USPTO Patent Applications - Health Informatics (G16H)

Ring-Type Wearable Device for Biometric Health Monitoring

USPTO published patent application US20260096777A1 on April 9, 2026, filed on May 8, 2025, by inventors Thomas Voigt and Swagatika Bhattacharya. The application covers a ring-type wearable device with multiple LEDs and photodetectors for biometric health monitoring. The published application does not grant any rights and represents an early-stage IP filing rather than an issued patent.

Pain Detection Method Using Sensors in Personal Care Devices

USPTO published patent application US20260096772A1 titled "Pain Detection Method" on April 9, 2026, filed by six inventors including Amir Hussein Rmaile, Mark Thomas Johnson, and others. The application (No. 19114825, filed September 26, 2023) covers a method for objectively measuring user pain during personal care sessions using force and/or motion sensors integrated into personal care devices such as oral care devices. The sensor data patterns can infer pain levels, pain reaction events at specific points or locations, and increased sensitivity in certain body areas.

Insulin Dosage Patent Application by Stone, Mikhno, Turksoy, and Grossman

USPTO published patent application US20260096775A1 titled 'Open Loop Intelligent Dosing' on April 9, 2026. The application covers systems and techniques for generating personalized insulin dosage recommendations by building a physiological model from a patient's historical glucose and insulin data, estimating the impact of meal consumption on glucose levels, and outputting a personalized meal bolus insulin dose recommendation. The application was filed on September 16, 2025 under Application No. 19330568, with four named inventors: Michael P. Stone, Arthur Mikhno, Kamuran Turksoy, and Benyamin Grossman. This is a published patent application record; it has not yet been granted and creates no immediate compliance obligations.

Bayesian Denoising for Retrospective PISA Detection in Analyte Traces

The USPTO published patent application US20260096780A1 for a Bayesian denoising method to detect pressure-induced sensor artifacts (PISA) in analyte traces. The method receives measured analyte data samples from a sensor, generates a reconstructed trace with confidence windows using a Bayesian algorithm, and compares the two to identify artifacts outside the confidence interval. Inventors include Andrea Facchinetti, Simone Del Favero, Giovanni Sparacino, Elena Idi, Eleonora Manzoni, and Nunzio Camerlingo. The application (No. 18908001) was filed October 7, 2024 and published April 9, 2026.

System to Monitor and Manage Patient Hydration via Plethysmograph Variability Index in Response to Passive Leg Raising

The USPTO published patent application US20260096773A1 on April 9, 2026, filed September 2, 2025, for a system using a processor to predict fluid responsiveness or unresponsiveness in patients based on plethysmograph variability parameters and limb-elevation fluid responsiveness parameters, then causing administration or termination of fluid delivery based on the combined prediction. The named inventors are Massi Joe E. Kiani, Bilal Muhsin, and Keith Ward Indorf. CPC classifications include A61B 5/4848 and G16H 20/17 (Health Informatics), with the application originating from US application number 19316927.

Systems and Methods for Automatic State Estimation of Current Imaging Exam Using User Actions on Console Screen

USPTO published patent application US20260096786A1 filed by seven inventors including SIVA CHAITANYA CHADUVULA, EKIN KOKER, OLGA STAROBINETS, RANJITH NAVEEN TELLIS, SANDEEP MADHUKAR DALAL, THOMAS ERIK AMTHOR, and YUECHEN QIAN. The application covers a system for automatically estimating the state of a medical imaging examination using video analysis and a state machine algorithm implemented in electronic processors, with results displayed on an electronic processing device. The filing date was September 19, 2023, with application number 19112526.

Anumana Patent for Non-Invasive ECG and Echo Monitoring with Machine Learning

Anumana, Inc. filed patent application US20260096783A1 on April 9, 2026 (filed October 4, 2024 as Application No. 18907123) for an apparatus and method combining non-invasive ECG and echocardiogram data with machine learning to generate simulated invasive measurement data. The system fuses ECG and echo measurements and uses trained ML models to produce data equivalent to invasive cardiac measurements, transmitting results to remote devices. CPC classifications span cardiac monitoring (A61B 5/7267, A61B 5/256, A61B 5/346), medical imaging (G06T 15/005), and health informatics (G16H 50/20, G16H 50/70).

