# US 2026/0294330 A1 · Predictive Neurological Monitoring and Clinical Decision Support System

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0294330 A1
- Document ID: US20260294330A1
- Status: Published application
- Kind: A1
- Application number: 19/572,624
- Filed: 2026-03-19
- Published: 2026-10-01
- Earliest claimed: 2025-04-01
- Assignees: Pahk Neurology Inc.
- Applicants: Pahk Neurology Inc.
- Inventors: Eugene Pahk; Kristians Veters; Artjoms Simutins
- Web page: https://patentgrep.com/patent/US20260294330A1

## CPC

- A61B 5/4088: {Diagnosing of monitoring cognitive diseases, e.g. Alzheimer, prion diseases or dementia}
- A61B 5/0022: {Monitoring a patient using a global network, e.g. telephone networks, internet}
- A61B 5/1101: {Detecting tremor}
- A61B 5/162: {Testing reaction times}
- A61B 5/6898: {Portable consumer electronic devices, e.g. music players, telephones, tablet computers}
- A61B 5/7267: {involving training the classification device}
- A61B 5/7275: {Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor}
- A61B 5/7435: {Displaying user selection data, e.g. icons in a graphical user interface}
- A61B 5/746: {Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms}
- A61B 5/7475: {User input or interface means, e.g. keyboard, pointing device, joystick}
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- A61B 2562/0219: Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

A system and method for predictive neurological assessment and clinical decision support utilizing mobile device sensor fusion. A mobile computing device presents digitized cognitive testing modules and simultaneously captures both foreground interaction data and background biometric sensor data, such as micro-oscillations, haptic pressure, and millisecond-level interaction latencies. The captured data is transmitted to a remote server where a trained machine learning model processes the background biometric data to detect sub-perceptual biometric anomalies. By comparing these anomalies against historical baseline patterns and supervised clinical diagnosis flags, the system generates a predictive trajectory vector to forecast the onset of a neurological condition prior to a macroscopic decline in foreground test scores. The system outputs early-warning alerts, temporal prognoses, and treatment efficacy tracking to a secure clinician portal to facilitate predictive medical interventions.

## Claims

### Claim 1

A method for predictive neurological assessment and clinical decision support via mobile device sensor fusion, the method comprising:
rendering a digitized cognitive testing module on a display screen of a mobile computing device;
capturing foreground interaction data via the mobile computing device as a user interacts with the digitized cognitive testing module;
simultaneously capturing background biometric sensor data while the user interacts with the digitized cognitive testing module, wherein capturing the background biometric sensor data comprises:
polling continuous multi-axis motion data utilizing an accelerometer and a gyroscope of the mobile computing device to extract micro-oscillation and device sway metrics;
logging spatial coordinate data and haptic pressure data utilizing a capacitive touch digitizer of the mobile computing device; and
timestamping individual screen interactions utilizing a native operating system event logger of the mobile computing device to extract millisecond-level inter-tap latency and motor planning metrics;
transmitting the foreground interaction data and the background biometric sensor data to a remote server;
calculating a foreground task completion score based on the captured foreground interaction data on the remote server;
processing the background biometric sensor data via a trained machine learning model executing on the remote server to detect a sub-perceptual biometric anomaly, wherein the machine learning model is trained on a historical dataset comprising background biometric sensor data from a plurality of patients correlated with supervised clinical diagnosis flags;
generating a predictive trajectory vector indicating a future onset of a neurological condition based on the detected sub-perceptual biometric anomaly prior to a macroscopic decline in the user's foreground task completion score; and
outputting an alert and a corresponding specific treatment plan recommendation to a secure clinician portal based on the generated predictive trajectory vector.

### Claim 2 (depends on 1)

The method of claim 1, wherein training the machine learning model comprises:
receiving a supervised clinical diagnosis flag associated with a specific patient of the plurality of patients, the clinical diagnosis flag establishing a date of formal neurological diagnosis; and
executing a retrospective analysis on the historical dataset of the specific patient to identify an early-warning biometric signature occurring in the background biometric sensor data prior to the date of formal neurological diagnosis.

