AI-Driven Biomarker Detection in Medical Diagnostics

Non-invasive monitoring of blood analyte levels using electrocardiograms (ECGs) and machine learning. The technique involves training a machine learning model using ECGs and corresponding analyte levels from multiple individuals. The model can then predict analyte levels from new ECGs without needing blood samples. The training data is processed to filter out potentially inaccurate analyte measurements that could negatively impact model accuracy. This involves analyzing the ECGs to identify features that correlate with analyte levels, as some subtle changes in the ECG may indicate analyte fluctuations that are not obvious to the human eye. The model is trained using these identified ECG features rather than just the analyte measurements. By leveraging the relationship between ECGs and analyte levels, the model can accurately predict analyte concentrations without invasive blood draws.

Source

Using graph neural networks to predict disease progression, survival, and therapy response for multi-focal diseases like cancer or COPD based on graph representations of the patient's disease lesions. The method involves applying a trained graph machine learning model to a graph representation of the patient's lesions to generate clinical information for predicting disease outcomes. The graph representation can include both local and global features of the lesions. The trained model considers the full distribution of lesions across organs rather than just isolated lesions. This holistic approach improves prediction accuracy compared to using isolated lesion features.

Source

A system for detecting dielectric changes in matter using microwave scattering, machine learning, and big data to detect diseases like stroke. The system involves collecting microwave scattering data from an array, analyzing it with a learning device, comparing results to imaging, and iteratively improving the learning. The learning uses correlations between microwave and imaging to predict disease states. It leverages big data to learn disease signatures from scattering. This provides a cheap, portable, scalable alternative to traditional imaging for rapid, intelligent diagnostics using microwaves.

Source

A multivariate classification system and method for identifying reliable biomarkers for diseases like hepatocellular carcinoma. The system reduces the dimensionality of biomarker features to improve classification accuracy by calculating effect sizes of differences between groups and using weights and stability metrics to group features. It then extracts weighted features above a threshold as biomarkers for classification. This reduces the number of features and improves robustness compared to direct univariate analysis.

Source

A biomarker identification method using genetic algorithms to find optimal subsets of genes that can predict and diagnose diseases from high-dimensional gene microarray data. The method involves filtering the genes initially using a technique like mRMR, then applying multiple machine learning algorithms to select features. The feature sets from each algorithm are combined and initialized the genetic algorithm population. Improvements like clustering balance global and local search. The genetic algorithm finds the best feature subset for biomarker identification.

Source

Model-based featurization and classifiers for predicting disease states from nucleic acid samples like blood tests. The technique involves using tissue models to determine the tissue origin of sequencing reads from cell-free DNA. The featurization process assigns each read to the tissue model with the highest likelihood of origin. This tissue-based feature vector is then used to train classifiers for detecting diseases. The classifiers generate signal vectors representing disease state detection. By training and testing on labeled samples, cutoff thresholds are determined for each disease state based on the labels. This allows detecting diseases from unknown samples by applying the labeled cutoffs to their signal vectors.

Source

A method for selecting biomarkers for machine learning models to accurately predict treatment response for diseases. The method involves partitioning the set of biomarker genes into subsets, applying multiple machine learning models to each subset to predict treatment response, generating an ensembled feature set of the top performing biomarkers, and then further selecting a final set of biomarkers from the ensembled set based on their impact on the final treatment prediction. This iterative process refines the biomarker selection for more accurate treatment response prediction.

Source

A multi-model feature selection method and system for machine learning applications in medical diagnosis that improves accuracy and reduces computational complexity by selecting relevant features with high contribution. The method involves using multiple machine learning models to analyze feature contribution, assign weights based on model accuracy, calculate feature scores, sort and select the top half as the optimal feature subset. This reduces the number of features while retaining the important ones.

Source

Detecting and monitoring cancer using DNA methylation analysis. The method involves identifying a panel of methylation signatures associated with multiple cancer types. A machine learning model is trained using this panel to distinguish between healthy individuals and those with cancer. By sequencing DNA from a subject and applying the model, cancer detection is possible. The model can also determine the tissue of origin for multi-cancer cases. The method aims to improve early cancer detection with high sensitivity and specificity using a universal panel of cancer-associated methylation signatures.