Ultrasonic Diagnostic Device with Blood Flow Processing

USPTO published patent application US20260096809A1 for an ultrasonic diagnostic device with one or more processors configured to acquire measurement data including both tissue-derived and blood-flow-derived information, compute a deformation amount between a plurality of frames, and generate display image data by aligning and synthesizing feature points across those frames. The application was filed on December 10, 2025 under Application No. 19415098, naming Shoya Sasaki, Naoya Iizuka, Takeshi Sato, and Hiroki Takahashi as inventors.

CT Polyp Detection Using Dual CNN Architecture for Medical Imaging

USPTO published patent application US20260096796A1 for a CT polyp detection system employing dual convolutional neural network (CNN) architectures. The method receives CT scan image data, segments colon portions using a first CNN to identify candidate polyps, then validates candidates through multiplanar analysis using a second model to classify each as a polyp or non-polyp. The four named inventors are Aly Farag, Mohamed Yousuf, Samir Harb, and Asem Ali. The application was filed October 8, 2025 and published April 9, 2026.

Systems and Methods for Drug Interaction Analysis

The USPTO published Patent Application US20260112508A1 for systems and methods analyzing drug-drug and gene-drug interactions. The disclosed invention utilizes a database comprising genes associated with a plurality of drugs based on clinical relevance, and provides recommendations including alternative drugs when problematic interactions are identified. Inventors: Bill Massey, Christopher Diaz, Nicholas Glimcher. Application No. 19117365 filed October 5, 2023.

AI Estimates Biological Age from Facial, Tongue, Retina

The USPTO published patent application US20260112504A1, filed December 10, 2025, for methods, devices, and systems that use a multi-modal transformer-based architecture with cross-attention combining facial, tongue, and retina images to estimate biological age (BA). The resulting AgeDiff metric (chronological age minus biological age) can function as a standalone biomarker or combine with other factors for chronic disease risk stratification and progression prediction. Inventors are Yuanxu GAO and Kang ZHANG, with application number 19414736.

AML JV Patents Biomarker Correlation System for Condition Identification

AML JV, LLC has filed patent application US20260112506A1 for systems and methods that correlate biomarkers to identify condition indicators. The application, filed on October 21, 2025, covers a system that obtains individual profile data, identifies profiles indicative of a condition based on a first biomarker, determines attributes among profiles, and finds correlations between attribute subsets indicative of different biomarkers. The technology applies CPC classification G16H 50/70 (Health Informatics) and names five inventors.

Machine-Learned Model Architecture for Generating Individual-Specific Instructions

USPTO published patent application US20260112505A1 filed by Bilikis J. Oladimeji, Reem A. Hussain, and Aaron C. Wacker on October 23, 2024, covering systems and methods for using generative language models to generate individual-specific health instructions. The invention maps a condition to an ontological representation, incorporates individual-specific information, determines comorbidity relationships, and generates prompts based on property data associated with condition and comorbidity concepts. CPC classifications G16H 50/70 and G16H 70/20 place this in health informatics.

NEC Planning Device Health Evaluation Patent Application

NEC Corporation has filed a patent application (US20260112502A1) for a planning device that evaluates health states by acquiring feasibility evaluation index values for each state and health-level-evaluation-index values. The device searches for optimal paths from an initial state to a target state by calculating evaluation values using state transitions, where better evaluations occur when the transition-destination state has higher feasibility and better health-level indices relative to the transition-source state. The application was filed on October 16, 2025, under CPC classification G16H 50/30 (Health Informatics).

BenevolentAI Patent Application, ML Identifies Biological Entities for Drug Discovery

BenevolentAI Technology Limited has filed US Patent Application US20260112507A1, published April 23, 2026, disclosing a computer-implemented method for training a machine learning model to identify biological entities for drug discovery. The method involves masking biological entity mentions in text sequences, encoding them into input representations, and training the model to predict unique entity identifiers, enabling automated target identification using full contextual information from biomedical text corpora. The application names five inventors including Dane Sterling Corneil, Maciej Ludwick Wiatrak, and Vinay Prashanth Subbiah, with Application No. 19380404 filed November 5, 2025.

