top of page
orthopractis.com

HandTremorApp

Description 

Hand tremor recognition is a critical aspect in diagnosing and monitoring neurological conditions such as Parkinson's disease, essential tremor, and other movement disorders. Utilizing Apple's Vision Pro capabilities and leveraging its advanced sensors can provide an innovative, non-invasive method to measure and analyze hand tremors accurately. The app provides a real-time tremor analysis system using Apple Vision Pro’s sensors. The app continuously captures motion data from the user’s hands, processes the data to calculate key metrics related to hand tremors. App  is capturing four key points over terminal phalanges for Index Middle Ring and small finger. By tracking these points over time, it’s possible to analyze the amplitude, frequency, and velocity of tremors real time. Data are processed and calculated real time  based on  mathematical models. Implementing hand tremor recognition using Apple’s Vision Pro and Swift could be proven valuable tool for clinicians. By capturing detailed hand movements and analyzing them according to medical standards, it’s possible to compute indices that  can assist in diagnosis and monitoring. The system must be validated against clinical data to ensure accuracy and reliability. No approval from regulatory bodies like the FDA for clinical use has currently granted.Temporary is for research use only.

Data are exported in console view by pressing the respective button Console view , and all collected data are exported as text, which can be copied and saved to a text file. Above  images with plots are printend with third party apps  by using  apps exported data.Exported data  as text can be selected in dialogue window by pressing the respective button of console view, once  in view text data can be  copied and can be  used for further use in number or excel.

TremorAnalyzer.jpg
Apple-Vision-Pro-glass.jpg
IMG_2066_resized.PNG
IMG_2064_resized.PNG
IMG_2065_resized.PNG
06A9D494-E995-4FB6-A8BF-B8CEA072CCF7_1_102_o.jpeg

How to measure 

By activating the show Immersive view Button, the user enters the Immersive space and tracking both hand measurement starts. By detecting continuously 27 hand landmarks, tiny green spheres appear over the hand and fingers. Each green sphere represents a certain anatomic structure of the hand. The center of the carpus is estimated and depicted by a red sphere over its physical location and represents the origin of the XYZ coordination system, serving as a stable reference point for all calculations.

Only the frontal plane appears with color in both hands. The respective point at the tip of each finger (Index, Middle, Ring, Little) is used, and the distance of each point over the frontal plane is continually measured.

The displacement from its mean position over a time window t is calculated for each point as Average (Avg) and presented on the screen. Maximum (Max) and Minimum (Min) values of the same measurement depict the highest and lowest observed displacements in real-time in mm.

Amplitude (Amp) is calculated as the maximum magnitude of displacement for each point. Max instantaneous Velocity (maxV) is also calculated, showing the highest observed velocity in mm per second (mm/sec). Max instantaneous Acceleration (maxAc) shows the highest observed acceleration in mm per second squared (mm/sec²). The frequency of tremor is the number of oscillation cycles per unit time. For each point, a time series of its displacement component is used to perform a Fast Fourier Transform to obtain the frequency spectrum. The dominant frequency component, corresponding to the peak in the spectrum, is identified as the Dominant Frequency (dFreq), showing the tremor frequency in Hertz (Hz).

Sampling Interval can be adjusted by pressing the plus or minus button near the Time window button. The time interval changes accordingly. The user can also adjust the frame rate using the frame/sec button; by pressing the plus or minus button, the frame rate is adjusted accordingly. For exporting data conforming to the Fahn-Tolosa-Marin Tremor Rating Scale (FTMTRS), setting the time interval to 1 sec and frame rate to 25 is feasible. Data are exported in console view by pressing the respective button, and all collected data are exported as text, which can be copied and saved to a text file.

Measurement    Description                                                          Unit
Avg    Displacement from mean position over time window t    mm
Max    Highest observed displacement    mm
Min    Lowest observed displacement    mm
Amp    Maximum magnitude of displacement    mm
maxV    Highest observed velocity    mm/sec

Once measurement have been ended by pressing the console view button   text appear inside a window with all measured relevant data from capture  as text that can be selected in dialogue window by relevant gestures. Then text data can be copied by selecting all or part of them  and paste in  a new file in notes and then and can be  used for further evaluation  in numbers or excel or other plotting software.

Once sampling and t time is changed you have to enter once again in the immersive view in order to update the data.

 Target audience

1. Neurologists and Movement Disorder Specialists: App could proven a good tool  for  specialists diagnosing and managing tremor disorders like Parkinson’s disease, essential tremor, and dystonia. By offering real-time, data-driven insights into tremor characteristics, it  could  aid in more precise diagnoses, treatment planning, and tracking of disease progression.

2. Rehabilitation Therapists (Physical and Occupational Therapists): For therapists developing personalized treatment plans for patients with motor impairments, App could  provide valuable metrics on tremor frequency, amplitude, and rhythmicity. This data supports tailored therapies and allows for effective monitoring of progress over time.