Source

Automatically detecting and quantifying pathology in medical imaging scans using AI techniques like unsupervised learning to identify and classify irregularities without requiring a human expert. The method involves training a computer system to analyze scanned images and quantify the extent of pathology patterns. It can be used to automatically identify pathologies like usual interstitial pneumonia (UIP) in lung CT scans for diagnosing conditions like idiopathic pulmonary fibrosis. The AI system learns to recognize pathology patterns without human supervision.

Source

A method and system for accurately predicting cervical cancer using deep feature selection to integrate multiple omics data. The method involves using a Lasso (least absolute shrinkage and selection operator) deep feature selection algorithm to extract relevant features from multi-omics data like genomic, epigenomic, transcriptomic, and proteomic data. The Lasso algorithm helps avoid overfitting and improves model generalization by shrinking and selecting important features. The extracted features are then used to train a prediction model for cervical cancer screening. The Lasso deep feature selection enables better accuracy compared to directly using the raw multi-omics data for prediction.

Source

Explainable AI models using CNN and attention for early, non-invasive diagnostics of Alzheimer's Disease by analyzing patient's speech patterns. The models can detect Alzheimer's using two types of features: parts-of-speech (PoS) and language embeddings. Attention layers capture the relative importance of features within each class and between classes. This provides explainability by showing which features are most important for the model's decision-making.

Source

Using machine learning algorithms trained on historical insurance claims data and biomarker levels to identify and monitor diseases like multiple sclerosis before symptoms become apparent. The algorithms can predict disease risk and progression based on patterns in insurance claims data like diagnosis codes, treatments, and demographics. Blood biomarkers like long non-coding RNA levels can also be used. This allows early detection, prognosis, and personalized treatment recommendations for diseases like MS.

Source

Method for diagnosing and staging liver diseases using DNA analysis. The method involves analyzing a DNA sample, like circulating cell-free DNA (cfDNA) or blood cell DNA, to determine methylation patterns at specific CpG sites. Methylation levels are calculated from these patterns and used to classify the liver disease. This allows non-invasive diagnosis and staging of conditions like fatty liver, NASH, cirrhosis, and liver cancer without biopsy. The methylation markers are selected based on tissue specificity and can distinguish between liver and non-liver tissue.

Source

Extracting biomarkers from subject response data like video and audio to determine disease severity. The method involves presenting stimuli to a subject and recording their response. Biomarkers are derived from the recorded data by extracting descriptors like facial expressions, speech patterns, and vocal characteristics. These descriptors are used to calculate biomarkers that indicate disease severity.

Source

Extracting biomarkers for disease diagnosis using a deep neural network from a limited number of samples. The method involves analyzing genomic data from a small set of samples to extract biomarkers for disease diagnosis. The method uses repeated random sampling of the limited data to train and evaluate neural networks. It determines important genomic features that affect network accuracy. By synthesizing features from multiple networks, it ranks and selects biomarkers from the limited data. This allows biomarker extraction using neural networks even with a small number of samples, unlike conventional methods needing large datasets.

Source

A method for identifying colorectal cancer molecular markers using multi-omics data that involves preprocessing the data, training neural network models, and identifying important features. The method aims to improve accuracy of colorectal cancer molecular marker identification compared to single-omics methods. It starts by screening the omics data based on correlation with colorectal cancer. Then, it trains neural network models using the filtered data. Finally, it identifies features from the models that have the greatest impact on classification to find potential colorectal cancer molecular markers.

Source

A blood test-based method using machine learning to diagnose and predict COVID-19 using routine blood test data. The method involves training a machine learning model using blood test data from COVID-19 patients and healthy controls to identify biomarkers that distinguish COVID-19 cases. This model can then be used to predict COVID-19 status from new blood test data. The method aims to provide a cheaper, faster, and more accessible diagnostic alternative to PCR testing, particularly in resource-limited settings where PCR testing may not be feasible.

Source

A method to predict diseases using biosignals like EEG, ECG, etc. that involves learning the biosignals using vectorization and deep learning techniques. The method involves expressing the biosignals as multiple symbols based on a condition, converting each symbol to an N-dimensional vector, summing the vectors to get a learnable vector, and using it to predict diseases in new biosignals by comparing against extracted vectors from those biosignals. This allows learning and analyzing biosignals in real time to predict diseases based on what's learned.

Source

Determining biomarkers and health conditions of patients related to cognitive health using multimodal health data and machine learning. The method involves receiving diverse health data like speech, gait, eye movements, etc. from multiple modalities. A neural network extracts latent variables from the health data. These latent variables are provided to another neural network to determine biomarkers and health conditions. The system aims to provide reliable, noninvasive, accurate, and cost-effective determination of cognitive impairment, optimized care, and treatment recommendations.