Anatomical Feature-Based Medical Prediction

USPTO published patent application US20260112496A1 on April 23, 2026, covering methods and systems for anatomical feature-based medical predictions. The application describes capturing and processing images to extract anatomical features, fitting 3D models to craniofacial and oral cavity structures, and generating predictions for medical conditions such as obstructive sleep apnea. The application was filed September 17, 2025, with four named inventors: Jatin Maniar, Narayanan Ramanathan, Ikshvanku Barot, and Hitesh Kalra.

Yutaro Miyagawa Patents Health Prediction System, 23rd Apr

Yutaro Miyagawa has filed USPTO Application US20260112499A1 for a health prediction system that uses a processor to collect biological data from users, store the data in a database, analyze the stored data to predict health conditions, provide lifestyle guidance based on prediction results, and issue warnings when abnormal values are detected. The application was published on April 23, 2026, with a filing date of October 14, 2025, under CPC classification G16H 50/30 (Health Informatics).

Fully Autonomous Medical Solution, US Patent US20260112495A1

Mazen A. Al-Sinan and Ho-Hsun David Kuo filed USPTO Patent Application US20260112495A1 on December 17, 2025, published April 23, 2026, for a fully autonomous medical platform integrating diagnosis, treatment determination, and patient monitoring into a single computing device. The platform compiles user medical records from external sources, uses a disease/symptom database to automatically diagnose ailments and recommend treatments, and tracks recovery timelines with automatic adjustments for treatment changes. CPC classifications indicate primary focus on health informatics (G16H) across diagnosis, data standards, therapies, and management domains.

NEC Multi-Modal Disease Risk Prediction Device and Method

NEC Corporation filed US Patent Application US20260112498A1 on April 23, 2026, covering a risk prediction device and method that integrates multi-modal data using probability distribution in a latent space. The system handles missing modality data by acquiring relevance information between probability distributions of different modalities, then generates integrated probability distribution data for disease risk prediction. The application (19350495) was filed on October 6, 2025, with inventors Chenhui HUANG, Kansuke Wagata, and Fumiyuki Nihey.

Diabetic Status Model Using Voice Patent Application

KVI BRAVE FUND I INC. has filed US Patent Application 19423110 with the USPTO for methods and systems that generate a Type-II diabetes (T2DM) diabetic status prediction model using voice samples. The application, published April 23, 2026 as US20260112494A1, describes collecting voice samples from training subjects at multiple time points, extracting voice feature values for a plurality of voice features, and training a diabetic status model from the samples and features. CPC classifications include G16H 50/20 and G10L 15/02. Inventors are Yan FOSSAT, Jaycee Morgan KAUFMAN, Jouhyun Clare JEON, and Anirudh THOMMANDRAM.

Multimodal Clinical System Patent Application US20260112493A1

USPTO published patent application US20260112493A1 for a multimodal clinical system on April 23, 2026, filed December 10, 2025 as Application No. 19414783. The application covers a method for training a model using bidirectional multimodal attention blocks that learn and fuse representations within and across modalities simultaneously. Inventors are Yuanxu GAO and Kang ZHANG.

Voice-Based Blood Glucose Prediction via Machine Learning

KVI BRAVE FUND I INC. filed patent application US20260112492A1 on April 23, 2026 for systems and methods predicting blood glucose levels using voice samples via machine learning. Inventors Yan FOSSAT and Jouhyun JEONG developed methods receiving voice samples at multiple time points paired with corresponding blood glucose measurements, sorting samples into glucose level categories, extracting voice features, and generating a blood glucose prediction model based on selected features. The application was filed December 2, 2025 under application number 19405524.

Feline Hypertrophic Cardiomyopathy Risk Prediction System

Mary Alexis Seguin filed US Patent Application US20260112500A1 for a system predicting feline hypertrophic cardiomyopathy risk. The system uses an ensemble of machine learning diagnostic models trained on medical data to screen new patient data, escalating to a knowledge-based diagnostic model when risk is indicated. The system includes a user interface that acquires additional clinician input and provides risk assessments with guidance for further testing and treatment. The application was filed October 15, 2025, published April 23, 2026, and covers CPC classifications G16H 50/30, G06N 20/20, G16H 10/20, G16H 20/00, G16H 50/20, and G16H 50/70. This patent application covers health informatics and AI/ML-based diagnostic systems for veterinary cardiac disease, specifically targeting feline hypertrophic cardiomyopathy. Entities developing or licensing veterinary diagnostic software, AI-based health risk assessment tools, or feline healthcare technology may wish to review this application's claims for potential overlap with their own IP or product development roadmaps. The publication of this application establishes a priority date of October 15, 2025 for the disclosed methods.