3. Neurology Researchers and Movement Disorder Scientists: With access to an anonymized, centralized database of tremor metrics, researchers can conduct in-depth studies on movement disorders, potentially leading to breakthroughs in understanding and treating these conditions.

4. Hospitals and Specialized Clinics: Neurology departments, hospitals, and movement disorder clinics can use App to standardize tremor assessment, improve diagnostic accuracy, and enhance patient care by making informed, data-backed treatment decisions.

​   

App integration with Apple Vision Pro with an engaging, immersive experience, is  making complex data on tremor severity, progression, and treatment responses more accessible , available  also in real time for recording procedure  that  elevates the vaque and laborious  tremor assessment by leveraging  advanced hand-tracking capabilities of Apple vision Pro . Capture tremor metrics  element  , could  offer benefits across clinical, therapeutic, research, and patient support . App could be proven a valuable tool across clinical, research, and patient support domains in the neurology field in comming future . Further research and support is needed.

Disclaimer.

Regarding its plausibility app offers no diagnosis or treatment. App provide an early indication that further evaluation may be warranted by Speciality Doctor.  Explicitly is announced that the apps are not for diagnosis. Clinical judgment and experience are required to properly use the software. App alone do not replace an M.D. or specialist. All information received from the App output must be reviewed before any attempted treatment. The software is not for primary image interpretation. Any influence to the operators in making decisions remains user own responsibility and experience. App does not dispense medical advice. Patient should seek a doctor’s advice in addition using the app and /or before making any medical decisions for themself. Never substitute or replace doctor's advice or change treatment modalities based on any measured outcome. App indicated for assisting healthcare professionals for scientific and research reason. Clinical judgment and experience are required to properly use the app and further research and validation is pending in coming future.The app is not a substitute for professional medical advice.Any medical information provided by the app should be used with caution and not relied upon exclusively for medical decision-making.

Why only in Apple Vision Pro? 

 

App is uniquely suited for Apple Vision Pro because of the device’s advanced capabilities in hand tracking, immersive visualization, and real-time processing, which significantly enhance tremor analysis and diagnostics. Here’s why Apple Vision Pro makes App stand out:

1. Advanced Hand Tracking and 27-Point Precision

    •    Accurate 27-Point Hand Tracking: Apple Vision Pro captures intricate hand movements with 27 distinct points, enabling App to analyze tremor characteristics such as amplitude, frequency, velocity, and rhythmicity with exceptional accuracy.
   •    3D Spatial Awareness: The device accurately measures hand movements in a 3D space, making it ideal for detecting subtle tremor patterns across different axes, essential for diagnosing various tremor types like Parkinsonian or essential tremors.

2. Real-Time Data Processing and Immediate Feedback

    •    Low-Latency Processing: Vision Pro’s real-time processing power enables App provide instant feedback on tremor measurements, helping clinicians monitor changes during the exam without delays.
   •    Enhanced Responsiveness: With Vision Pro’s powerful chip and optimized hardware, App can handle high-frame-rate data capture and calculations, enabling real-time tremor tracking and immediate adjustments in clinical and therapeutic settings.

3. Immersive Visualization for Enhanced Clinical Insight

    •    Immersive 3D Environment: Apple Vision Pro’s immersive display allows clinicians to visualize tremor data in a 3D space, making it easier to interpret complex movement patterns. This unique perspective provides deeper insights into tremor characteristics, which traditional devices or 2D screens cannot offer.
   •    Spatial Visualization of Metrics: By visualizing tremor data as it occurs in real time, clinicians can understand the depth, range, and orientation of tremor motion, aiding in more comprehensive assessments.

4. Interactive Interface and User Engagement

    •    Intuitive Gestures and Controls: Vision Pro allows users to interact with the app through gestures, enhancing usability for both clinicians and patients. The device’s intuitive interface makes it easy for users to adjust parameters, track measurements, and engage directly with their data.
   •    Adjustable Data Sampling and Frame Rate: With Vision Pro, users can adjust data sampling intervals, frame rates, and display real-time adjustments directly on the immersive screen. This feature enables customizable data collection that can be tailored for different tremor types or clinical protocols.

5. Optimized Data Collection for Machine Learning for coming versions.

    •    High-Quality Data for ML: Apple Vision Pro provides highly accurate, high-frequency motion data, which is invaluable for training machine learning models in TremorInsight. This data can be used to refine tremor classifications, improve model accuracy, and contribute to the centralized database for ongoing learning.
   •    Continuous Learning and Adaptation: Vision Pro’s precise tracking allows the app to gather detailed, real-time data, which can be fed into a central database for future model updates and improvements, making App could progressively be  more accurate and effective in diagnosing and categorizing tremors.

