System and method for tremor detection

Abstract

A system and method for diagnosing a medical condition are provided. The system includes a smart pen having a magnetoelastic tip, ferrofluid, and a coil for converting handwriting movements into electrical signals. The system further includes a processor configured to analyze the electrical signals to distinguish handwriting patterns of the medical condition from handwriting patterns without the medical condition.

Description

SYSTEM AND METHOD FOR TREMOR DETECTIONCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 789,027, filed April 15, 2025, which is hereby incorporated by reference in its entirety.GOVERNMENT FUNDING STATEMENT

[0002] This invention was made with government support under CA287326 awarded by the National Institutes of Health, and N00014-24-1-2065 awarded by the Office of Naval Research. The government has certain rights in the invention.BACKGROUND

[0003] Parkinson’s disease (PD) is one of the fastest-growing neurodegenerative diseases and affects approximately 10 million people worldwide. Symptoms of PD, including tremors, rigidity, and bradykinesia, can significantly impair mobility and reduce quality of life. Because PD is progressive and chronic, timely and accessible diagnosis is important for improving patient outcomes and enabling earlier intervention. However, conventional approaches for diagnosing and monitoring PD progression remain limited. Existing clinical procedures often rely on observation of motor symptoms, such as tremors and rigidity, by trained clinicians. These approaches are highly user-dependent, inefficient, and lack objective quantitative standards. In addition, access to subspecialty resources for PD diagnosis is limited, particularly in low-income regions where the number of neurologists may be as low as 0.03-0.13 per 100,000 population. These constraints make conventional PD evaluation difficult to scale for large populations and resource-limited settings.

[0004] Recent advances have identified biological changes that may occur before the onset of motor symptoms, including the presence or absence of pathological a-synuclein in tissues or body fluids, cerebrospinal fluid levels of L-3,4-dihydroxyphenylalanine decarboxylase (DDC), and morphological abnormalities in the retina. While such biomarkers may support earlier detection, identification of these biomarkers typically requires sophisticated instrumentation, centralized healthcare infrastructure, and specialized expertise. As a result, these approaches are often costly,difficult to access, and impractical for broad screening or routine monitoring outside specialized clinical facilities.

[0005] Thus, there is an ongoing need for improved systems and methods for PD diagnosis and monitoring that are quantitative, low-cost, and / or accessible, and that can be deployed across large populations, including in non-clinical and resource-limited environments.SUMMARY

[0006] The present disclosure overcomes the aforementioned drawbacks by providing systems and methods for personalized Parkinson’s disease (PD) detection. In one non-limiting example, a smart pen is provided that may include a magnetoelastic tip, a ferrofluid ink reservoir, and a coil disposed around at least a portion of a pen body. During use, motion of the smart pen and pressure applied during handwriting can cause deformation of the magnetoelastic tip and dynamic movement of the ferrofluid ink, thereby generating variations in magnetic flux that are converted by the coil into electrical signals representative of handwriting activity. In some examples, the smart pen is configured to capture both on-surface handwriting motions and in-air handwriting motions, thereby enabling detection of multiple motor characteristics associated with PD, including tremor-related, bradykinesia-related, and pressure-related abnormalities. Unlike conventional handwriting analysis tools and inertial sensing devices that primarily measure trajectory or acceleration alone, the disclosed smart pen combines a deformable magnetoelastic structure with magnetized ferrofluid ink to transduce both pressure-dependent and motiondependent handwriting events in a self-powered manner, without reliance on an external power source. Accordingly, the disclosed systems and methods provide a quantitative, low-cost, portable, and accessible platform for objective PD-related motor assessment, with potential utility in large populations, home settings, and resource-limited environments.

[0007] In accordance with one aspect of the present disclosure, a diagnostic writing utensil for determining a presence or quantifying symptoms of a medical disorder is provided that may include a barrel forming the diagnostic writing utensil. The diagnostic writing utensil may further include a magnetoelastic tip coupled to the barrel and having a shape that changes during use of the writing utensil, and wherein the magnetoelastic tip may generate a magnetic field configured to extend along the barrel that may vary with changes in the shape of the magnetoelastic tip. The diagnostic writing utensil may further include a ferrofluid arranged in the barrel to be subjected tothe magnetic field of the magnetoelastic tip. The diagnostic writing utensil may further include a coil extending around the ferrofluid, wherein changes in the shape of the magnetoelastic tip may cause changes in the magnetic field generated by the magnetoelastic tip that may induce motion of the ferrofluid, and wherein motion of the ferrofluid may induce a current in the coil that is indicative of a presence or quantification of symptoms of a medical disorder.

[0008] According to another aspect of the disclosure, a writing utensil is provided that may include a barrel forming the diagnostic writing utensil. The writing utensil may further include a magnetoelastic tip coupled to the barrel and having a shape that changes during use of the writing utensil, and wherein the magnetoelastic tip may generate a magnetic field configured to extend along the barrel that may vary with changes in the shape of the magnetoelastic tip. The writing utensil may further include a ferrofluid arranged in the barrel to be subjected to the magnetic field of the magnetoelastic tip. The diagnostic writing utensil may further include a coil extending around the ferrofluid, and wherein changes in the shape of the magnetoelastic tip may cause changes in the magnetic field generated by the magnetoelastic tip that may induce motion of the ferrofluid and wherein motion of the ferrofluid may induce a current in the coil.

[0009] According to yet another aspect of the present disclosure, a method is provided. The method may include providing a writing utensil having a magnetoelastic tip, ferrofluid, and a coil. The method may further include converting handwriting movements using the writing utensil into magnetic flux variations using the magnetoelastic tip and ferrofluid. The method may further include generating electrical signals from the magnetic flux variations using the coil.

[0010] According to another aspect of the disclosure, a system is provided that may include a smart pen having a magnetoelastic tip, ferrofluid, and a coil for converting handwriting movements into electrical signals. The system may further include a processor configured to analyze the electrical signals to distinguish handwriting patterns of the medical condition from handwriting patterns without the medical condition.BRIEF DESCRIPTION OF THE DRAWINGS

[0011] FIG. l is a schematic diagram of a diagnostic system, in accordance with some aspects of the disclosure.

[0012] FIG. 2 is a flow chart of the operation of the diagnostic system, in accordance with some aspects of the disclosure.

[0013] FIG. 3 is a schematic drawing of a diagnostic writing utensil, in accordance with some aspects of the disclosure.

[0014] FIG. 4 is a schematic illustration of a magnetoelastic tip of the diagnostic writing utensil, in accordance with some aspects of the disclosure.

[0015] FIG. 5 illustrates cross-sectionals view of the diagnostic writing utensil, in accordance with some aspects of the disclosure.

[0016] FIG. 6 is an illustration of a reservoir of the diagnostic writing utensil, in accordance with some aspects of the disclosure.

[0017] FIG. 7 illustrates electrical outputs of the diagnostic writing utensil, in accordance with some aspects of the disclosure.

[0018] FIG. 8 illustrates the diagnostic writing utensil in a clinical-use context, in accordance with some aspects of the disclosure.

[0019] FIG. 9 illustrates analysis workflow of handwriting-derived signals acquired by the diagnostic writing utensil, in accordance with some aspects of the disclosure.DETAILED DESCRIPTION

[0020] Effectively diagnosing Parkinson’ s Disease (PD) promptly and accessibly is crucial for improving patient outcomes, yet reaching this goal has been a significant challenge. The present disclosure provides systems and methods for detecting, quantifying, and / or monitoring motor neurological symptoms through handwriting-based analysis. Handwriting is a complex process that intertwines cognitive, perceptual, and fine motor abilities, all of which are substantially affected by PD. When motor symptoms affect the dominant hand, analyzing handwriting patterns can offer crucial insights and quantitative biometric markers for diagnosing and assessing the disease. Although conventional handwriting analysis tools like digital tablets / screens are utilized, they typically specialize in tracking on-surface handwriting trajectories, often overlooking the motor symptoms that manifest during writing. Moreover, their bulkiness, high cost, and complicated setups not only limit these traditional tools when applied within clinical environments but also hinder continuous monitoring in non-clinical settings.