NEC Multi-Modal Risk Estimation Device

NEC Corporation has filed patent application US20260112497A1 for a multi-modal risk estimation device that acquires data from multiple modalities, converts each into probability distributions in a latent space via encoder, predicts risks per modality, and integrates them using modality-specific weights. The system is designed to support disease risk estimation and decision-making regarding lifestyle interventions for subjects. The application was filed on October 6, 2025, under application number 19350429.

Baekburn Inc Gamification Smart Weight Device Patent Application

Baekburn Inc has filed US Patent Application US20260112501A1 for a gamification-based smart weight management device, published April 23, 2026. The application covers a system with six functional units: a next-day mission generation unit, lifestyle pattern data collection unit, victory achievement prediction unit, mission performance detection unit, action plan change unit, and user level determination unit. The invention aims to help users achieve weight management goals through personalized game-like missions that adapt based on analyzed lifestyle patterns.

Automated Chat-Based Care Coordination Agents Patent

The USPTO has published patent application US20260112483A1 for an AI-based system and method to coordinate care and device management for neurostimulation devices. The system uses a pre-trained large language model (LLM) and conversational agent to facilitate natural language interactions among multiple human users, including patients and healthcare providers, allowing coordinated conversations about device configuration and usage.

NEC Corporation Link Prediction Using Knowledge Graph Embeddings and NLP

NEC Corporation has filed patent application US20260112488A1 (Application No. 19352593) for an information processing apparatus and method that performs link prediction by combining knowledge graph embeddings with sentence embeddings from a trained natural language model. The system generates triples from a query, creates sentences from those triples, calculates scores using both NLP embeddings and knowledge graph embeddings, then aggregates the scores for prediction. The invention targets improved accuracy at lower computational cost for link prediction tasks. Filing date was October 8, 2025; publication date April 23, 2026.

Brain State Optimization Methods - Santarnecchi

Emiliano Santarnecchi filed patent application US20260112487A1 on June 5, 2023, which published on April 23, 2026, covering methods and apparatus for transitioning a brain from an initial state to a target state using non-invasive stimulation. The application discloses protocols estimated from the brain's structural composition and functional architecture, and also describes using brain information to inform computational general AI agent design. The publication represents new intellectual property in the neurotechnology and AI sectors.

Dassault Systemes Patent, Predicting Newborn Cardiovascular Behavior from Fetus Parameters

Dassault Systemes has obtained US Patent Application US20260112490A1, published April 23, 2026, covering a computer-implemented method for predicting cardiovascular behavior of a newborn at birth from fetus physiological parameters. The method involves obtaining fetus physiological data, applying a non-calibrated surrogate cardiovascular model, calibrating the model using data assimilation algorithms, and triggering physiological birth changes to predict newborn cardiovascular behavior. The application (No. 19362244) was filed October 17, 2025, with inventors Hernán Morales Varela and Philipp Christoph Weder.

Baby Changing Pad With On-Device Voice Command Processing

USPTO published patent application US20260112491A1 for a baby changing pad with on-device machine-learning voice command processing. The device includes at least one microphone, at least one speaker, and one or more processors that identify user commands by extracting acoustic features from audio streams and generate responses output via the speaker. The application was filed on 2025-10-17 under application number 19362259, with inventor Shaker Rawanbakhsh.

NEC Rehabilitation Planning System Using Machine Learning

NEC Corporation has filed USPTO patent application US20260112473A1 for a machine learning-based rehabilitation planning system. The system includes a rehabilitation pattern selection unit, an ability value prediction unit that predicts physical ability values after rehabilitation, a repetition control unit, and a determination unit that identifies optimal rehabilitation patterns. The invention aims to efficiently create personalized rehabilitation plans by iteratively selecting rehabilitation patterns and evaluating predicted ability outcomes against predetermined conditions.