### Claim 3 (depends on 1)

The method of claim 1, wherein the background biometric sensor data is captured continuously by the mobile computing device during non-clinical, gamified application usage independently of the digitized cognitive testing module.

### Claim 4 (depends on 1)

The method of claim 1, further comprising a method for predictive treatment efficacy tracking, comprising:
receiving a treatment initiation flag via the clinician portal indicating the recommended specific treatment plan has been prescribed to the user;
continuously receiving subsequent foreground interaction data and subsequent background biometric sensor data from the user's mobile computing device following the treatment initiation flag;
receiving a positive outcome flag via the clinician portal indicating a clinical improvement in the user;
executing a second retrospective analysis via the machine learning model on the subsequent background biometric sensor data captured between the treatment initiation flag and the positive outcome flag;
identifying a physiological stabilization pattern in the subsequent background biometric sensor data indicating treatment efficacy; and
applying the identified physiological stabilization pattern to actively treated patients to dynamically predict the efficacy of a prescribed treatment prior to an improvement in the foreground task completion score.

### Claim 5 (depends on 1)

The method of claim 1, wherein generating the predictive trajectory vector further comprises:
calculating a specific temporal prognosis indicating a projected timeframe before the onset of severe neurological degradation based on the historical dataset.

### Claim 6 (depends on 1)

The method of claim 1, wherein the machine learning model is further trained by:
receiving a normal health flag via the clinician portal indicating a baseline evaluation period of a predefined duration wherein a patient demonstrates stable health; and
modeling a baseline biometric signature of a normal brain based on the background biometric sensor data associated with the normal health flag to increase the machine learning model's accuracy in contrasting healthy users against neurologically compromised users.

### Claim 7 (depends on 1)

The method of claim 1, wherein the sub-perceptual biometric anomaly comprises an increase in the millisecond-level inter-tap latency occurring simultaneously with a targeted subset of the spatial coordinate data indicating hemispatial interaction neglect.

### Claim 8 (depends on 1)

The method of claim 1, wherein processing the background biometric sensor data further comprises:
combining the captured foreground interaction data with the background biometric sensor data to form a synergistic multidimensional dataset; and
wherein the trained machine learning model processes the synergistic multidimensional dataset to detect the sub-perceptual biometric anomaly and generate the predictive trajectory vector.

### Claim 9 (depends on 4)

The method of claim 4, wherein executing the second retrospective analysis comprises:
initiating an algorithmic review of the subsequent background biometric sensor data starting chronologically at a date associated with the treatment initiation flag and advancing forward in time until the machine learning model identifies the physiological stabilization pattern, thereby determining a minimum time threshold required to identify treatment efficacy prior to the receipt of the positive outcome flag.

### Claim 10 (depends on 2)

The method of claim 2, wherein executing the retrospective analysis on the historical dataset of the specific patient comprises:
initiating an algorithmic review of the historical dataset starting chronologically from an earliest recorded baseline session of the specific patient and advancing forward in time until the machine learning model recognizes a pattern constituting the early-warning biometric signature.