[0021] The present disclosure provides a smart pen that may be self-powered and have a configuration capable of converting handwriting motions into analyzable electrical signals. In one non-limiting example, the smart pen or other hand-held device can include a magnetoelastic tip, aferrofluid material, and an electrical sensing structure, such as a coil positioned around or proximate to at least a portion of a pen body (e.g., a barrel). During operation, handwriting-related motion and / or applied pressure may cause deformation of the magnetoelastic tip and dynamic movement of the ferrofluid material. These mechanical actions can produce corresponding variations in magnetic flux, which can be detected by the electrical sensing structure and converted into voltage and / or current signals representative of the user’ s handwriting activity. In this manner, the disclosed systems and methods can enable high-fidelity capture of both on-surface handwriting movements and in-air handwriting movements in a self-powered manner, without requiring an external energy source to generate the sensing signal. As a result, the disclosed smart pen configuration can detect motor abnormalities that may not be adequately captured by systems that measure only trajectory, only acceleration, or only final handwriting output.

[0022] In some examples, the systems and methods described herein are configured for use in connection with PD, although the disclosed technology may also be applicable to other neurological or movement-related conditions associated with multiple sclerosis (MS), bradykinesia, impaired fine motor control, tremors, other neurodegenerative diseases that may include tremors or the like, or other handwriting-related abnormalities. The present disclosure addresses a significant need for prompt, accessible, and quantitative assessment of motor symptoms, particularly in contexts where conventional clinical evaluation may be subjective, infrequent, costly, or difficult to access.

[0023] The present disclosure therefore provides a practical and beneficial platform for objective motor assessment. In some examples, the disclosed systems and methods may be used for early-stage detection of neurological conditions, for ongoing monitoring of symptom progression, and / or for repeated measurements outside of a conventional clinical setting. This may be particularly beneficial because symptom severity in many neurological conditions can fluctuate over time, including between clinical visits, and such fluctuations may not be readily captured through episodic in-clinic observation alone. By providing a low-cost, portable, and quantitative sensing platform that can be used in home settings and resource-limited environments, the present disclosure supports broader access to neurological monitoring and more effective longitudinal tracking of disease-related motor changes. By enabling quantitative conversion of handwriting activity into compact analyzable signals, the present disclosure provides a sensing platform that isnot only privacy-conscious and computationally efficient, but also suitable for repeated and scalable use across broad populations.

[0024] While Parkinson’ s disease is used throughout the disclosure as a non-limiting example, the disclosed systems and methods are not limited to PD and may be used in connection with a variety of neurological, neuromuscular, or movement-related disorders in which fine motor symptoms are present. Accordingly, references to PD herein are provided for explanatory purposes and should not be construed as limiting the scope of the present disclosure.

[0025] Referring to FIG. 1, a system 100 is shown for detecting, quantifying, and / or monitoring handwriting-associated motor symptoms of a subject. In some examples, the system 100 may include a hand-held device or smart pen 10, a signal module 12, a handwriting input or handwriting activity 14, one or more processors 16, and a user interface 18. In other configurations, one or more of these components may be combined, omitted, distributed, or implemented in different form factors.

[0026] In some configurations, the pen 10 may include a magnetoelastic tip, a ferrofluid material, and an electrical sensing structure, such as a coil disposed around or proximate to at least a portion of a pen body such as a barrel, as described in greater detail below. The pen 10 may be configured to operate as both a writing instrument and a self-powered motion transducer that can convert handwriting-related mechanical inputs into electrical signals. In some examples, such mechanical inputs may include on-surface writing motions, in-air writing motions, pressure variations, acceleration, rotation, and / or other fine motor movements associated with handwriting.

[0027] In the illustrated non-limiting configuration, the handwriting input 14 may represent one or more writing tasks, drawing tasks, freeform handwriting activities, or other hand-executed motion sequences performed using the pen 10. In some examples, the handwriting input 14 may include writing on a surface, drawing geometric or repeating patterns, writing letters or symbols, and / or performing in-air gestures without contact with a writing surface. The handwriting input 14 may generate motion and pressure conditions that interact with the pen 10 to produce one or more analyzable electrical signals.

[0028] In some configurations, the signal module 12 may be operatively associated with the pen 10 and may be configured to acquire, receive, condition, amplify, digitize, filter, store, and / or transmit one or more handwriting-derived electrical signals generated by the pen 10. In one nonlimiting example, the signal module 12 may include one or more analog front-end circuits, suchas one or more amplifiers, filters, analog-to-digital converters, and / or signal conditioning circuits. In other examples, at least a portion of the signal module 12 may be integrated within the pen 10, the processor 16, or another computing device.

[0029] In some examples, operation of the pen 10 may be based on a dual transduction mechanism. For example, pressure applied during handwriting may deform the magnetoelastic tip and thereby alter magnetic flux through a magnetoelastic effect, while motion of magnetized ferrofluid material within the pen 10 may generate additional time-varying magnetic flux associated with handwriting movement. A coil or other electrical sensing structure associated with the pen 10 may convert such magnetic flux changes into voltage and / or current signals representative of the handwriting activity 14. In this manner, the pen 10 may capture both pressuredependent and motion-dependent characteristics of handwriting in a self-powered manner, without requiring an external power source for signal generation.

[0030] In some configurations, the signal module 12 may communicate one or more acquired signals to the processor 16 for analysis. Communication between the signal module 12 and the processor 16 may be wired and / or wireless, including by way of non-limiting example Bluetooth, near-field communication, Wi-Fi, cable-based communication, or other communication pathways. In some examples, the processor 16 may be implemented within the signal module 12, within the pen 10, within a mobile device, within a desktop or laptop computer, within a wearable device, and / or within a remote server or cloud-based computing system.

[0031] The processor 16 may be configured to determine one or more handwriting-related parameters based on the signals acquired from the pen 10. In non-limiting examples, such parameters may include one or more features associated with tremor, writing time, cycle duration, pressure variation, acceleration, frequency content, signal amplitude, peak-to-peak timing, and / or other kinematic or motor characteristics. In some configurations, the processor 16 may compare acquired data to one or more baseline data sets, reference patterns, population data sets, and / or subject-specific longitudinal data to identify changes indicative of motor abnormalities, disease progression, or a presence or quantification of symptoms of a medical disorder.

[0032] In some configurations, the processor 16 may preprocess the signals before analysis. By way of non-limiting example, preprocessing may include amplification, denoising, filtering, normalization, artifact reduction, segmentation, frequency-domain transformation, time-frequency analysis, and / or extraction of one or more engineered or learned features. In some examples, theprocessor 16 may convert a one-dimensional handwriting-derived signal into a two-dimensional representation, such as a time-frequency scalogram or other transformed image representation, for subsequent machine-learning analysis.

[0033] In some examples, the processor 16 may analyze the handwriting-derived signals using one or more computational models, including artificial intelligence and / or machine-learning models. Non-limiting examples may include one-dimensional convolutional neural networks, recurrent neural networks, decision-tree-based models, logistic regression models, and / or other supervised or unsupervised classifiers. In some configurations, such models may be trained or updated using subject-specific data, longitudinal monitoring data, and / or population-scale training data.

[0034] In some configurations, the processor 16 may be configured to classify or otherwise assess neurological status based on the handwriting-derived signals. In one non-limiting example, the processor 16 may distinguish between healthy subjects and subjects having PD. In other examples, the processor 16 may detect, quantify, and / or longitudinally track one or more motor symptoms associated with PD or other neurological conditions, including but not limited to tremors, bradykinesia, impaired motor planning, segmented movement patterns, abnormal pressure control, and / or other handwriting-related abnormalities.

[0035] The user interface 18 may be configured to present information associated with the analysis to a user, such as a subject, caregiver, clinician, researcher, or technician. In some examples, the user interface 18 may include a display, a mobile application interface, a web-based interface, a terminal, a clinical workstation, a wearable device interface, and / or other output device. The user interface 18 may present one or more raw signals, processed signals, extracted handwriting parameters, trend data, classification outputs, confidence values, status indicators, and / or alerts associated with the monitored condition.

[0036] In some examples, the system 100 may support repeated, intermittent, or continuous monitoring over time, including use outside of a conventional clinical environment. Because the pen 10 may operate in a self-powered manner and may generate compact time-series signals that are suitable for efficient storage and computational analysis, the system 100 may provide a low-cost, portable, privacy-conscious, and widely disseminable platform for early detection, ongoing monitoring, and longitudinal tracking of neurological motor symptoms.