KARL STORZ Hospital Report System Using Sensors

KARL STORZ SE & Co. KG filed US Patent Application US20260112468A1 on October 18, 2024 (Application No. 18920489) for a system and method of generating hospital reports using sensor data. The system captures procedure data from at least two sensors, detects and timestamps multiple events during the procedure, synchronizes the events based on their respective timestamps, and generates a machine-readable report containing the merged events. The patent application was published on April 23, 2026.

Medical Equipment Location and Identification System

USPTO published patent application US20260112478A1, filed October 20, 2025, for systems and methods of identifying and locating medical equipment and supplies. The invention involves transforming clinical names of medical equipment and supplies into inventory names, transmitting those names to an inventory database, and identifying at least one location of the medical equipment or supplies. The publication is an administrative notice of a pending patent application with no immediate compliance obligations for healthcare providers or medical device manufacturers.

Systems and Methods for Identifying Predictive Parameters Based on Clinical Studies

USPTO published patent application US20260112462A1 on April 23, 2026, filed February 6, 2023 under application number 18164877. The invention by inventors Andrew Busey, Amicia Devin Elliott, and Claire Aldridge discloses a system and method for identifying predictive parameters based on clinical study results using AI/machine learning to analyze biological and historical parameters across subject subsets. CPC classifications G16H 10/20 (health informatics) and G06N 20/00 (machine learning) indicate healthcare and AI applications.

Generative AI Method for Patient-Specific Electrical Stimulation Prescriptions

USPTO published patent application US20260112472A1 titled 'Method and System for Generating Electrical Stimulation Prescription' filed October 20, 2025, naming Fang-Cheng Liu, Chuan-Feng Lin, and Tung-Shu Yang as inventors. The application discloses a generative AI module that creates exclusive patient-specific electrical stimulation prescriptions by analyzing patient feedback and a database of reference prescriptions, with doctor confirmation before treatment delivery. CPC classifications are G16H 20/30 and G16H 40/63 under health informatics.

Health Assessment Via Music Media, CHIME AI

The USPTO published patent application US20260112480A1 (A1 kind, filed September 10, 2023, application No. 19117046) for CHIME HEALTH AI INC., with inventors David M. GREENBERG and Igor RADOVANOVIC. The application discloses a processor-executed method for assessing user health by obtaining input data, extracting characteristics, determining health scores based on variation from predetermined values, comparing against health condition thresholds, and generating personalized health messages plus health scores for output devices. CPC classification: G16H 40/63 (Health Informatics).

Cardiac Monitoring Device Management Patent Application US20260112481A1

USPTO published patent application US20260112481A1, filed June 28, 2022 (Application No. 19142797), titled 'Systems and Methods for Managing a Cardiac Monitoring Device with a Monitoring Interval Tracker,' naming inventors Richard Todd Butka, Christopher S. Irving, and Patrick Beaulieu. The invention covers a monitoring interval tracker that generates monitoring intervals for multiple patient care modalities including in-office care, remote monitoring, and heart failure care, and that determines whether to extend monitoring intervals, generate dates of service, or create new monitoring intervals based on transmission data and docket report approvals.

AI Avatar Interface for Medical Queries Using Regulatory-Approved Content

The USPTO published patent application US20260111682A1, filed July 30, 2025 (Application No. 19284743), disclosing a method and system for an AI avatar interface that responds to medical queries using regulatory-approved curated content. The system configures a content delivery platform to receive approved content from a content creator and regulatory body, couples the platform to an LLM operating on vector embeddings of tokenized linguistic units, and generates an AI avatar for user interaction. When a user question cannot be answered, the system marks it as unanswerable in a database and sends an indication to both the content creator and the regulatory body.

US20260112463A1 Clinical Data Unified Platform System

USPTO published patent application US20260112463A1 for a unified clinical trial data management system invented by Avik Kumar Pal and Yerramalli Subramaniam. The system comprises a centralized database storing clinical trial data and interconnected modules with automated data linking mechanisms that maintain associations between data entries across modules based on predefined rules. The system enforces data dependencies between modules to ensure integrity, automatically creates placeholders in linked modules upon data entry, prevents certain actions until prerequisites are met, and enables real-time data reconciliation without manual intervention or separate application transfers.