Source

Improving lung cancer diagnosis by using AI to accurately assess lung nodules based on both lung scans and blood tests. The AI classifier identifies image markers like vascular convergence and air bronchograms in lung scans, as well as blood markers like genomic, epigenomic, and proteomic biomarkers. It also considers factors like discrepancies between scans over time and comorbidities. The AI combines this multimodal data to provide a more accurate lung nodule risk score than using just lung scans.

Source

A method to accurately predict Alzheimer's Disease (AD) from Mild Cognitive Impairment (MCI) patients by using a combination of brain MRI scans and cognitive assessment scores. The method involves training a machine learning model using MRI images containing hippocampus structure and cognitive function scores. The model learns to predict AD based on both structural and functional biomarkers. The hippocampus is segmented into sections and features like volume, surface area, and curvature are extracted. These features are combined with cognitive scores for AD prediction.

Source

Using CT scans to predict molecular biomarkers like PD-L1 and EGFR mutations in lung cancer without invasive biopsies. The method involves processing CT images to estimate molecular data related to biomarkers in the genome, transcriptome, proteome, and metabolome. Nodule patches are identified from the CTs, and nodule-based data is calculated for the biomarkers. This nodule-based data is used as input to trained functions along with CT features to predict molecular data without needing CT images of the entire lung.

Source

A device that uses artificial intelligence to predict if a person who doesn't yet have Alzheimer's disease will develop the condition. The device takes inputs like age, gender, genotype, and brain scan results to train an AI model that can predict beta-amyloid conversion status. This allows identifying individuals who are at higher risk of developing Alzheimer's at an earlier stage compared to waiting for amyloid deposits. The goal is primary prevention by targeting interventions to those at higher risk of progression.

Source

Detecting health issues like heart conditions and brain disorders by analyzing electromagnetic fields (EMF) from the body. A machine learning algorithm trained on EMF data from known conditions is used to identify abnormalities in EMFs from unknown individuals. The algorithm extracts features from the EMFs, associates them with other data, generates a hypothesis function, and determines the presence of abnormalities. This allows diagnosing conditions like heart ischemia or brain disorders without direct imaging or testing.

Source

Automated method for diagnosing and staging non-alcoholic steatohepatitis (NASH) and liver fibrosis using deep learning and image analysis. The method involves developing automated algorithms to recognize and score histological features like ballooning, inflammation, fibrosis, and steatosis from liver biopsy images. This allows fully automated diagnosis and staging of NASH and fibrosis without subjective manual interpretation. The algorithms use deep learning models to predict pathological patterns and morphometric measurements from tissue images. The automated method improves reproducibility and accuracy compared to manual scoring by pathologists.

Source

Method for diagnosing Alzheimer's disease progression using brain network analysis. The method involves extracting features from neuroimaging data like fMRI, converting brain networks to feature vectors using graph embedding, selecting relevant features, and classifying between Alzheimer's, mild cognitive impairment, and normal controls using machine learning techniques like SVM and ELM. The graph embedding technique called node2vec helps transform the brain network graphs into vector representations.

Source

Cancer diagnosis device using machine learning to aid in cancer diagnosis by analyzing DNA methylation, gene expression, and copy number variation data. The device receives these cancer biomarker datasets from a patient, fuses the features using attention mechanisms, reduces the dimensionality, and maps the data to probabilities of different cancer types. It can also be used as a training method for the device. The device aims to improve accuracy of cancer diagnosis by leveraging multi-omics data and reducing feature redundancy.

Source

Early warning system for epidemic and pandemic diseases using personalized biomarkers and AI to monitor health and identify diseases. The system involves establishing a baseline of biomarkers for an individual's genes, proteins, and cell-free DNA. Changes in these biomarkers over time are monitored and correlated with diseases. An AI algorithm compares biomarker patterns across multiple individuals to predict diseases. By leveraging personalized biomarkers and AI, the system aims to provide early indications of diseases and epidemics by continuously monitoring health data.

Source

Generating aptamers using machine learning to design DNA sequences that bind to target biomolecules. The method involves training a restricted Boltzmann machine (RBM) neural network using datasets of aptamers that bind to a target biomolecule. The RBM learns the probability distribution of aptamer sequences that bind. It can then generate new aptamer sequences with high binding probability. This allows synthesizing new aptamers without requiring repeated SELEX selection rounds. The RBM can also classify aptamers into binders vs non-binders based on their sequences.