[0037] Various components of the control circuitry may be implemented on one or more processors 16. The processor 16 may be configured to send and / or receive information (e.g., including instructions, data, values, signals, or the like) to / from the various components of the system 100 and / or external components. The processor 16 may include, for example, a processor suitable equivalent, a DSP, CPU, APU, GPU, microcontroller, application-specific integrated circuit, programmable gate array, and the like, any other digital and / or analog components, as well as combinations of the foregoing (whether distributed, networked, locally connected, or the like), and may further comprise inputs and outputs for receiving and providing control instructions, control signals, drive signals, power signals, sensor signals (e.g., current or voltage sensor output, image sensor output, volumetric scanner output, and the like), digital signals, analog signals, and the like. All such computing devices and environments are intended to fall within the meaning of the terms “controller,” “control unit,” “processor,” “processor unit,” “processing device,” or “processing circuitry” as used herein unless a different meaning is explicitly provided or otherwise clear from the context. In some examples, the processor 16 may be located outside of the server network.

[0038] The processor 16 may include processing circuitry configured to execute operating routine(s) stored in a memory. The memory may include any suitable volatile memory, nonvolatile memory, storage, any other suitable type of storage medium, or any suitable combination thereof. For example, the memory may include RAM, ROM, EEPROM, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical drives, etc. In some embodiments, the memory may have encoded thereon a computer program for controlling operation of the processor 16, the pen 10, or any other component of the control circuitry. In some configurations, the various components of the system 100 may be implemented as software, as hardware, or any suitable combination thereof. In some embodiments, the operating routine(s) may comprise firmware.

[0039] In some configurations, the processor 16 may be communicatively coupled with the user interface 18, either directly or indirectly. As such, the user interface 18 may be considered part of the control circuitry, or, in some configurations, may be considered a separate component outside of the control circuitry. In some examples, user interface 18 may receive PD indication data from the pen 10 and PD status to an external monitoring system.

[0040] In some examples, the user interface 18 may include a housing that contains a display (e g., for showing device identification information and / or PD conditions), an LED indicator (e.g., to provide a visual status indication of PD status, or other information), and an input connector for receiving signals from the pen 10 and / or the processor 16. In some embodiments, the user interface 18 may include internal circuitry housed within the housing, such as a dedicated processor and memory components. The processor may execute program instructions stored in the memory to analyze sensor signals received from the pen 10, enabling the pen 10 to detect the PD conditions. However, in other embodiments, such functionality may be executed by the processor 16. Furthermore, in some embodiments, the user interface 18 can establish communication with external monitoring systems through various protocols including CANbus, Modbus, BACNet IP, ethernet, and / or other wired or wireless connections, enabling integration with management systems and providing indicators or digital alarm outputs when PD conditions are detected.

[0041] Referring to FIG. 2, a method 200 is shown for converting handwriting activity into electrical signals that may be used to detect, determine a presence of, quantify, and / or monitor symptoms of a medical disorder. In some examples, all or part of the method 200 may be performed by the system 100 of FIG. 1, including the pen 10, the signal module 12, the processor 16, and / or the user interface 18. In other configurations, the method 200 may be performed by other systems or devices configured to convert handwriting-related movement into magnetic flux variations and corresponding electrical signals.

[0042] In block 202, a writing utensil having a magnetoelastic tip, ferrofluid, and a coil may be provided. In one non-limiting example, the writing utensil may include a barrel forming the writing utensil, a magnetoelastic tip coupled to the barrel and may have a shape that changes during the use of writing utensil, a ferrofluid arranged within the barrel, and a coil extending around the ferrofluid. In some configurations, the writing utensil may further include a replaceable reservoir configured to house the ferrofluid, and the replaceable reservoir may include a cap configured to reduce and / or prevent leakage and evaporation. In some examples, the magnetoelastic tip may include magnetic particles embedded in a deformable matrix, such as a silicone matrix, and the ferrofluid may include magnetic nanoparticles dispersed in a carrier solution.

[0043] In some configurations, the magnetoelastic tip may be configured to generate a magnetic field that extends generally along at least a portion of the barrel and may vary with changes in the shape of the magnetoelastic tip or otherwise acts on the ferrofluid during use of thewriting utensil. During handwriting, the writing utensil may be used on a writing surface and / or in air, such that the utensil experiences one or more handwriting movements including pressure, acceleration, translation, rotation, tilting, stroke transitions, and / or other fine motor actions. These handwriting movements may be associated with normal motor activity and / or with symptoms of a medical disorder, including but not limited to tremor-related symptoms, bradykinesia-related symptoms, impaired pressure control, impaired motor planning, and / or segmented motion patterns.

[0044] In block 204, handwriting movements using the writing utensil may be converted into magnetic flux variations using the magnetoelastic tip and ferrofluid. In some examples, converting handwriting movements into magnetic flux variations may include deforming the magnetoelastic tip in response to handwriting movements and moving the ferrofluid in response to handwriting movements. For example, when the magnetoelastic tip contacts a writing surface and experiences writing pressure, the shape of the magnetoelastic tip may change. This deformation may alter the orientation, spacing, and / or relative position of magnetic particles within the magnetoelastic tip, thereby causing a corresponding change in the magnetic field and magnetic flux associated with the tip. In some examples, the resulting magnetic flux variation may be coupled into the ferrofluid, which is subjected to the magnetic field of the magnetoelastic tip.

[0045] In some configurations, movement of the writing utensil may also induce motion of the ferrofluid, including during on-surface handwriting movements and / or in-air handwriting movements. For example, acceleration, direction changes, stroke transitions, and / or rotational movement of the writing utensil may induce displacement, agitation, redistribution, or other dynamic movement of the ferrofluid within the barrel and / or reservoir. The magnetoelastic-tip deformation and ferrofluid motion may each contribute to time-varying magnetic flux, such that the writing utensil operates using a combined transduction mechanism that captures both pressuredependent and motion-dependent characteristics of handwriting.

[0046] In block 206, electrical signals may be generated from the magnetic flux variations using the coil. In one non-limiting example, motion of the ferrofluid relative to the coil and / or magnetic flux changes associated with deformation of the magnetoelastic tip may induce a current and / or voltage in the coil. The generated electrical signals may be representative of the handwriting movements and may encode one or more features associated with handwriting pressure, writing speed, acceleration, tremor, frequency content, signal amplitude, timing, and / or other motor characteristics. In some configurations, the writing utensil may generate such electrical signals ina self-powered manner, such that the act of handwriting itself provides the mechanical input used to generate the sensing signal.

[0047] In some examples, the electrical signals generated in block 206 may be acquired, conditioned, and / or analyzed to detect symptoms associated with a medical disorder. For example, the electrical signals may be amplified, filtered, digitized, segmented, normalized, transformed, or otherwise preprocessed prior to analysis. One or more features may then be extracted from the signals, including by way of non-limiting example peak amplitude, peak-to-peak timing, writing duration, cycle duration, pressure-correlated signal features, acceleration-correlated signal features, and / or frequency-domain features. In some configurations, a processor may compare such features to one or more reference data sets, baseline data, subject-specific data, and / or longitudinal monitoring data to identify changes indicative of a medical disorder or changes in symptom severity over time. In some configurations, the electrical signals may be analyzed by delivering the electrical signals, or one or more representations derived therefrom, into a machinelearning model trained to classify a medical disorder or disease. In some examples, the machinelearning model may include a one-dimensional convolutional neural network, although other models may additionally or alternatively be used. In one non-limiting example, the machinelearning model may be trained to distinguish handwriting patterns associated with a neurodegenerative disease or disorder, or another medical condition that causes tremors or fine motor impairment, from handwriting patterns not associated with the medical condition. In some examples, the analysis may be used to determine a presence of the medical condition, quantify symptom severity, distinguish between subject populations, and / or longitudinally track symptom progression.