Dual Transcription Generates Health Records From Voice

Acuity Holding Corp. has filed US Patent Application US20260112467A1, published April 23, 2026, covering a dual-transcription AI system that automatically generates health records from vocalized expressions. The system uses two independent transcription models to produce separate transcripts, which are reconciled via a large language model before machine learning determines record data. Desktop automation then populates a record-keeping application based on the extracted data, filed September 19, 2025.

Central Patient Access System for Distributed Electronic Health Record Network

USPTO published patent application US20260111598A1 on April 23, 2026, disclosing a system and method for central patient access to a distributed electronic health record (EHR) network. The system comprises multiple EHR systems connected via secure communication links and a central network node that authenticates patient identities and establishes pre-authorized central patient accounts capable of accessing and modifying patient data across all connected EHR systems. The application (No. 19363561) was filed on October 20, 2025, with six named inventors.

Streamlining Prescription Fulfillment for REMS and Controlled Substances

USPTO published patent application US20260112471A1, filed October 17, 2025, for a prescription fulfillment system that electronically handles documentation for regulated products subject to REMS (Risk Evaluation and Mitigation Strategy) and controlled substance scheduling requirements. The system connects healthcare provider offices with pharmacies via a platform that automatically populates required forms, eliminating paper and redundant documentation while reducing administrative burden. This application names Robert Chuba, Danielle Suzanne Burroughs, Meggan Curta Sullivan, and Joseph V. Fenton as inventors, with CPC classifications in G16H 20/10, G06F 40/174, and G16H 10/60.

Medical Device With Radio Frequency Identification Tear Tag

The USPTO published patent application US20260100258A1 for a medical device with an RFID tear tag system. The application claims a medical device with an RFID circuit and separable electrical component that changes signal output when separated. Inventors are Ashley Rachel Rothenberg and Kevin M. Ryan.

Mental Health Prediction via Conversational Speech Analysis

USPTO published patent application US20260100259A1 by Ellipsis Health, Inc. covering systems and methods for identifying mental health condition severity through passive processing of conversational speech and language. The application discloses language models and acoustic models that process conversation data to generate outputs for mental health assessment, as well as methods for tracking analytics on self-report questionnaires administered during conversations.

Electronic Illuminator System with Fiber and Cap Color Detection

The USPTO published patent application US20260100260A1 for an Electronic Illuminator System, invented by Hans UTZ and Dragan Nebrigic. The system includes a fiber detection assembly, cap color detection assembly, and printed circuit board, configured to illuminate a cable with an effect corresponding to the cable upon authenticating it.

Systems And Methods For Infusion Pump Formulary Validation

USPTO published patent application US20260100261A1 for systems and methods validating infusion pump operations using configuration files to generate operational instructions. The system monitors output signals to determine infusion pump performance across combinations of infusible materials, dosages, rates, and synthesized subjects. The application was filed December 11, 2025, under CPC classifications G16H 20/17 and G16H 40/67.

Domain Extension Learning Device Patent by NEC Corporation

Domain Extension Learning Device Patent by NEC Corporation

ML Smart Array for Healthcare Providers - Patent Application US20260100286A1

USPTO published patent application US20260100286A1 for a machine learning smart array system for healthcare providers. Inventors Praveen Bhat Gurpur and Suchitra Joyce Phillips filed the application covering AI-driven healthcare suggestions using historical provider data and external health information sources. The A1 publication marks the application entering the examination queue.

AI-Powered Ultrasound Scan Review Generator Patent Application

USPTO published patent application US20260100287A1 by Global Ultrasound Institute Inc. on October 4, 2024. The application covers an AI model trained on ultrasound scan datasets that generates automated assessment facts and written feedback for ultrasound reviews, with multi-language support and configurable per-exam review interfaces.

HEALTHCARE PROVIDER ASSISTANT SYSTEM VOICE AI PATENT

USPTO published patent application US20260100288A1 on April 9, 2026, covering a healthcare provider assistant system using voice AI. The system receives voice data from a microphone, distinguishes healthcare provider from patient voices, generates transcriptions, and processes them through a trained language model to complete tasks and generate documentation from healthcare encounters.