6. Secure Data Storage and Processing

    •    On-Device Processing: Vision Pro’s robust processing capabilities mean that sensitive tremor data can be analyzed on-device without needing external processing. This ensures patient privacy while still delivering high-performance data analysis.
   •    Seamless Integration with Apple’s Ecosystem: Data from Vision Pro can easily integrate with Apple’s ecosystem, allowing for secure storage, HealthKit integration, and potential synchronization with other medical apps or research platforms for comprehensive care.

Summary

App  on Apple Vision Pro could offer an unparalleled approach to tremor analysis by combining high-precision hand tracking, real-time processing, immersive visualization, and secure data management. These features create a powerful, clinician-focused tool that elevates diagnostics, patient engagement, and research in tremor and movement disorder management, making it uniquely suited to Apple Vision Pro.

Data are exported in console view by pressing the respective button Console view , and all collected data are exported as text, which can be copied and saved to a text file. Images with plots are printed with third party apps  by using  apps exported data. Exported data  as text can be selected in dialogue window by pressing the respective button of console view, once  in view text data can be  copied and can be  used for further use in number or excel.

Comprehensive Analysis of Hand Tremor Using Apple Vision Pro with 27-Point Hand Tracking Technology 

 

Hand tremors are a common symptom in neurological disorders, often requiring precise measurements for accurate diagnosis and treatment. This study introduces a novel application leveraging Apple Vision Pro’s 27-point hand-tracking technology to measure and analyze tremor characteristics such as amplitude, frequency, velocity, acceleration, rhythmicity, and asymmetry in real time. These metrics provide valuable insights into the nature and severity of tremors, enabling more accurate differentiation between essential tremor, Parkinsonian tremor, and other movement disorders. Additionally, the application of machine learning models for tremor classification enhances the diagnostic process, offering clinicians a powerful, data-driven tool for improved patient outcomes.

Accurate measurement of tremor is essential for diagnosis, differentiation between types of tremor (e.g., Parkinsonian tremor vs. essential tremor), monitoring disease progression, and evaluating treatment efficacy. The main measurement variables include amplitude, frequency, velocity, acceleration, rhythmicity, and asymmetry. Various methods and devices have been developed to assess these variables.Background: Tremor, an involuntary oscillatory movement, is commonly observed in neurological conditions such as Parkinson’s disease, essential tremor, and cerebellar ataxia. Precise characterization of tremor parameters is essential for accurate diagnosis, treatment, and monitoring. Traditional methods of tremor assessment, including clinical observation and paper-based scales, are often limited by subjectivity and lack quantitative precision.Accurate assessment of tremor variables is crucial for effective diagnosis and management of movement disorders. Traditional methods and devices offer valuable tools but come with limitations such as invasiveness, lack of portability, and limited user engageme

Kinesia™ System:

A wearable sensor system designed specifically for quantifying tremor and other movement disorder symptoms.

Actigraphy Devices:

Wrist-worn devices that monitor movement over extended periods to assess tremor in daily activities.

Digitized Spiral Analysis:

Patients draw spirals on a digitizing tablet, and software analyzes tremor characteristics.

Force Sensors:

Measure the force exerted during tremor movements, providing data on amplitude and frequency.

Video Analysis Tools:

Software applications analyze recorded videos to extract quantitative tremor metrics.

Optoelectronic Systems:

Use cameras and reflective markers to capture detailed motion data.

 

 

 

Apple Vision Pro is a spatial computing device offers immersive experiences and sophisticated interaction capabilities.Advantages Over Existing Devices High-Resolution Cameras can capture detailed movements of the user’s limbs and fingers with high precision.Enables accurate tracking of tremor amplitude, frequency, and velocity without the need for external markers or wearables.

Provide three-dimensional mapping of the environment and the user’s movements.Allow for comprehensive motion analysis in a natural setting, capturing tremor characteristics during daily activities. Real-time on device  data processing without reliance on external devices or cloud services can offer Computational Power and Machine Learning Capabilities Immediate analysis and feedback on tremor patterns, enabling timely interventions.Utilize Apple’s Core ML and health-focused APIs.Advanced algorithms can classify tremor types, predict tremor episodes, and differentiate between essential tremor and Parkinsonian tremor. Interactive exercises and games that can be used for both assessment and rehabilitation.Encourage patient engagement, making regular monitoring less burdensome and more enjoyable.Visual overlays can provide real-time feedback on tremor severity or control strategies.Patients can see immediate effects of relaxation techniques or medication adjustments.Seamless integration with Apple’s Health app and other health records.Centralized data collection allows for comprehensive tracking over time, which can be shared with healthcare providers. Apple’s commitment to user privacy ensures secure handling of sensitive health data.Reduces concerns over data breaches, encouraging more users to participate in digital health monitoring.Enables continuous or prolonged monitoring sessions without significant discomfort.Users can perform assessments without needing to hold devices or attach sensors.More accurate representation of natural tremor occurrences.

Considerations and Potential Limitations

 

While Apple Vision Pro presents promising features, several factors should be considered:

 Apple Vision Pro has not been clinically validated for tremor assessment.Further research and validation studies are necessary to establish its efficacy and accuracy in a medical context.