Source

Developing machine learning systems for disease diagnostics using peptide array data from multiple conditions. The systems are trained on peptide sequence and binding data from a diverse range of conditions, rather than just one condition. This improves performance for diagnosing individual conditions compared to models trained on just that condition. The method involves representing peptide sequences as dense compact vectors and using regressors to learn correlations between sequence and binding values. The regressors are then used in classifiers to diagnose conditions based on peptide data.

Source

Detecting health abnormalities like cancer using liquid biopsy and machine learning. The method involves training a machine learning classifier to detect health abnormalities from liquid biopsy data. The training uses a combination of original data and synthetic data generated by kernel density estimation to balance the number of samples with and without the abnormality. This modified training set is used to select the best classifier for detecting the abnormality in new liquid biopsy samples. The selected classifier provides a diagnostic test result indicating if a patient has the health abnormality based on their liquid biopsy data.

Source

Method for diagnosing and subtyping complex diseases using small RNA biomarkers. The method involves identifying small RNA sequences associated with disease states from a discovery sample set. Machine learning is then used to build classifiers that can accurately diagnose and subtype diseases based on the presence or abundance of those small RNA sequences in new samples. This allows classification of diseases that may have similar symptoms or pathologies, including early stages, and identification of previously unlabeled disease subtypes.

Source

A method for identifying prognostic markers in complex diseases using fusion of multi-source biological data. The method involves integrating multiple biological networks, embedding the nodes into low-dimensional vectors, constructing a two-layer heterogeneous network, denoising it, and using network propagation to rank features as prognostic biomarkers. This leverages multiple types of biological data to identify more accurate and interpretable prognostic markers compared to single network methods.

Source

Automated tool using machine learning and generative adversarial networks (GANs) to improve diagnosis of Parkinson's disease (PD) from MRI scans. The tool leverages GANs to generate synthetic MRI images that are used to augment the training dataset. This helps the classification models learn better features from the real scans. The tool achieved a 16% increase in accuracy compared to traditional methods. The augmented dataset allowed the models to diagnose PD with a higher accuracy than using just the real scans. The tool demonstrates the feasibility of using GANs to generate unseen medical images for improving classification accuracy.

Source

Building a diagnostic classifier for colorectal cancer using machine learning on specific miRNAs found in colorectal cancer patient serum. The classifier accurately distinguishes CRC samples from healthy samples based on the expression levels of four miRNAs (miR-654-5p, miR-126, miR-10b, and miR-144) identified from exosomes in serum. This provides a potential early diagnostic tool for colorectal cancer using blood tests.

Source

Determining movement biomarkers from video of a subject using optical flow analysis and an autoencoder. The method involves generating optical flows from the video frames, encoding the flows using an autoencoder to extract movement features, and interpreting the encoded features as movement biomarkers. This allows automated, objective quantification of movement characteristics like tremor frequency from video. The autoencoder learns to separate and quantify movement types and values from the optical flows. It can be trained using labeled videos to accurately determine movement biomarkers. The encoded features can then be used for diagnosis, monitoring, and progression assessment of movement disorders like Parkinson's.

Source

Automatically identifying biomarkers that are associated with physiological variables like muscle mass, by leveraging machine learning models to rank and screen candidate biomarkers based on their predictive power. The method involves inputting a target physiological variable and a set of potential biomarkers into multiple machine learning models. The biomarkers are then ranked based on their model outputs. Importance scores are calculated using a screening condition to select top-ranked biomarkers as candidates. Finally, the correlation between the candidates and the target physiological variable is calculated to identify the actual biomarker(s) that influence it.

Source

A method for detecting Alzheimer's disease using brain fMRI scans, especially when the available training data is limited. The method involves transforming the fMRI feature spaces of different datasets into a common subspace using a data space transformation technique. This allows combining and expanding the training data from multiple sources into a larger, more diverse dataset. The transformed features are then used to train a machine learning model for Alzheimer's diagnosis.

Source

Predicting clinical parameters like vision loss or progression of retinal diseases like diabetic macular edema (DME) using fluid volumes measured from optical coherence tomography (OCT) images. The method involves extracting subretinal and intraretinal fluid volumes from OCT scans, generating metrics based on total retinal volume, fluid volumes, and segmentation, and using machine learning models to predict clinical parameters like visual acuity, DME progression risk, etc. from these metrics.

