patients from the convenience of their homes so that they do not overburden the healthcare system and also do not spread the virus unknowingly by continuously monitoring themselves. Method-A Cambridge University research group shared such a dataset of cough and breath sound samples from 582 healthy and 141 COVID-19 patients. Among the COVID-19 patients, 87 were asymptomatic while 54 were symptomatic (had a dry or wet cough). In addition to the available dataset, the proposed work deployed a real-time deep learning-based backend server with a web application to crowdsource cough and breath datasets and also screen for COVID-19 infection from the comfort of the user's home. The collected dataset includes data from 245 healthy individuals and 78 asymptomatic and 18 symptomatic COVID-19 patients. Users can simply use the application from any web browser without installation and enter their symptoms, record audio clips of their cough and breath sounds, and upload the data anonymously. Two different pipelines for screening were developed based on the symptoms reported by the users: asymptomatic and symptomatic. An innovative and novel stacking CNN model was developed using three base learners from of eight state-of-the-art deep learning CNN algorithms. The stacking CNN model is based on a logistic regression classifier meta-learner that uses the spectrograms generated from the breath and cough sounds of symptomatic and asymptomatic patients as input using the combined (Cambridge and collected) dataset. Results-The stacking model outperformed the other eight CNN networks with the best classification performance for binary classification using cough sound spectrogram images. The accuracy, sensitivity, and specificity for symptomatic and asymptomatic patients were 96.5%, 96.42%, and 95.47% and 98.85%, 97.01%, and 99.6%, respectively. For breath sound spectrogram images, the metrics for binary classification of symptomatic and asymptomatic patients were 91.03%, 88.9%, and 91.5% and 80.01%, 72.04%, and 82.67%, respectively. Conclusion-The web-application QUCoughScope records coughing and breathing sounds, converts them to a spectrogram, and applies the best-performing machine learning model to classify the COVID-19 patients and healthy subjects. The result is then reported back to the test user in the application interface. Therefore, this novel system can be used by patients in their premises as a pre-screening method to aid COVID-19 diagnosis by prioritizing the patients for RT-PCR testing and thereby reducing the risk of spreading of the disease.

Detection and Severity Classification of COVID-19 in CT Images Using Deep Learning.

Detecting COVID-19 at an early stage is essential to reduce the mortality risk of the patients. In this study, a cascaded system is proposed to segment the lung, detect, localize, and quantify COVID-19 infections from computed tomography images. An extensive set of experiments were performed using Encoder-Decoder Convolutional Neural Networks (ED-CNNs), UNet, and Feature Pyramid Network (FPN), with different backbone (encoder) structures using the variants of DenseNet and ResNet. The conducted experiments for lung region segmentation showed a Dice Similarity Coefficient (DSC) of 97.19% and Intersection over Union (IoU) of 95.10% using U-Net model with the DenseNet 161 encoder. Furthermore, the proposed system achieved an elegant performance for COVID-19 infection segmentation with a DSC of 94.13% and IoU of 91.85% using the FPN with DenseNet201 encoder. The proposed system can reliably localize infections of various shapes and sizes, especially small infection regions, which are rarely considered in recent studies. Moreover, the proposed system achieved high COVID-19 detection performance with 99.64% sensitivity and 98.72% specificity. Finally, the system was able to discriminate between different severity levels of COVID-19 infection over a dataset of 1110 subjects with sensitivity values of 98.3%, 71.2%, 77.8%, and 100% for mild, moderate, severe, and critical, respectively.