### Claim 11

A system for predictive neurological monitoring and clinical decision support via mobile device sensor fusion, the system comprising:
a mobile computing device comprising a display screen, a capacitive touch digitizer, a multi-axis accelerometer, a gyroscope, a native operating system event logger, and a wireless transceiver;
a remote cloud-based server in communication with the mobile computing device; and
a secure clinician portal in communication with the remote cloud-based server;
wherein the mobile computing device is configured to render a digitized cognitive testing module on the display screen, capture foreground interaction data as a user interacts with the digitized cognitive testing module, simultaneously capture background biometric sensor data, and transmit the foreground interaction data and the background biometric sensor data to the remote cloud-based server;
wherein capturing the background biometric sensor data comprises polling continuous multi-axis motion data utilizing the accelerometer and the gyroscope to extract micro-oscillation and device sway metrics, logging spatial coordinate data and haptic pressure data utilizing the capacitive touch digitizer, and timestamping individual screen interactions utilizing the native operating system event logger to extract millisecond-level inter-tap latency and motor planning metrics;
wherein the remote cloud-based server comprises a machine learning architecture configured to calculate a foreground task completion score based on the foreground interaction data, and process the background biometric sensor data via a trained machine learning model to detect a sub-perceptual biometric anomaly;
wherein the machine learning model is trained on a historical dataset comprising background biometric sensor data from a plurality of patients correlated with supervised clinical diagnosis flags;
wherein the remote cloud-based server is further configured to generate a predictive trajectory vector indicating a future onset of a neurological condition based on the detected sub-perceptual biometric anomaly prior to a macroscopic decline in the user's foreground task completion score; and
wherein the secure clinician portal is configured to output an alert and a corresponding specific treatment plan recommendation based on the generated predictive trajectory vector.

### Claim 12 (depends on 11)

The system of claim 11,
wherein the secure clinician portal is configured to receive a supervised clinical diagnosis flag establishing a date of formal neurological diagnosis for a specific patient; and
wherein the remote cloud-based server is configured to train the machine learning model by executing a retrospective analysis on the historical dataset of the specific patient, the retrospective analysis comprising initiating an algorithmic review of the historical dataset starting chronologically from an earliest recorded baseline session of the specific patient and advancing forward in time until the machine learning model recognizes a pattern constituting an early-warning biometric signature occurring prior to the date of formal neurological diagnosis.

### Claim 13 (depends on 11)

The system of claim 11,
wherein the secure clinician portal is further configured to receive a treatment initiation flag indicating the specific treatment plan has been prescribed, and subsequently receive a positive outcome flag indicating a clinical improvement in the user; and
wherein the remote cloud-based server is configured to execute a second retrospective analysis on subsequent background biometric sensor data continuously received following the treatment initiation flag, the second retrospective analysis comprising initiating an algorithmic review starting chronologically at a date associated with the treatment initiation flag and advancing forward in time until the machine learning model identifies a physiological stabilization pattern indicating treatment efficacy prior to an improvement in the foreground task completion score.

### Claim 14 (depends on 11)

The system of claim 11,
wherein the remote cloud-based server is configured to combine the captured foreground interaction data with the background biometric sensor data to form a synergistic multidimensional dataset; and
wherein the trained machine learning model processes the synergistic multidimensional dataset to detect the sub-perceptual biometric anomaly and generate the predictive trajectory vector.

### Claim 15 (depends on 11)

The system of claim 11,
wherein the secure clinician portal is configured to receive a normal health flag indicating a baseline evaluation period of a predefined duration wherein a patient demonstrates stable health; and
wherein the machine learning architecture of the remote cloud-based server is configured to model a baseline biometric signature of a normal brain based on the background biometric sensor data associated with the normal health flag to increase the machine learning model's accuracy in contrasting healthy users against neurologically compromised users.

### Claim 16 (depends on 11)

The system of claim 11, wherein the remote cloud-based server is configured to generate the predictive trajectory vector by calculating a specific temporal prognosis indicating a projected timeframe before the onset of severe neurological degradation based on the historical dataset.

### Claim 17 (depends on 11)

The system of claim 11, wherein the mobile computing device is configured to continuously capture the background biometric sensor data during non-clinical, gamified application usage independently of the active rendering of the digitized cognitive testing module.

### Claim 18 (depends on 11)

The system of claim 11, wherein the sub-perceptual biometric anomaly detected by the remote cloud-based server comprises an increase in the millisecond-level inter-tap latency occurring simultaneously with a targeted subset of the spatial coordinate data indicating hemispatial interaction neglect.