[0048] In some examples, the method 200 may further include performing a calibration procedure to account for variations in handwriting. For example, a user-specific calibration may be performed prior to use of the writing utensil, during initial use, and / or at one or more later times, such that signal amplitude, writing angle, writing style, holding angle, stroke intensity, or other individualized writing characteristics can be accounted for in subsequent analysis. In some configurations, the calibration procedure may include instructing a user to perform one or more predefined writing tasks, comparing resulting signals to one or more stored reference signals, and determining one or more correction factors or normalization parameters for later use.

[0049] After block 206, and optionally after one or more analysis and / or calibration operations, the method 200 may return to block 204 such that additional handwriting movements can be converted into additional magnetic flux variations and corresponding electrical signals. This iterative process may support repeated, intermittent, continuous, and / or longitudinal monitoring over time.

[0050] Referring to FIG. 3, the system 100 is shown as a smart pen 10 configured to convert handwriting activity into electrical signals that may be used for detecting, determining a presence of, quantifying, and / or monitoring symptoms of a medical disorder. In the illustrated non-limiting configuration, the smart pen 10 may include a magnetoelastic tip 104, a barrel 108, a coil 112, a ferrofluid 116, a reservoir 120, a cap 124, and a grip 128. In some examples, the smart pen 10 may generate one or more handwriting-derived signal features 132 that can be analyzed by the system 100, as discussed above with respect to FIGS. 1 and 2.

[0051] In some configurations, the magnetoelastic tip 104 may be coupled to a first end of the barrel 108 and may be configured to engage a writing surface during on-surface handwriting. In one non-limiting example, the magnetoelastic tip 104 may include an internal threaded structure configured to engage a corresponding external threaded structure of the barrel 108, such that the magnetoelastic tip 104 may be removably secured to the barrel 108 by twisting engagement. In other configurations, the magnetoelastic tip 104 may be coupled to the barrel 108 by press fit, adhesive bonding, snap fit, overmolding, mechanical interlock, or any other suitable coupling arrangement. The magnetoelastic tip 104 may be formed from a deformable magnetoelastic composite configured to change shape during writing and thereby change a magnetic field generated by the magnetoelastic tip 104.

[0052] In some examples, the magnetoelastic tip 104 may include magnetic particles embedded in a polymeric matrix. In one non-limiting example, the magnetic particles may be neodymium-iron-boron magnetic particles disposed in a silicone matrix. In another non-limiting example, the silicone matrix may include a soft elastomeric material, such as a silicone-based elastomer, such that the magnetoelastic tip 104 may deform under handwriting pressure while maintaining magnetic functionality. In other configurations, the magnetic particles may include iron oxide magnetic particles, ferrite particles, coated magnetic particles, or other magnetically responsive particulates, and the matrix may include silicone, elastomer, viscoelastic polymer, biocompatible polymer, or other deformable materials. In some examples, after fabrication, themagnetoelastic tip 104 may be magnetized, such as by impulse magnetization, to establish a remanent magnetic field extending generally along the barrel 108 and into the region containing the ferrofluid 116.

[0053] The barrel 108 may be coupled to the magnetoelastic tip 104 and may support one or more additional components of the smart pen 10. In the illustrated non-limiting configuration, the barrel 108 may extend longitudinally between the magnetoelastic tip 104 and the cap 124 and may define an internal space configured to receive the ferrofluid 116 and / or the reservoir 120. In one non-limiting example, the barrel 108 may be formed by additive manufacturing, such as 3D printing. In other configurations, the barrel 108 may be formed by molding, machining, extrusion, or other fabrication techniques. The barrel 108 may be formed from a polymeric material, composite material, resin, thermoplastic, or other suitable material providing structural support for handwriting while permitting assembly of the magnetoelastic tip 104, ferrofluid 116, reservoir 120, coil 112, and grip 128.

[0054] In some configurations, the ferrofluid 116 may be disposed within the barrel 108 and may be subjected to the magnetic field of the magnetoelastic tip 104. In one non-limiting example, the ferrofluid 116 may form an ink and may be housed within the reservoir 120, which may be received by the barrel 108. In some examples, the reservoir 120 may be a replaceable reservoir, such as a plastic ink cartridge, allowing replenishment and replacement after prolonged use. The cap 124 may be coupled to the reservoir 120 and / or the barrel 108 and may be configured to reduce and / or prevent leakage and evaporation of the ferrofluid 116. In some configurations, the cap 124 may be removably attached to permit reservoir replacement, refilling, maintenance, or storage. In other configurations, the reservoir 120 may be integral with the barrel 108, the cap 124 may be omitted, or the ferrofluid 116 may be contained by other sealing arrangements.

[0055] In some examples, the ferrofluid 116 may include magnetic nanoparticles dispersed in a carrier solution. In one non-limiting example, the ferrofluid 116 may include single-domain nanomagnets, such as nanoparticles, optionally coated with a surfactant such as oleic acid and dispersed in a liquid carrier. In some examples, the magnetic nanoparticles may have a mean particle diameter between about 5 nm and about 10 nm, and in one non-limiting example may have a mean particle diameter of about 7.13 nm. Such particles may exhibit superparamagnetic behavior, enabling rapid response of the ferrofluid 116 to changes in the magnetic field produced by the magnetoelastic tip 104 and to motion-induced dynamics during handwriting. In otherconfigurations, the ferrofluid 116 may include other magnetic nanoparticle compositions, carrier fluids, surfactants, and / or stabilizers.

[0056] The coil 112 may be disposed around at least a portion of the barrel 108 and positioned proximate the ferrofluid 116 such that changes in magnetic flux associated with the magnetoelastic tip 104 and / or the ferrofluid 116 may induce an electrical signal in the coil 112. In one non-limiting example, the coil 112 may be formed by winding conductive yarns or wires around an exterior of the barrel 108. In other configurations, the coil 112 may be formed from copper wire, conductive fiber, conductive thread, printed conductive traces, flexible circuitry, metallic film, or other conductive structures. In some examples, the coil 112 may include one or more turns extending helically around the barrel 108 to convert flux variations into voltage and / or current signals according to electromagnetic induction. In other configurations, the coil 112 may be partially embedded in the barrel 108, integrated into a flexible sleeve, formed on a substrate coupled to the barrel 108, or otherwise arranged to detect magnetic flux variations generated during writing.

[0057] The grip 128 may be coupled over at least a portion of the barrel 108 and / or the coil 112. In one non-limiting example, the grip 128 may include a polyolefin material configured to enclose the coil 112 and provide a slip-resistant surface for a user. In other configurations, the grip 128 may include silicone, elastomer, polymer foam, shrink tubing, or other materials selected to improve ergonomics, comfort, durability, and / or protection of the coil 112. In some examples, the grip 128 may also assist in retaining the coil 112 in position relative to the barrel 108.

[0058] In operation, handwriting pressure and motion may act on the magnetoelastic tip 104 and the ferrofluid 116 to generate the handwriting-derived signal features 132. For example, pressure applied to the magnetoelastic tip 104 during on-surface writing may deform the magnetoelastic tip 104 and alter the orientation and spacing of magnetic particles within the magnetoelastic tip 104, thereby changing the magnetic field and magnetic flux associated with the tip 104. At the same time, handwriting motion, including on-surface movement and in-air movement, may induce motion of the ferrofluid 116 within the reservoir 120 and / or barrel 108. The combined flux variations associated with deformation of the magnetoelastic tip 104 and dynamic movement of the ferrofluid 116 may induce an electrical signal in the coil 112, which may then be analyzed to extract the handwriting-derived signal features 132 and assess motor abnormalities associated with a medical disorder.

[0059] The structural arrangement shown in FTG. 3 may provide several advantages. In some examples, the reservoir 120 may facilitate maintenance and long-term use, the cap 124 may improve storage stability of the ferrofluid 116, the coil 112 may provide direct electrical readout of flux variations, and the grip 128 may improve user handling. Additionally, the structure may be compatible with scalable manufacturing processes and low-cost materials, making the smart pen 10 suitable for broad deployment, including home settings and resource-limited environments.