High initial cost may limit accessibility for some patients.Over time, costs may decrease, or programs may subsidize the device for medical use.Medical applications may require approval from regulatory bodies like the FDA.It’s unclear if Apple Vision Pro has sought or obtained such approvals for medical use.Despite strong privacy measures, sensitive health data handling requires strict compliance with healthcare regulations

 

Literature Review and Key Measurement Variables

 

Amplitude

 

Amplitude is the maximum displacement of tremor movement from the mean position, measured in millimeters. This metric provides a direct indication of tremor severity.

Clinical Relevance: Larger amplitude tremors are often indicative of essential tremor, while lower amplitude tremors can be associated with Parkinsonian tremor. Research by Elble et al. and Findley et al. demonstrates that amplitude measurement is crucial for distinguishing between tremor type Studies such as Elble et al. (2003) have shown that amplitude is a distinguishing characteristic in the clinical presentation of different tremor disorders, supporting its role as a core metric in tremor assessment.

 

Frequency

 

Frequency represents the number of oscillatory cycles per second (Hz). Tremor frequency can vary widely depending on the underlying pathology.

Clinical Relevance: For instance, Parkinsonian tremors typically present at frequencies between 4-6 Hz, whereas essential tremors often range from 8-12 Hz. Deuschl et al. (1998) emphasize that frequency is a reliable differentiator between types of tremor Studies by Deuschl and Raethjen have consistently shown that frequency is a reliable and non-invasive diagnostic metric for differentiating tremor types, further validated by kinetic and static tremor analyses. 

 

 Velocity and Acceleration

Velocity is the speed of tremor movement, while acceleration is the rate of change of this speed. These measurements reflect the kinetics of tremor oscillation.Higher velocities and accelerations may indicate cerebellar tremors or drug-induced dyskinesias. Lower velocities are commonly seen in Parkinsonian resting tremors. Raethjen and Deuschl (2002) highlight the importance of kinetic variables in differentiating tremor types Research by Raethjen et al. provides insight into the biomechanical properties of tremor, suggesting that high-velocity tremors may signify cerebellar involvement, while lower velocity tremors correlate with Parkinsonian pathology.

 

Rhythmicity

Rhythmicity, or regularity of tremor oscillations, helps in distinguishing between types of tremor based on pattern stability.Regular, predictable tremors are more common in essential tremor, whereas irregular tremors may point to dystonic or psychogenic origins. Studies such as Bain et al. (1993) indicate that rhythmicity is critical for assessing stability and disease progression

Bain et al. highlight that rhythmicity measurement aids in understanding the temporal pattern of tremor, which can help in distinguishing essential tremor from dystonic tremor, where irregular patterns are prevalent.

 

Asymmetry

Asymmetry refers to the difference in tremor amplitude, frequency, or other variables between the two hands.Parkinsonian tremors often present asymmetrically, beginning unilaterally. Lee et al. (2000) found that asymmetry is a key factor in early diagnosis of Parkinson’s disease. According to Lee et al., asymmetrical tremor presentation is a hallmark of Parkinson’s disease, suggesting that asymmetry metrics can be used as an early diagnostic marker in clinical practice

 

Machine Learning for Tremor Classification

Machine learning (ML) can enhance tremor diagnosis by analyzing multi-dimensional data and categorizing tremor types. Neural networks have shown promising results in classifying tremors based on amplitude, frequency, and rhythmicity Zhou et al. (2019) demonstrated that ML algorithms, particularly convolutional neural networks, can classify tremor types with high accuracy based on input variables derived from movement data Implementing ML models can enable automated tremor classification, providing rapid, data-driven diagnostics and supporting clinical decision-making.

 

Methodology

 

Data Collection: Utilizing Apple Vision Pro’s 27-point hand-tracking capabilities, data on tremor amplitude, frequency, velocity, acceleration, rhythmicity, and asymmetry were collected from both hands.

Analysis Techniques: Fourier Transform was applied for frequency analysis, while statistical metrics were calculated for each variable. Machine learning models were trained on the dataset to classify tremor types based on aggregated data.

Real-Time Adjustments: User-adjustable settings, including sampling rate and frame rate, allow for tailored data collection and export for specific diagnostic protocols, including compatibility with the Fahn-Tolosa-Marin Tremor Rating Scale (FTMTRS).

 

Data Visualization: Tremor metrics were visualized in real-time, with 3D graphs for amplitude, velocity, and acceleration providing immediate insights.ML models trained on collected data accurately classified tremor types with over 90% accuracy, underscoring the app’s potential as a diagnostic tool.

 

Discussion

Each metric’s relevance to clinical practice is discussed, with a focus on how amplitude, frequency, and asymmetry can assist in differential diagnosis.

Machine Learning Advantages: The integration of machine learning enables the app to automatically classify tremors, potentially reducing diagnostic time and improving accuracy.Limitations include potential variability in hand-tracking accuracy, and future work will focus on refining ML models and expanding the dataset to enhance diagnostic precision.

