Medical Data Mining TRIZ Cases
- Electronic Health Record System for Patient-Controlled Data Sharing
Patient-controlled electronic health records consolidate fragmented medical data, eliminating transfer delays and reducing duplicate tests.
- Physiological Monitoring for Drug Profile Characterization
A physiological monitoring system acquires sensor data to generate drug profiles characterizing substances affecting a subject.
- Medical Image Search Feature Space Projection
Computer system projects medical images into multi-dimensional feature space for visual and semantic similarity retrieval.
- CGM Sensor Tracking Insertion Removal Times
A continuous glucose monitoring system tracks sensor insertion and removal times using stored identifiers.
- Exponentially Decaying Memory for Neural Signal Classification
An exponentially decaying memory model classifies neural signals while preserving long-term characteristics within limited implantable device memory.
- Automatic Electroanatomic Map Generation from Morphologic Beat Classification
Automated system classifies intracardiac beats by morphology to generate distinct electroanatomic maps for simultaneous arrhythmia analysis.
- Neural Network Predictive Computing System for Clinical Trial Patient Screening
Automated neural network training eliminates manual feature engineering bottlenecks, increasing prediction accuracy for patient health conditions.
- Physiological Information Processing Abnormality Index
A physiological information processing apparatus calculates an abnormality index from autonomic nerve parameters to display diagnostic data.
- Learning Assistance Device for Confidential Medical Model Aggregation
A learning assistance device aggregates learned discriminators from terminal devices to improve discrimination performance.
- Wearable Device EMG Signal Processing for Motion Artifact Removal
A wearable device system filters acceleration and bio-impedance signals to generate accurate motion data.
- Multi-Objective Optimization Using Composite Novelty Selection
A multi-objective algorithm uses composite functions and novelty measures to guide search toward useful tradeoffs.
- Machine Learning Infection Classification System
Machine learning algorithms classify primary and latent herpesvirus infections from antibody profiles, resolving ambiguity in infection timing determination.
- Attention Module Machine Learning Model for Fetal Gestational Age Estimation
A trained machine learning model with an attention module aggregates weighted feature vectors from ultrasound images to estimate fetal gestational age.
- Synthetic Electronic Health Records via Generative Adversarial Networks
Generative adversarial networks produce synthetic patient data vectors that replicate authentic medical record patterns for machine learning applications.
- Patient Monitoring via Code Intermediary for Secure Data Transfer
Generating measurement data as a 1D or 2D code allows a user terminal to recognize and transmit this data to a server, eliminating wireless approval hurdles.
- ECG Respiratory Effort Detection via Signal Modulation Analysis
Analyzing ECG signal amplitude and frequency modulations to determine patient respiratory effort.
- Cardiovascular Parameter Calculation Using Multivariate Models
Multivariate statistical models determine arterial pressure waveform data to calculate cardiovascular parameters.
- Intermediary Knowledge Base for Cross-Standard Data Mapping
A central knowledge base mediates translations between distinct coding systems to establish one-to-one entity mappings.
- Medical Care Assistance System for Efficient Examination Result Input
A medical care assistance system arranges relevant examination options on an input screen to streamline the reporting workflow.
- Medical Information Navigation Engine for Clinical Knowledge Extraction
A Medical Information Navigation Engine extracts coded elements from clinical history to calculate reimbursement potential.
- Informed Consent Codification for Human Specimen Research Compliance
Automates regulatory compliance analysis by tracking consent rule changes against a global intelligence knowledgebase.
- Multi-Graph Attention Network for TCM Syndrome Classification
A multi-graph attention network aggregates symptom and syndrome element features to classify traditional Chinese medicine syndromes.
- Comorbidity-Stratified Cancer Treatment via Gene Expression Analysis
Gene expression analysis identifies comorbidity-linked genes, resolving the trade-off between prediction accuracy and assessment complexity.
- Surgical Guideline Indicator Mapping for Automated Bill Adjudication
Pre-computed mappings between procedure codes and clinical guidelines enable automated adjudication recommendations, reducing manual review errors.
- Multi-Task Learning with Dynamic Imputation for Missing Data
Dynamic imputation fills missing features during training to preserve dataset representativeness and reduce bias in multi-task learning models.
- Acoustic Anhedonia Classification via Speech Feature Extraction
Audio sensors capture speech samples for automated acoustic feature extraction and anhedonia status classification.
- Medical Image Display Segmentation for Diagnostic Accuracy
A medical image display system segments case areas by disease name to align similar images in columns.
- Unsupervised Template Extraction for Question Answering Systems
Automated relationship template generation via hierarchical clustering and distributional semantics.
- Sleep Stage Classification Using Heart Rate Variability Parameters
A sleep determination apparatus extracts heart rate variability parameters to classify non-REM stages using machine learning.
- Multistage AI Model Integrating CTG and Maternal Data for Labor Predictions
A multistage AI model combines cardiotocography and maternal health data to generate labor predictions.
- N-Dimensional Well-Being Profile System for Personalized Intervention
A recommendation system creates n-dimensional profiles to track user well-being changes over time through personalized intervention cycles.
- Deep Learning Encoder for Multi-Modal Patient Data Similarity
A deep learning encoder transforms multi-modal patient data into compact signatures for efficient similarity assessment.
- Adaptive T-Wave Detection Algorithm for Deviated ECG Waveforms
A physiological information processing apparatus classifies electrocardiogram data to select an appropriate analysis algorithm from a plurality of options.
- Biofluid Diagnostic Screening System Using Imaging Analysis
Lateral flow test strip paired with an imaging device captures detection zone images to calculate analyte concentrations for patient biofluids.
- GAN Imputation Model for Missing Medical Diagnosis Data
Generative adversarial network imputes missing medical diagnosis data using binary search optimization.
- Vehicle Hypoglycemic Detection via Multi-Sensor Fusion
A vehicle hypoglycemic detection system fuses physiological and driving data to predict events.
- Deep Neural Network for ELISPOT Spot Detection
A deep neural network analyzes ELISPOT images to distinguish true signals from background noise without manual parameter tuning.
- Orthostatic Hypotension Detection Using Heart Rate Data Analysis
An AI model analyzes heart rate and activity data to detect orthostatic hypotension, eliminating the need for time-consuming tilt tests.
- Disease Prediction System Integrating Biosignal Data and Medical Knowledge
A disease prediction system merges biosignal data with a medical knowledge base to generate personalized health analysis results.
- Risk Score Determination Using Sensor Data and Machine Learning
Machine learning models analyze sensor data to determine user risk scores, replacing demographic reliance with behavioral pattern accuracy.