Source

Predicting progression of diseases like idiopathic pulmonary fibrosis (IPF) using medical images and machine learning. The method involves selecting important features from CT scans to stratify patients into groups likely to progress versus stable. A machine learning algorithm trained on CT images with labeled regions indicating progression is used to identify regions in new scans that are expected to reflect progressive pulmonary fibrosis. The algorithm is trained using a quantum particle swarm optimization technique to select the optimal feature subset for classification. This allows early identification of patients at high risk of progression for intervention.

Source

Detecting the etiological factors of diseases like cancer by using machine learning techniques that can identify signatures specific to factors like aging, obesity, smoking, and genetic mutations. The signatures are generated using a supervised learning approach where mutational patterns are trained on labeled data with known factors. This allows more accurate prediction of exposure compared to unsupervised methods. The signatures can vary by tissue type and provide insight into tissue-specific etiology. The method involves feature engineering with low-variance, variable-length mutations, and training a model on these features.

Source

Computer-implemented system and method for diagnosing liver diseases like hepatocellular carcinoma using neural networks. The system involves using dense block neural networks to analyze CT liver images and determine the presence of liver lesions indicative of HCC. The dense block architecture allows direct transmission of features from input to output layers. The neural networks are trained to distinguish HCC from non-HCC CT images. The systems can provide diagnostic assistance for HCC detection. The dense block models outperformed other architectures in accuracy, specificity, and positive predictive value for HCC classification.

Source

Automatically building reliable and disease-specific analysis models for predicting disease status using digital biomarkers. The method involves collecting historical digital biomarker data, splitting it into training and test sets, training a model on the training set, and evaluating the model's accuracy on the test set. This process is repeated with different models and hyperparameters to find the best one for predicting disease status based on digital biomarkers. The trained model can then be used to predict disease status from new digital biomarker data.

Source

Automated estimation of clinical biomarkers using machine learning on continuous sensor data from everyday activities to replace specialized tests. The system involves extracting relevant sensor segments using ML models, then estimating biomarkers using further ML models. It trains ML models for biomarker extraction and classification on labeled sensor data, then trains biomarker estimation models on labeled estimated biomarkers. This allows calculating biomarkers continuously from everyday sensor data without specialized equipment or frequent tests.

Source

An overall screening method for identifying proteomic biomarkers for clinical applications. The method involves data preprocessing, feature pre-screening, feature extraction, and multiple machine learning feature selection algorithms to identify the best biomarker combinations. It aims to improve the practical application results of biomarkers by using a combination of techniques to screen for high-quality biomarkers that can accurately classify and predict clinical samples.

Source

Identifying individuals who are potentially affected by a medical condition using AI. The method involves training an AI classifier using positive and negative indications of a specific medical criterion directly observed in a group of test subjects. The classifier is then applied to medical profiles of a larger population to identify individuals who are potentially affected by the same condition. This allows identifying people who may have the condition even if they haven't been directly tested for it. The AI uses learned patterns from the test group to make predictions for the larger population.

Source

A colorectal cancer detection system using machine learning techniques based on the XGBoost algorithm. The system has modules for data acquisition, preprocessing, feature selection, model building, and result prediction. It leverages XGBoost's gradient boosting algorithm to accurately analyze and predict colorectal cancer based on medical data. The feature selection module uses recursive feature elimination with logistic regression to select the most representative features. This helps improve algorithm performance, reduce complexity, and ensure accurate predictions with a large number of input features.

Source

Accurately diagnosing cancer using artificial intelligence and liquid biopsy of exosomes. The method involves training a deep learning model using SERS signals of cultured cell exosomes to analyze blood exosome SERS signals. The trained model allows more accurate classification of cancer exosomes in patient blood compared to traditional methods. By calculating the relative similarity of blood exosome data to cancer cell exosome data, the model can diagnose cancers.

Source

A computer-based method to identify genetic biomarkers for diseases like Alzheimer's by integrating quantitative imaging and genome analysis. It starts with accurate quantitative imaging of disease phenotypes like amyloid PET, then does genome-wide association (GWAS) to find genetic biomarkers. Clinical response predictions against these biomarkers validate them. The method uses quantitative imaging accuracy, genome analysis depth, and clinical response prediction to find disease biomarkers with just hundreds of subjects, compared to thousands needed for GWAS alone. It generates reports with identified biomarkers and therapeutic targets.

Source