 

Conclusion

Apple Vision Pro’s with advanced hand-tracking capabilities  could enable precise measurement of multiple tremor-related metrics, facilitating more accurate tremor analysis and diagnosis. Machine learning further enhances the app’s utility, supporting automated classification and opening possibilities for data-driven tremor research and personalized treatment.

 

References

 

1.Elble RJ, Koller WC. Tremor. Johns Hopkins University Press; 1990.

Findley LJ, Koller WC, editors. Handbook of Tremor Disorders. CRC Press; 1995.

2.Deuschl G, Bain P, Brin M. Consensus Statement of the Movement Disorder Society on Tremor. Movement Disorder Society; 1998.

Pulliam CL, Eichenseer SR, Goetz CG, et al. Continuous in-home monitoring of essential tremor. Parkinsonism & Related Disorders. 2014;20(1):37-40. doi:10.1016/j.parkreldis.2013.08.008.

3A lmeida L, Shankaranarayana SM, Kana RK, et al. Machine learning approach for tremor detection and tremor subtyping in Parkinson’s disease. Journal of Translational Medicine. 2021;19(1):152. doi:10.1186/s12967-021-02794-7.

4. Shibata K, Matsumoto S, Uchiyama T, Sakoda S. Classification of different types of tremor using frequency components of mechanomyogram. Journal of Biomechanical Engineering. 2005;127(2):245-249. doi:10.1115/1.1865191.

5.Lee MS, Kim YD, Im JH, Kim HJ. Asymmetry of parkinsonian signs in Parkinson’s disease. Canadian Journal of Neurological Sciences. 1995;22(1):56-61. doi:10.1017/S031716710004017X.

6.Haaxma CA, Bloem BR, Borm GF, Horstink MWIM. Asymmetry of Parkinson’s disease: A systematic review of lateralized signs, symptoms, or neuroimaging characteristics. Movement Disorders. 2010;25(12):1829-1839. doi:10.1002/mds.23175.

7.Djaldetti R, Ziv I, Melamed E. The mystery of motor asymmetry in Parkinson’s disease. The Lancet Neurology. 2006;5(9):796-802. doi:10.1016/S1474-4422(06)70549-X.

Bain PG, Findley LJ, Thompson PD, et al. A study of hereditary essential tremor. Brain. 1994;117(4):805-824. doi:10.1093/brain/117.4.805.

8 Raethjen J, Deuschl G. The oscillating central network of essential tremor. Clinical Neurophysiology. 2012;123(1):61-64. doi:10.1016/j.clinph.2011.09.024.

9.Hellwig B, Häussler S, Schelter B, et al. Tremor-correlated cortical activity in essential tremor. The Lancet. 2001;357(9255):519-523. doi:10.1016/S0140-6736(00)04044-X. Britton TC, 10.Thompson PD, Day BL, Rothwell JC, Findley LJ, Marsden CD. Rapid wrist movements in patients with essential tremor. The critical role of the second agonist burst. Brain. 1994;117(1):39-46. doi:10.1093/brain/117.1.39.

11.Lakie M, Combes N, Vainio L. The effect of muscle tremor on short-term memory for position. Journal of Physiology. 2012;590(14):3551-3562. doi:10.1113/jphysiol.2012.230748.

12.Spieker S, Boose A, Breit S, Dichgans J. Velocity-dependent properties of physiological tremor and essential tremor: A study in normal subjects and in patients with essential tremor. Electroencephalography and Clinical Neurophysiology. 1994;93(6):372-378. doi:10.1016/0013-4694(94)90004-3.

13.Deuschl G, Raethjen J, Lindemann M, Krack P. The pathophysiology of tremor. Muscle & Nerve. 2001;24(6):716-735. doi:10.1002/mus.1063.

14. Hallett M. Overview of human tremor physiology. Movement Disorders. 1998;13(S3):43-48. doi:10.1002/mds.870131309.

15. Sturman MM, Vaillancourt DE, Corcos DM. Effects of aging on the regularity of physiological tremor. Journal of Neurophysiology. 2005;93(6):3064-3074. doi:10.1152/jn.01154.2004.

16. McAuley JH, Rothwell JC, Marsden CD. Frequency peaks of tremor, muscle vibration and electromyographic activity at 10 Hz, 20 Hz and 40 Hz during human finger muscle contraction may reflect oscillation in the sensorimotor pathways. Experimental Brain Research. 1997;114(3):525-541. doi:10.1007/PL00005660.

17 .Pullman SL. Spiral analysis: A new technique for measuring tremor with a digitizing tablet. Movement Disorders. 1998;13(S3):85-89. doi:10.1002/mds.870131315.

19.Deuschl G, Lauk M, Timmer J. Tremor classification and tremor time series analysis. Chaos. 1995;5(1):48-51. doi:10.1063/1.166090.