[0060] While FIG. 3 illustrates one non-limiting configuration of the smart pen 10, alternative implementations are also contemplated. For example, the reservoir 120 may be omitted in favor of direct containment of the ferrofluid 116 within the barrel 108, the coil 112 may be internal or external to the barrel 108, the grip 128 may be omitted or integrated with the barrel 108, the cap 124 may be implemented as a plug, valve, seal, or threaded closure, and the magnetoelastic tip 104 may have different shapes, stiffnesses, magnetic particle loadings, and / or surface geometries depending on the desired writing performance and sensitivity. Similarly, the smart pen 10 may be implemented as a pen, stylus, marker, diagnostic writing utensil, or other hand-held device configured to convert fine motor movements into electrical signals.

[0061] Referring to FIG. 4, a portion of the smart pen 10 is shown schematically to illustrate the magnetoelastic tip 104 and the ferrofluid 116 that cooperatively convert handwriting movements into magnetic flux variations and corresponding electrical signals. In the illustrated non-limiting configuration, the magnetoelastic tip 104 may include magnetic particles embedded within a silicone matrix, and the ferrofluid 116 may be disposed adjacent to and / or above the magnetoelastic tip 104 such that the ferrofluid 116 is subjected to a magnetic field generated by the magnetoelastic tip 104. As described herein, the magnetoelastic tip 104 and the ferrofluid 116 may function together as a coupled magnetic-mechanical transduction system within the smart pen 10.

[0062] The magnetoelastic tip 104 may provide at least two functions within the smart pen 10. First, the magnetoelastic tip 104 may generate a magnetic field, such as a generally longitudinal and / or vertical magnetic field extending along the barrel 108, to sustain magnetization within the ferrofluid 116. Second, the magnetoelastic tip 104 may be deformable such that, under external mechanical stress during handwriting, the shape of the magnetoelastic tip 104 changes. This deformation may alter the orientation and spacing of the embedded magnetic particles and thereby shift the magnetic flux generated by the magnetoelastic tip 104. In some examples, handwritingpressure applied during on-surface writing may produce such deformation, while other movements, including tilting, tapping, or dynamic contact changes, may also contribute to deformation-dependent flux changes.

[0063] During operation, deformation of the magnetoelastic tip 104 and dynamic movement of the ferrofluid 116 may each contribute to time-varying magnetic flux. For example, when handwriting pressure deforms the magnetoelastic tip 104, the magnetic field distribution generated by the magnetoelastic tip 104 may change. In addition, movement of the smart pen 10 during handwriting may cause motion, redistribution, agitation, or other displacement of the ferrofluid 116. The combined effect of the deformation-induced magnetic flux variation from the magnetoelastic tip 104 and the motion-induced magnetic flux variation associated with the ferrofluid 116 may produce a changing net magnetic flux through the coil 112. In response, the coil 112 may generate a voltage and / or current according to electromagnetic induction, such as according to Faraday’s law, thereby converting handwriting activity into analyzable electrical signals.

[0064] While FIG. 4 illustrates one non-limiting configuration in which the magnetoelastic tip 104 includes magnetic particles embedded in a silicone matrix and the ferrofluid 116 includes superparamagnetic nanoparticles dispersed in a carrier fluid, other implementations are also contemplated. For example, the magnetoelastic tip 104 may include different particle loadings, particle sizes, particle distributions, matrix stiffnesses, and / or magnetic field strengths to tune sensitivity and writing performance. Similarly, the ferrofluid 116 may be formulated with different nanoparticle concentrations, carrier fluids, viscosities, surfactants, or magnetic characteristics depending on desired responsiveness, durability, and manufacturability. The present disclosure therefore is not limited to the particular material selections illustrated in FIG. 4.

[0065] Referring to FIG. 5, a schematic cross-sectional view of the smart pen 10 is shown to illustrate the spatial relationship among the magnetoelastic tip 104, the ferrofluid 116, the coil 112, the barrel 108, and the cap 124. In the illustrated non-limiting configuration, the magnetoelastic tip 104 may be arranged at a lower end of the smart pen 10, the ferrofluid 116 may be positioned above the magnetoelastic tip 104 within the barrel 108, and the coil 112 may extend around at least a portion of the barrel 108 in a region proximate the ferrofluid 116. The cap 124 may be positioned at an upper end of the smart pen 10 to close or seal the upper portion of the pen structure. This axial arrangement may permit magnetic interaction between the magnetoelastic tip 104 and theferrofluid 116, while also positioning the coil 112 to detect flux changes occurring within the pen during use.

[0066] In some configurations, the cross-sectional arrangement shown in FIG. 5 may enable the smart pen 10 to respond to different categories of handwriting movement through different, but coordinated, internal mechanisms. For example, when the smart pen 10 is pressed against a writing surface, force applied at the lower end of the smart pen 10 may act primarily through the magnetoelastic tip 104, while motion of the smart pen 10 along a stroke path or between strokes may induce movement of the ferrofluid 116 within the barrel 108. Because the ferrofluid 116 can be disposed in the magnetic field region associated with the magnetoelastic tip 104 and within the sensing region of the coil 112, both kinds of handwriting input may influence the net flux captured by the coil 112. In some configurations, the smart pen 10 may operate through at least two interacting mechanisms. A first mechanism may involve dynamic movement of the ferrofluid 116 within the barrel 108, and a second mechanism may involve deformation of the magnetoelastic tip 104 through a magnetoelastic effect. Both mechanisms may contribute to magnetic field variation and corresponding flux variation through the coil 112, thereby inducing voltage and / or current signals representative of handwriting activity. Because the two mechanisms may be active to different extents depending on the type of handwriting movement being performed, the arrangement shown in FIG. 5 may enable the smart pen 10 to track multiple motion patterns during handwriting using a single integrated structure.

[0067] Hand movements during handwriting may generally be categorized into at least two classes. A first class may include in-air movements, in which the smart pen 10 moves between strokes without contacting a writing surface. A second class may include on-surface movements, in which the magnetoelastic tip 104 contacts the writing surface and is subjected to writing pressure while forming one or more primary strokes. In some examples, these different movement classes may generate different internal responses within the smart pen 10, such that the smart pen 10 may detect both transitional motions and contact-based writing motions during a full writing sequence.

[0068] During in-air movements, the operating principle may be based primarily on movement of the magnetized ferrofluid 116. For example, acceleration of the smart pen 10 during stroke transitions, direction changes, rotation, spinning, or other non-contact movements may induce displacement, redistribution, or other dynamic movement of the ferrofluid 116 within the barrel 108. Because the ferrofluid 116 may be disposed within the magnetic field region established bythe magnetoelastic tip 104, such motion may produce time-varying magnetic flux through the coil 112. In this manner, the smart pen 10 may function during in-air writing as a handwriting-driven magnetic motion sensor capable of capturing acceleration-related and movement-related characteristics even in the absence of surface contact.

[0069] During on-surface movements, the smart pen 10 may operate through a combined mechanism. In addition to dynamic movement of the ferrofluid 116, pressure applied to the magnetoelastic tip 104 during contact with the writing surface may deform the magnetoelastic tip 104. Such deformation may shift the magnetic field and magnetic flux associated with the magnetoelastic tip 104 through the magnetoelastic effect. The resulting deformation-induced magnetic flux variation may combine with the motion-induced magnetic flux variation associated with the ferrofluid 116, thereby producing a changing net magnetic flux through the coil 112. The coil 112 may then generate a corresponding electrical signal, which may reflect both pressure-related and motion-related components of the handwriting movement.

[0070] FIG. 5 further illustrates that the smart pen 10 may respond not only to substantially linear writing movements, but also to additional motion patterns such as rotating and spinning movements. In some configurations, such movements may alter the relative position and orientation of the ferrofluid 116 within the magnetic field region, thereby changing the flux linkage with the coil 112. As a result, the smart pen 10 may capture handwriting-associated motor behavior across a broader range of fine motor actions than systems limited to surface-contact trajectory tracking alone. This may be beneficial in assessing medical disorders in which motor abnormalities are expressed not only in stroke formation, but also in transitional, rotational, or irregular motion patterns occurring during handwriting.

[0071] While FIG. 5 illustrates one non-limiting cross-sectional configuration, other internal layouts are also contemplated. For example, the longitudinal spacing between the magnetoelastic tip 104 and the ferrofluid 116 may be adjusted, the axial length and placement of the coil 112 may be varied, the barrel 108 may include one or more internal chambers, baffles, or flow-guiding features for controlling movement of the ferrofluid 116, and the cap 124 may be configured in a variety of sealing or venting arrangements. Similarly, the ferrofluid 116 may occupy more or less of the interior volume of the barrel 108 depending on desired signal sensitivity, dynamic response, writing feel, and manufacturability. Accordingly, the particular cross-sectional relationship shownin FIG. 5 should be understood as one example of an internal smart pen configuration capable of implementing the functions described herein.