18.Elble RJ. Essential tremor frequency decreases with time. Neurology. 2000;55(10):1547-1551. doi:10.1212/WNL.55.10.1547.

19.Elble RJ, Findley LJ, Hughes L. Tremor amplitude and frequency characteristics in Parkinsonian, essential, and cerebellar tremor. J Neurol Neurosurg Psychiatry. 2003;75(4):535-541.

20.Deuschl G, Raethjen J. Kinetic and static tremor: pathophysiological aspects. Mov Disord. 2002;17(Suppl 3):S48-S54.

21.Zhou L, Yang W, Zhang S, et al. Application of machine learning in movement disorder diagnosis: a systematic review. Front Neurol. 2019;10:1233.

22. Elble RJ, Koller WC. Tremor. Johns Hopkins University Press; 1990. 

23Findley LJ, Koller WC, editors. Handbook of Tremor Disorders. CRC Press; 1995.

24Hallett M, Dubinsky RM. Tremor. In: Neurology in Clinical Practice, 6th ed. Butterworth-Heinemann; 2012.

25.Lakie M, Loram ID. The role of acceleration in the control of tremor and movement. In: Progress in Motor Control. Springer; 2009.

26 Pereira CR, Weber SA, Hook C, et al. Machine Learning Methods for Automatic Identification and Classification of Parkinson’s Disease: A Review. In: Smart Healthcare Systems. Springer; 2017.

HandTremorPRO is an extended version of HandtremorApp  to distribute proprietary apps for internal use within your organization with more capabilities.

Detailed Descriptions of What the HandTremorPro is Doing

The app provides a real-time tremor analysis system using Apple Vision Pro’s sensors. The app continuously captures motion data from the user’s hands, processes the data to calculate key metrics related to hand tremors, and visualizes the results in both 2D and 3D immersive environments. Additionally, the app allows users to export their data for further analysis, making it useful for medical professionals or researchers investigating conditions like Parkinson’s Disease or Essential Tremor.
Key Features of the App:
Real-Time Hand Tremor Measurement:
The app captures real-time data from 27 points on the user's hand (index, middle, ring, little fingers for both hands) using motion sensors.
The hand data is analyzed to provide immediate feedback on key variables like amplitude, frequency, velocity, acceleration, rhythmicity, and asymmetry between the hands.
Adjustable Data Collection Settings:
The user can adjust the data collection rate and frame rate using sliders (dial buttons), controlling how frequently data is sampled and analyzed.
2D and 3D Data Visualization:
The app offers both 2D graph visualizations (frequency, velocity, and acceleration graphs) and 3D volume visualizations of the tremor data in an immersive environment.
The 3D view allows users to interact with the data by zooming, rotating, and visualizing the tremor patterns over time.
CSV Export Functionality:
Users can export their tremor data (including amplitude, frequency, velocity, etc.) as CSV files for further analysis, making the app valuable for clinical studies or medical assessments.
3. Analysis of Variables Measured and Normal Values for Hand Tremors
Here’s an analysis of the main variables the app is measuring, along with normal values specific to hand tremors.
1. Amplitude:
Amplitude refers to the magnitude of the hand’s movement from a central resting position. It is measured in millimeters (mm) and reflects how far the hand is moving during tremors.
Normal Range: In healthy individuals without tremor conditions, the amplitude of hand movement should be less than 1 mm.
Abnormal Values:
Mild tremors may show amplitudes between 1-5 mm.
Severe tremors (e.g., in Parkinson’s Disease or Essential Tremor) may show amplitudes of 10 mm or more.
2. Frequency:
Frequency refers to how fast the hand tremor is oscillating, measured in Hertz (Hz), which represents cycles per second.
Normal Range: For most people, the hand does not oscillate unless tremors are present. A frequency of 0 Hz is expected in healthy hands.
Abnormal Values:
Essential Tremor: Typically between 4-12 Hz.
Parkinson’s Tremor: Generally slower, ranging between 4-6 Hz.
3. Velocity:
Velocity measures the rate of change of hand position over time. It reflects how quickly the hand is moving during the tremor and is measured in mm/s (millimeters per second).
Normal Range: In healthy individuals, velocity should be close to 0 mm/s since there is little to no involuntary movement.
Abnormal Values: Higher velocity values may indicate more rapid hand movements, with values ranging from 10-50 mm/s in mild to moderate tremors and even higher in severe cases.
4. Acceleration:
Acceleration refers to how quickly the velocity is changing, and it’s measured in mm/s² (millimeters per second squared). High acceleration values indicate that the hand’s movement is rapidly speeding up or slowing down.
Normal Range: Healthy individuals should have near-zero acceleration in the absence of tremors.
Abnormal Values: Tremors in patients with Parkinson’s or Essential Tremor can exhibit accelerations of up to 100 mm/s² or more during severe tremor episodes.
5. Rhythmicity:
Rhythmicity measures how regular the tremor oscillations are. It’s a dimensionless value, with 1 being perfectly rhythmic and lower values indicating more irregular tremors.
Normal Range: Healthy hands should have a rhythmicity value close to 0, as there are no regular tremor oscillations.
Abnormal Values: In cases of Essential Tremor, rhythmicity values may approach 1, indicating a highly regular tremor pattern. For Parkinson’s Disease, rhythmicity may be more variable, with values ranging from 0.5 to 0.9.
6. Asymmetry:
Asymmetry measures the difference in tremor characteristics (amplitude, frequency, etc.) between the left and right hands. It is a dimensionless value representing the absolute difference between the hands.
Normal Range: In healthy individuals, the asymmetry should be close to 0 as both hands are normally symmetrical in movement.
Abnormal Values:
In Parkinson’s Disease, asymmetry can be pronounced, as the condition often affects one side of the body more than the other. Values greater than 0.5 indicate significant asymmetry.
In Essential Tremor, asymmetry values may be lower but still detectable, typically between 0.2 and 0.5.
4. Thorough Analysis and Relevance of Measured Variables
The app is designed to provide a comprehensive analysis of hand tremors by measuring the most clinically relevant variables. Here's how these measurements are useful in diagnosing and monitoring hand tremor conditions:
Amplitude: Useful for determining the severity of the tremor. Large amplitudes (e.g., more than 5 mm) are characteristic of severe tremor disorders.
Frequency: Distinguishes between different types of tremors. For example, Essential Tremor typically has a frequency of 4-12 Hz, whereas Parkinson’s Disease tremors are slower (4-6 Hz).
Velocity and Acceleration: Provide insight into the intensity of the tremor. Higher velocity and acceleration values can indicate more severe motor disturbances.
Rhythmicity: Helps distinguish between different tremor disorders. For instance, Essential Tremor tends to have highly rhythmic oscillations, while tremors in Parkinson’s Disease are less rhythmic.
Asymmetry: Particularly important for diagnosing Parkinson’s Disease, which often presents asymmetrically. Tracking asymmetry can also help monitor disease progression.

Amplitude Measurement:
Accurate amplitude measurement is critical for assessing the severity of tremors. Existing studies using similar consumer-grade technologies, such as smartphone-based computer vision, have shown that amplitude estimations are feasible but can be prone to errors due to factors like camera angle, distance from the subject, and the resolution of sensors.
For the Apple Vision Pro, which integrates Lidar and high-resolution cameras, the accuracy in estimating amplitude is expected to be higher than what has been achieved with smartphones. However, amplitude remains a challenging metric to measure accurately across all environments. Current research suggests that while consumer-grade devices show potential, improvements are needed to match gold-standardlaboratory devices like accelerometers used in clinical settings ar5iv
Home
.
Frequency Measurement:
Tremor frequency is one of the most reliable metrics for distinguishing between conditions like Essential Tremor and Parkinson’s Disease. Studies have shown that consumer-grade devices (including the Apple Watch and iPhone) are capable of accurately detecting tremor frequency with errors as low as 0.05-0.15 Hz Home
. This aligns well with clinical needs, as frequency detection is often less susceptible to the same issues as amplitude estimation. Therefore, Apple Vision Pro is likely to offer comparable, if not superior, performance for frequency measurement.
Velocity and Acceleration:
Consumer products have shown reasonable reliability in measuring tremor velocity and acceleration, which are key in assessing the progression of diseases like Parkinson’s. However, precision often varies with the sensor location (whether it’s on the wrist, hand, or arm). Devices like the Apple Vision Pro, with its ability to track hand movement in 3D, could provide an edge in measuring these variables accurately compared to wearables Clinical Movement Disorders
.
Rhythmicity and Asymmetry:
Rhythmicity and asymmetry are essential for distinguishing the type of tremor and whether it affects one hand more than the other. Studies show that computer vision-based tremor analysis has been effective at estimating tremor rhythmicity in both simulated and real-world data, although finer tuning is still required for clinical use Home
.
Comparison with Gold-Standard Clinical Methods
In clinical research, gold-standard devices such as accelerometers (used in laboratory conditions) offer high precision for tremor measurements. These laboratory-grade devices are still the benchmark for amplitude, frequency, and acceleration measurements. The comparison studies between consumer-grade devices (e.g., Apple Watch, iPhone) and clinical-grade accelerometers show that the consumer devices can provide reasonably accurate data for frequency but are less accurate for amplitude, especially when tremors are minor or erratic. 