[0072] Referring to FIG. 6, the smart pen 10 is shown with the ferrofluid 116 stored in a replaceable reservoir 120. In the illustrated non-limiting configuration, the reservoir 120 may include a cartridge body configured to contain the ferrofluid 116 and to be removably received by the barrel 108. In this manner, FIG. 6 illustrates one example of a replaceable ink-reservoir configuration in which the ferrofluid 116 may be securely housed while remaining serviceable and replaceable within the smart pen 10.

[0073] In some configurations, the reservoir 120 may be configured as a cartridge that can be inserted into and removed from the barrel 108 without disassembling the remainder of the smart pen 10. This may facilitate replacement of depleted ferrofluid 116, maintenance of the smart pen 10, and long-term use of the device in repeated monitoring applications. In one non-limiting example, the reservoir 120 may be formed from a plastic material and may define an internal chamber sized to hold a selected volume of the ferrofluid 116 while maintaining compatibility with the magnetic field region established by the magnetoelastic tip 104 and the sensing region of the coil 112.

[0074] In some examples, the ferrofluid 116 may be loaded into the reservoir 120 using an automated liquid-handling system. By way of non-limiting example, an automated liquid handler may dispense a predetermined amount of the ferrofluid 116 into the cartridge body 136 under controlled manufacturing conditions, after which the cap 124 may be applied to seal the reservoir 120. Such loading procedures may be advantageous for repeatable fdling volume, reduced contamination, improved process throughput, and compatibility with industrial-scale manufacturing. Accordingly, the replaceable reservoir 120 shown in FIG. 6 may support both practical day-to-day use and scalable fabrication of the smart pen 10.

[0075] The reservoir arrangement illustrated in FIG. 6 may also contribute to stability and reliability of the smart pen 10. As the ferrofluid 116 may be physically contained within a replaceable cartridge rather than being permanently fixed within the smart pen 10, the useful life of the smart pen 10 may be extended by replacing only the reservoir 120 when needed. This may be beneficial in longitudinal monitoring applications in which repeated handwriting measurements are acquired over extended periods.

[0076] While FIG. 6 illustrates one non-limiting replaceable-reservoir configuration, other implementations are also contemplated. For example, the reservoir 120 may be rigid or flexible, transparent or opaque. In some configurations, the reservoir 120 may be pre-filled and disposable. In other configurations, the reservoir 120 may be refillable. The filling operation may be manual, semi-automated, or fully automated, and may employ pipetting systems, metering pumps, syringebased dispensers, robotic filling platforms, or other liquid-delivery equipment suitable for introducing the ferrofluid 116 into the reservoir 120.

[0077] Referring to FIG. 7, the system 100 is shown during angular manipulation of the smart pen 10 to illustrate the effect of rotation angle on the electrical output of the smart pen 10. In the illustrated non-limiting example, the smart pen 10 is shown at different rotated positions relative to a vertical reference orientation, including positive and negative angular displacements. During handwriting, a user may naturally rotate or tilt the smart pen 10 in order to adjust writing comfort, stroke direction, or pen handling. FIG. 7 demonstrates that the smart pen 10 may continue to generate a detectable electrical output and convert handwriting into high-fidelity sensing signals under such rotated conditions, thereby supporting use of the smart pen 10 under realistic handwriting postures rather than only at a single fixed orientation.

[0078] In some examples, rotation of the smart pen 10 may alter the current generated by the coil 112. By way of non-limiting example, the smart pen 10 may be rotated to angles of about ±22.5 degrees, about ±45 degrees, and about ±67.5 degrees relative to a reference orientation. Experimental evaluation showed that the magnitude of the output current may increase as the angle of rotation increases, regardless of the rotational direction. In some configurations, this increase in current with greater rotation angle may be associated with increased displacement or redistribution of the ferrofluid 116 within the barrel 108 as the gravitational relationship between the ferrofluid 116 and the smart pen 10 changes.

[0079] In addition to discrete rotation angle testing, the smart pen 10 may also be evaluated at different holding angles relative to vertical. In some examples, holding angles between about 0 degrees and about 60 degrees may be examined. Such testing showed that time-domain features, including peak-to-peak timing, may remain substantially consistent across different holding angles, indicating that timing-based handwriting information may be largely unaffected by the angle at which the smart pen 10 is held. At the same time, signal amplitude may increase with increasing holding angle. In one non-limiting explanation, steeper holding angles may causegravity to shift the ferrofluid 116 farther from the magnetoelastic tip 104, thereby increasing motion of the ferrofluid 116 during handwriting and enhancing the amplitude of the induced signal in the coil 112.

[0080] In some configurations, the relationship between holding angle and signal amplitude may be used in a calibration procedure. For example, because signal amplitude may vary predictably with holding angle, one or more correction factors may be determined for a given user and applied during later analysis of handwriting-derived signals. In one non-limiting example, a calibration procedure may include measuring the response of the smart pen 10 at one or more known holding angles, determining a relationship between holding angle and signal amplitude, and using that relationship to normalize subsequently acquired signals. Such calibration may permit the system 100 to account for inter-user differences in writing posture while preserving the utility of the smart pen 10 for quantitative handwriting analysis.

[0081] The results illustrated in FIG. 7 therefore can indicate that the smart pen 10 may maintain sensing functionality across a range of rotational and holding orientations encountered during natural handwriting. This may be beneficial because it allows the smart pen 10 to capture handwriting-related motor information without requiring the user to maintain a rigid or highly constrained grip posture. In this manner, FIG. 7 can support the use of the smart pen 10 as a practical writing utensil for real-world neurological monitoring while also providing a basis for user-specific angle compensation when desired.

[0082] In some examples, the magnetoelastic tip 104 can demonstrate a high remanence value, enabling it to emerge as a stable and permanent magnetic field source. After impulse magnetization, the magnetoelastic tip can provide a magnetic field accompanied by a surface magnetic flux (e.g., capable of reaching 70 mT). This vertically distributed magnetic field can effectively magnetize the ferrofluid ink 116. Moreover, after the ferrofluid ink 116 in the smart pen 10 is periodically used over an extended period, both the viscosity of the ferrofluid ink 116 and the magnetic properties of the magnetoelastic tip 116 can remain stable, showing no noticeable variation. This confirms the long-term stability of system 100 for practical applications.

[0083] In some examples, the magnetoelastic effect in the magnetoelastic tip 104 can be quantified by measuring its magnetomechanical coupling factor (e.g., dsa), defined as the magnetic flux variation under applied stress. Within the range of forces applied during handwriting, the magnetoelastic tip 104 can deform and exhibit a d33 of up to 4.9* 1 O'8T / Pa in the vertical directionas a non-limiting example. These results confirm that the magnetoelastic tip 104 can effectively convert handwriting pressure into magnetic flux variations, inducing voltage signals in the coil 112.

[0084] In some examples, that handwriting pressure can effectively deform the magnetoelastic tip 104, confirmed by strain-stress testing results. Moreover, prolonged handwriting may raise the temperature of the smart pen 10 close to body temperature, and ferrofluid ink 116 can remain stable under temperature fluctuations. The results show that the viscosity-shear rate curves measured at room temperature (e.g., approximately 25°C) and typical hand temperature (e.g., approximately 33°C) can be nearly identical. Similarly, magnetic flux measurements of the smart pen 10 revealed no significant variation between the two temperatures. These findings confirm the stability of the ferrofluid ink 116 across the tested temperature range, ensuring reliable performance during prolonged handwriting analysis in practical applications.

[0085] In some examples, the smart pen 10 can detect and / or capture subtle, high-frequency motions during handwriting. The system 100 can exhibit high sensitivity and accuracy in converting subtle biomechanical forces into real-time voltage signals. The results demonstrated that the smart pen 100 can accurately record high-frequency movements (e g., in the range of 10-12 Hz) with a high signal-to-noise ratio (SNR) (e.g., up to 39 dB). Since human handwriting typically occurs at frequencies below 5 Hz, these findings validate the smart pen’s 10 capability to precisely record fast handwriting movements without noticeable lag, confirming its suitability for high-frequency motion detection.