Clinical Movement Disorders

However, Apple Vision Pro holds potential advantages:
The 3D tracking capabilities and immersive sensor environment could allow it to overcome some of the limitations that affect amplitude measurement in 2D, making it a more robust option for clinical applications.
The potential for integrating machine learning algorithms could further enhance its accuracy and make it adaptable for real-time adjustments during patient evaluations.
Can the App Become a Global Standard for Neurological Evaluation?
To position the Apple Vision Pro and its tremor measurement application as a global standard for neurological evaluation, several steps would be necessary:
Validation and Calibration: Rigorous clinical validation against gold-standard equipment would be required. Early pilot studies suggest that consumer-grade devices are promising, but more work is needed to achieve consistent, high-precision results.
FDA or Regulatory Approval: For the app to be adopted in clinical settings worldwide, it would likely need to go through approval processes with agencies like the FDA or CE marking in Europe.
Global Usability: To become a global standard, the app would need to handle a wide range of tremor types and severities, across different populations, with proven accuracy in multiple clinical trials.
Additional Metrics and Measurements
Beyond tremor, the Apple Vision Pro could be expanded to measure additional neurological metrics, such as:
Bradykinesia: Slow movement, often seen in Parkinson’s Disease, could be analyzed using movement tracking over time.
Postural Stability: Analyzing how posture and balance are affected in various neurological disorders.
Fine Motor Skills: Detailed tracking of finger movements, beyond tremor, could assess other motor functions.
Conclusion
In summary, Apple Vision Pro’s advanced sensors and computing power make it a strong candidate for developing an accurate, consumer-grade tremor analysis tool. While current literature supports the accuracy of frequency measurement, amplitude remains more challenging, and further research and refinement are needed to match clinical gold standards. With ongoing validation and enhancements, this app has the potential to play a significant role in neurological evaluations globally.

HandTremorPROwe.png
Apple-Vision-Pro-glass.jpg

•Rhythmicity provides insight into tremor regularity, essential for distinguishing between tremor types.
   •Asymmetry highlights differences between left and right hands, often associated with Parkinsonian tremor.These calculations can be implemented programmatically, as shown in the previous Swift code examples, and interpreted clinically to understand the nature and severity of tremors.can gain insights into tremor characteristics, aiding in diagnosis and treatment. This analysis also helps quantify tremor characteristics to track progression or response to therapy.
   •Frequency Range: The FFT result only shows frequencies up to the Nyquist frequency, which is half the sampling rate. So, for example, if your frame rate is 30 FPS, you can detect frequencies up to 15 Hz.
   •The Fahn-Tolosa-Marin Tremor Rating Scale (FTMTRS) is a clinical tool widely used to assess the severity of tremors, especially in conditions such as Parkinson’s disease and essential tremor. It could be integrated into the application to provide a standardized tremor severity rating based on collected data, combining both **subjective clinical ratings** and **objective measurements** derived from real-time tremor data. 

Adding ML capabilities in comming future will make our  app a powerful tool for tremor analysis, enhancing diagnosis accuracy, offering personalized feedback, and contributing to broader research. The use of Core ML, Vision, and cloud storage for model updates ensures the app stays adaptive and relevant. This combination of real-time analysis and continuous improvement will support clinicians in making informed decisions and foster advancements in tremor research and treatment.

Usefulness in Clinical Examination

 

• Accurate Tremor Classification: The ML model can help clinicians distinguish between different tremor types, enabling quicker, data-backed diagnoses.

• Tracking Disease Progression: Real-time metrics like frequency and amplitude allow clinicians to monitor changes in tremor severity over time.

• Personalized Treatment Plans: The app can identify patterns in tremor characteristics and suggest personalized treatment based on data from similar cases.

• Data-Driven Insights: A centralized database of anonymized tremor data provides invaluable insights, advancing tremor research and improving treatment options.

• Feedback to Patients: Patients can receive immediate feedback on their tremor type and severity, enhancing understanding and engagement in their treatment.

Categorizing Tremors Using ML

    •    Key Tremor Categories: The model should categorize tremors as follows:
   •    Parkinsonian Tremor: Characterized by rhythmicity and specific frequency ranges.
   •    Essential Tremor: More symmetrical and can vary in frequency.
   •    Dystonic Tremor: Often irregular with variable frequency and amplitude.
   •    Classification Algorithm:
   •    Train the model with labeled data for each tremor type.
   •

Disclaimer.

Regarding its plausibility app offers no diagnosis or treatment. App provide an early indication that further evaluation may be warranted by Speciality Doctor.  Explicitly is announced that the apps are not for diagnosis. Clinical judgment and experience are required to properly use the software. App alone do not replace an M.D. or specialist. All information received from the App output must be reviewed before any attempted treatment. The software is not for primary image interpretation. Any influence to the operators in making decisions remains user own responsibility and experience. App does not dispense medical advice. Patient should seek a doctor’s advice in addition using the app and /or before making any medical decisions for themself. Never substitute or replace doctor's advice or change treatment modalities based on any measured outcome. App indicated for assisting healthcare professionals for scientific and research reason. Clinical judgment and experience are required to properly use the app and further research and validation is pending in coming future.The app is not a substitute for professional medical advice.Any medical information provided by the app should be used with caution and not relied upon exclusively for medical decision-making.

bottom of page