[0086] In some examples, the smart pen 10 can remain stable in practical handwriting applications. For example, after 10,010 cycles of continuous mechanical excitation, current signals of the smart pen 10 can remain stable, indicating minimal variation in signal output and demonstrating its prolonged durability. Moreover, the smart pen 10 can provide waterproof properties, ensuring stable performance despite hand perspiration. In some configurations, the smart pen 10 can maintain stable functionality without any performance degradation, even after being sprayed with water while wet. Additionally, the smart pen 10 can stay reliable across diverse writing environments. The results indicate that the roughness and softness of typical writing surfaces have a minimal impact on signal generation. These findings demonstrate the smart pen’s robust sensing performance and resilience against environmental variations.

[0087] Referring to FIG. 8, the system 100 is illustrated in a clinical-use context in which the smart pen 10 may be used for handwriting-based assessment of a neurological condition, such as PD. In the illustrated non-limiting example, a subject can use the smart pen 10 to perform a handwriting task, such as Task 1, which may include drawing continuous wavy lines. FIG. 8 further schematically illustrates neurological and handwriting-related differences that may be associated with PD and that may be reflected in signals acquired by the smart pen 10 during use. In some configurations, handwriting may serve as a useful source of motor information because handwriting involves coordinated cognitive, perceptual, kinesthetic, and fine motor processes. As a result, abnormalities affecting fine motor control may manifest during writing and may be detected through analysis of handwriting-derived signals. In one non-limiting example, the smart pen 10 may be used to record one or more standardized writing tasks, drawing tasks, freeform writing tasks, or other pen-based motion tasks in order to generate signal features 132 that are representative of the subject’s handwriting biomechanics.

[0088] FIG. 8 also schematically illustrates a neurological basis for such handwriting differences. In some examples, PD may be associated with gradual loss or dysfunction of dopamine-producing neurons in the brain, which may adversely affect motor control and fine motor skill. Because handwriting depends on precise control of hand and finger motion, such neurological changes may influence the subject’s ability to perform writing tasks smoothly and consistently. The smart pen 10 may therefore be used as a tool for capturing handwriting-related consequences of such motor impairment in a quantitative manner.

[0089] In some examples, handwriting produced by subjects affected by PD may differ from handwriting produced by subjects not affected by PD. By way of non-limiting example, handwriting associated with PD may exhibit irregularities related to tremor, bradykinesia, segmented movement, altered stroke timing, altered pressure control, and / or other fine motor abnormalities. Such differences may appear in the written output itself and may also be reflected in the electrical signals generated by the smart pen 10 during handwriting. In some configurations, the system 100 may detect these differences by analyzing one or more time-domain features, amplitude-domain features, frequency -domain features, and / or localized waveform characteristics of the signal features 132.

[0090] In one non-limiting example, tremor-related handwriting behavior may produce additional minor peaks, oscillations, or other localized signal variations in the handwriting-derivedsignals acquired by the smart pen 10, whereas smoother handwriting movements may generate comparatively less irregular signal structure. Similarly, bradykinesia-related behavior may be reflected by prolonged writing duration, increased cycle time, segmented movement patterns, or pauses between sub -movements. Because the smart pen 10 can capture both on-surface and in-air handwriting activity, the system 100 may be configured to detect abnormalities occurring during stroke formation as well as abnormalities occurring during movement transitions between strokes. In some configurations, the smart pen 10 may be used in a human study, screening workflow, monitoring program, or other evaluation protocol in which one or more subjects perform one or more designated handwriting tasks and the resulting signals are analyzed by the processor 16. A human study involving both PD patients and healthy subjects was conducted using the pen 10, and the system can distinguish PD patients from healthy subjects with an average accuracy of 96.22%.

[0091] The use context illustrated in FIG. 8 may be beneficial because handwriting is a familiar daily activity that can be performed with minimal user training, while the smart pen 10 can generate compact analyzable signals without requiring large imaging equipment or other specialized clinical instrumentation.

[0092] Referring to FIG. 9, the system 100 is shown in a configuration in which handwriting-derived signals acquired by the smart pen 10 are analyzed by the processor 16 using a machinelearning model for personalized handwriting analysis and diagnosis of PD. In the illustrated nonlimiting example, the processor 16 may receive handwriting signals generated during use of the smart pen 10 and may provide such signals as input to a one-dimensional convolutional neural network (ID CNN) configured to distinguish handwriting patterns associated with PD from handwriting patterns not associated with PD. In some configurations, the processor 16 may therefore implement an end-to-end analysis pipeline that can convert handwriting-derived electrical signals into a diagnostic output.

[0093] In some examples, the handwriting-derived signals may be preprocessed before being provided to the machine-learning model. By way of non-limiting example, the processor 16 may standardize the signals, segment the signals into selected windows or slices, and optionally perform one or more additional preprocessing operations, such as normalization, filtering, or artifact reduction. The resulting signal segments may then be supplied to the ID CNN, which may be configured to identify localized waveform features associated with handwriting biomechanics and motor abnormalities.

[0094] In one non-limiting example, the ID CNN shown in FIG. 9 may include a plurality of convolutional layers, such as convolutional layers 1-5, each may be configured to extract features from the handwriting-derived signals. In some configurations, one or more of the convolutional layers may be followed by batch normalization and a rectified linear unit (ReLU) activation, as well as one or more pooling layers, such as max-pooling layers, to reduce dimensionality while preserving feature information. The extracted features may then be passed to one or more fully connected layers, including in one example a fully connected layer having 128 features and a final fully connected layer configured to output a diagnostic classification. Such a configuration may enable the processor 16 to learn localized time-domain patterns, including waveform irregularities associated with tremor, bradykinesia, and other handwriting-related abnormalities.

[0095] In some examples, the output of the ID CNN may include a classification indicating whether the acquired handwriting signals are more consistent with PD-related handwriting patterns or with non-PD handwriting patterns. In other configurations, the processor 16 may additionally output one or more confidence values, probabilities, severity indicators, or related diagnostic metrics. Because the model can operate on signals generated directly by the smart pen 10, the system 100 may provide a compact and computationally efficient approach for handwriting-based neurological assessment without requiring large imaging systems.

[0096] FIG. 9 may further illustrate comparison of performance across multiple machinelearning models. In one non-limiting example, the processor 16 may evaluate a plurality of candidate models, including a ID CNN, a ID CNN combined with long short-term memory (ID CNN+LSTM), an LSTM model, a random forest (RF) model, an extreme gradient boosting (XGBoost or XGB) model, and a logistic regression (LR) model. In some examples, performance may be evaluated using one or more testing metrics, including accuracy and Fl score, across repeated experiments. Such comparison may be useful for selecting a model architecture that provides suitable discrimination of PD-related handwriting signals.

[0097] In one non-limiting example, evaluation across repeated experiments indicated that the ID CNN can achieve high testing accuracy and Fl score. For example, the ID CNN can achieve an average testing accuracy of about 0.9622 ± 0.0251 and an average Fl score of about 0.9266 ± 0.0523, other evaluated models may also be utilized, including ID CNN+LSTM, LSTM, RF, XGB, and LR.

[0098] In some configurations, the processor 16 may train, validate, and / or update the machine-learning model using subject-specific data, population data, longitudinal monitoring data, or combinations thereof. In some examples, signals from different subjects may be separated into training and testing sets in order to reduce data contamination and evaluate performance on previously unseen subjects. In other configurations, one or more transfer-learning or hybrid-model approaches may additionally be employed. Accordingly, FIG. 9 illustrates that the system 100 may integrate the smart pen 10 with processor-based machine-learning analysis to provide personalized handwriting analysis for PD diagnosis and related neurological assessment.

[0099] The above discussion is presented to enable a person skilled in the art to make and use embodiments of the invention. Various modifications to the illustrated embodiments will be readily apparent to those skilled in the art, and the generic principles herein can be applied to other embodiments and applications without departing from embodiments of the invention. Thus, embodiments of the invention are not intended to be limited to embodiments shown, but are to be accorded the widest scope consistent with the principles and features disclosed herein. The following detailed description is to be read with reference to the figures, in which like elements in different figures have like reference numerals. The figures, which are not necessarily to scale, depict selected embodiments and are not intended to limit the scope of embodiments of the invention. Skilled artisans will recognize the examples provided herein have many useful alternatives and fall within the scope of embodiments of the invention.

[0100] For any embodiments of the invention are explained in detail, it is to be understood that the invention is not limited in its application to the details of construction and the arrangement of components set forth in the following description or illustrated in the attached drawings. The invention is capable of other embodiments and of being practiced or of being carried out in various ways. Also, it is to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting. For example, the use of “including,” “comprising,” or “having” and variations thereof herein is meant to encompass the items listed thereafter and equivalents thereof as well as additional items.

[0101] In some implementations, devices or systems disclosed herein can be utilized, manufactured, or installed using methods embodying aspects of the invention. Correspondingly, any description herein of particular features, capabilities, or intended purposes of a device or system is generally intended to include disclosure of a method of using such devices for theintended purposes, a method of otherwise implementing such capabilities, a method of manufacturing relevant components of such a device or system (or the device or system as a whole), and a method of installing disclosed (or otherwise known) components to support such purposes or capabilities. Similarly, unless otherwise indicated or limited, discussion herein of any method of manufacturing or using for a particular device or system, including installing the device or system, is intended to inherently include disclosure, as embodiments of the invention, of the utilized features and implemented capabilities of such device or system.

[0102] The above discussion is presented to enable a person skilled in the art to make and use embodiments of the invention. Various modifications to the illustrated embodiments will be readily apparent to those skilled in the art, and the generic principles herein can be applied to other embodiments and applications without departing from embodiments of the invention. Thus, embodiments of the invention are not intended to be limited to embodiments shown, but are to be accorded the widest scope consistent with the principles and features disclosed herein. The above detailed description is to be read with reference to the figures, in which like elements in different figures have like reference numerals. The figures, which are not necessarily to scale, depict selected embodiments and are not intended to limit the scope of embodiments of the invention. Skilled artisans will recognize the examples provided herein have many useful alternatives and fall within the scope of embodiments of the invention.

[0103] It is to be understood that the invention is not limited in its application to the details of construction and the arrangement of components set forth in the following description or illustrated in the attached drawings. The invention is capable of other embodiments and of being practiced or of being carried out in various ways. Also, it is to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting. For example, the use of “including,” “comprising,” or “having” and variations thereof herein is meant to encompass the items listed thereafter and equivalents thereof as well as additional items.

[0104] Additionally, unless otherwise specified or limited, the terms “about” and “approximate,” as used herein with respect to a reference value, refer to variations from the reference value of ± 15% or less, inclusive of the endpoints of the range. Similarly, the term “substantially equal” (and the like) as used herein with respect to a reference value refers to variations from the reference value of less than ± 30%, inclusive. Where specified, “substantially” can indicate in particular a variation in one numerical direction relative to a reference value. Forexample, “substanti lly less” than a reference value (and the like) indicates a value that is reduced from the reference value by 30% or more, and “substantially more” than a reference value (and the like) indicates a value that is increased from the reference value by 30% or more.

[0105] Also as used herein, ordinal numbers are used for convenience of presentation only and are generally presented in an order that corresponds to the order in which particular features are introduced in the relevant discussion. Accordingly, for example, a "first" feature may not necessarily have any required structural or sequential relationship to a "second" feature, and so on. Further, similar features may be referred to in different portions of the discussion by different ordinal numbers. For example, a particular feature may be referred to in some discussion as a "first" feature, while a similar or substantially identical feature may be referred to in other discussion as a "third" feature, and so on.

[0106] The description of the different advantageous embodiments has been presented for purposes of illustration and description and is not intended to be exhaustive or limited to the embodiments in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. Further, different advantageous embodiments may provide different advantages as compared to other advantageous embodiments. The embodiment or embodiments selected are chosen and described in order to best explain the principles of the embodiments, the practical application, and to enable others of ordinary skill in the art to understand the disclosure for various embodiments with various modifications as are suited to the particular use contemplated.

Claims

CLAIMS1. A diagnostic writing utensil for determining a presence or quantifying symptoms of a medical disorder, comprising:a barrel forming the diagnostic writing utensil;a magnetoelastic tip coupled to the barrel and having a shape that changes during use of the writing utensil, and wherein the magnetoelastic tip generates a magnetic field configured to extend along the barrel that varies with changes in the shape of the magnetoelastic tip;a ferrofluid arranged in the barrel to be subjected to the magnetic field of the magnetoelastic tip; anda coil extending around the ferrofluid,wherein changes in the shape of the magnetoelastic tip cause changes in the magnetic field generated by the magnetoelastic tip that induces motion of the ferrofluid, andwherein motion of the ferrofluid induces a current in the coil that is indicative of a presence or quantification of symptoms of a medical disorder.

2. The diagnostic writing utensil of claim 1, wherein the magnetoelastic tip comprises magnetic particles embedded in a silicone matrix.

3. The diagnostic writing utensil of claim 2, wherein the magnetic particles are neodymium-iron-boron magnets.

4. The diagnostic writing utensil of claim 1, wherein the ferrofluid comprises single-domain magnetic nanoparticles dispersed in a carrier solution.

5. The diagnostic writing utensil of claim 4, wherein the magnetic nanoparticles have a mean particle diameter between 5-10 nm.

6. The diagnostic writing utensil of claim 1, further comprising a replaceable ink reservoir for housing the ferrofluid and wherein the ferrofluid forms an ink.

7. A writing utensil comprising:a barrel forming a writing utensil;a magnetoelastic tip coupled to the barrel and having a shape that changes during use of the writing utensil, and wherein the magnetoelastic tip generates a magnetic field configured to extend along the barrel that varies with changes in the shape of the magnetoelastic tip;a ferrofluid arranged in the barrel to be subjected to the magnetic field of the magnetoelastic tip;a coil extending around the ferrofluid; andwherein changes in the shape of the magnetoelastic tip causes changes in the magnetic field generated by the magnetoelastic tip that induces motion of the ferrofluid and wherein motion of the ferrofluid induces a current in the coil.

8. A method comprising:providing a writing utensil having a magnetoelastic tip, ferrofluid, and a coil; converting handwriting movements using the writing utensil into magnetic flux variations using the magnetoelastic tip and ferrofluid; andgenerating electrical signals from the magnetic flux variations using the coil.

9. The method of claim 8, wherein converting handwriting movements into magnetic flux variations comprises:deforming the magnetoelastic tip in response to handwriting movements; and moving the ferrofluid in response to handwriting movements.

10. The method of claim 9, wherein deforming the magnetoelastic tip alters an orientation and spacing of magnetic particles embedded within the magnetoelastic tip.

11. The method of claim 8, further comprising analyzing the electrical signals to detect symptoms associated with a medical disease or disorder.

12. The method of claim 8, further comprising analyzing the electrical signals by delivering the electrical signals into a machine learning model trained to classify a medical disorder or disease.

13. The method of claim 12, wherein the medical disorder or disease includes a neurodegenerative disease or disorder, or medical condition that causes tremors.

14. The method of claim 8, further comprising performing a calibration procedure to account for variations in handwriting.

15. A system for diagnosing a medical condition, comprising:a smart pen having a magnetoelastic tip, ferrofluid, and a coil for converting handwriting movements into electrical signals; anda processor configured to analyze the electrical signals to distinguish handwriting patterns of the medical condition from handwriting patterns without the medical condition.

16. The system of claim 15, wherein the processor is configured to apply the electrical signals to a machine learning model to analyze the electrical signals.

17. The system of claim 16, wherein the machine learning model is a one-dimensional convolutional neural network.

18. The system of claim 15, wherein the smart pen further comprises a replaceable ink reservoir for housing the ferrofluid.

19. The system of claim 18, wherein the replaceable ink reservoir includes a cap configured to prevent leakage and evaporation of the ferrofluid.

20. The system of claim 19, wherein the magnetoelastic tip comprises magnetic particles embedded in a silicone matrix, and the ferrofluid comprises single-domain magnetic nanoparticles dispersed in a carrier solution.

Images