Help Build an A.I. Model to Predict Myasthenia Gravis Symptom Patterns and Flares

Myasthenia Gravis Clinical Trial

Official title:

A Digital Health Trial That Assesses Participant-driven Data Collection Using Smartphone Modules to Characterize Myasthenia Gravis Symptoms and Develop an A.I. Model to Predict Flares

Clinical Trial Details — Status: Completed

Administrative data

• NCT number: NCT04590716
• Other study ID #: DOC-005-2020
• Status: Completed
• Start date: October 2, 2020
• Est. completion date: July 26, 2021

Study information

• Verified date: July 2021
• Source: doc.ai inc
• Contact: n/a
• Is FDA regulated: No
• Study type: Observational

Clinical Trial Summary

There are limited objective measurements of MG symptoms as well as a dearth of data at a granular level of MG (myasthenia gravis) symptoms and triggers occurring longitudinally.

This study is designed to use the strengths of mobile smartphones which enable participant-driven real time capture of data manually and through augmented sensors such as video and audio, in order to better characterize MG symptoms and flares.

The study aims to enroll approximately 200 participants for approximately 9 months until analyzable data is available from at least 100 participants. Participants will complete in-app surveys for 3 months with, audiovisual recording of symptoms. This will take approximately 35 minutes per week after the initial survey.

Description:

Using their smartphones, potential participants will download the doc.ai research mobile app. There will be a web pre-screening link where potential participants will self-screen to see if they meet the basic eligibility criteria for this study. The recruitment tool for this trial is developed for diversity, fairness, and inclusion. With the aim to ensure diversity in the demographics of the trial to better understand the health needs of different populations. So, while some interested potential participants do qualify, they may not be invited into the trial due to these diversity requirements.

Patients with myasthenia gravis (MG) who meet the inclusion criteria will be invited to join this digital health trial. Participants will sign the e-consent and self-enroll into the study. Once their eligibility is confirmed by the study team (to ensure eligibility criteria and validity of participant i.e. eliminate robo sign-ins) they will be asked to take a selfie, provide documented proof of MG diagnosis, respond to a series of survey questions regarding their demographics, current health, medical history, and other MG related information.

Enrolled participants will have a daily brief check-in, 2 weekly check-ins and a weekly check-in which will include an audiovisual check-in, and will maintain an audiovisual diary to keep track of their symptoms, connect data, record their voice (to detect vocal symptoms: weakness, nasality) and take videos of their face (to detect facial symptoms: ocular, mouth droop) on a daily to weekly basis through the various data collecting modules in the doc.ai research app for the duration of their study participation. doc.ai's digital health trial platform will be leveraged to collect this data.

The study aims to enroll approximately 200 participants in approximately 9 months. It is expected that a minimum of 100 participants will be included in the final analysis as at any given time there will be a lag between potential participants expressing interest in the study, their eligibility being assessed by the PI, and participants completing all required study procedures.

At the end of their participation, participants will be asked to complete a questionnaire. After the participant has completed their final survey, they will be able to connect to a link redeemable as an Amazon.com gift card worth $250. All participants will also receive an end-of-trial-summary of the data that they had collected during the study. No medical advice or direction will be given based on this study.

In addition, in the final survey participants will be asked if they would be willing to complete a usability interview after their participation in this trial has ended. This subset of participants invited to be part of a follow-up usability interview will include those who complete all study required procedures and some who may not have completed all study required procedures, in order to assess usability experience of the app for the duration of their participation. Participants will be contacted at their end of their period of participation until a total of 10-15 participants successfully complete the usability interview. Participants who successfully complete the usability interview will receive a link for a $50 in Amazon.com gift card via the app.

For this study the data and, audio and video recordings will be captured directly on the doc.ai research app and securely stored on a HIPAA compliant cloud provider (Google Cloud Platform).

This data will be used to understand the patterns of symptoms and triggers in order to better characterize factors such as the length and timing of flares and any unique symptom patterns in order to create more objective measures of MG symptoms. Ultimately this data would be used to build a machine learning model that could predict MG symptom flares.

Primary Objective:

Use a collection of digital health modules on the smartphone to collect myasthenia gravis (MG) symptoms and triggers to better characterize symptom patterns and flares.

Secondary Objective:

Use the data collected to develop an A.I. model to detect and/or predict symptom flares.

Recruitment information / eligibility

• Status: Completed
• Enrollment: 113
• Est. completion date: July 26, 2021
• Est. primary completion date: July 26, 2021
• Accepts healthy volunteers: No
• Gender: All
• Age group: 18 Years and older

Eligibility

Inclusion Criteria:

1. Must have a documented diagnosis of Myasthenia Gravis

2. Must have ocular (eye drooping) and/or bulbar (speech) symptoms

3. Must be over the age of 18

4. Must reside in the US for the duration of the study

5. Must be able to read, understand, and write in English

6. Must have a smartphone supported by the doc.ai research app (iOS and Android)

Exclusion Criteria:

None

Related Conditions & MeSH terms

• Muscle Weakness
• Myasthenia Gravis

Intervention

Other:
• Data Collection
This is a non-interventional study conducted on the participant's smartphones to record MG related symptoms and conditions.

Locations

Country	Name	City	State
United States	Doc.Ai Mobile Based	Palo Alto	California

Sponsors (2)

Lead Sponsor	Collaborator
doc.ai inc	UCB Biopharma SRL

Country where clinical trial is conducted

United States, 

References & Publications (10)

Borza D, Darabant AS, Danescu R. Real-Time Detection and Measurement of Eye Features from Color Images. Sensors (Basel). 2016 Jul 16;16(7). pii: E1105. doi: 10.3390/s16071105. — View Citation

Duffy, JR: Motor Speech Disorders. Substrates, Differential Diagnosis and Management (1st ed). St. Louis, 1995, Mosby.

Duffy, JR: Motor Speech Disorders. Substrates, Differential Diagnosis and Management (2nd ed). New York, 2005, Elsevier Health Sciences.

Hegde S, Shetty S, Rai S, Dodderi T. A Survey on Machine Learning Approaches for Automatic Detection of Voice Disorders. J Voice. 2019 Nov;33(6):947.e11-947.e33. doi: 10.1016/j.jvoice.2018.07.014. Epub 2018 Oct 11. Review. — View Citation

Kang H. The prevention and handling of the missing data. Korean J Anesthesiol. 2013 May;64(5):402-6. doi: 10.4097/kjae.2013.64.5.402. Epub 2013 May 24. — View Citation

Kent RD, Kent JF, Rosenbek JC. Maximum performance tests of speech production. J Speech Hear Disord. 1987 Nov;52(4):367-87. Review. — View Citation

Konopka BM, Lwow F, Owczarz M, Laczmanski L. Exploratory data analysis of a clinical study group: Development of a procedure for exploring multidimensional data. PLoS One. 2018 Aug 23;13(8):e0201950. doi: 10.1371/journal.pone.0201950. eCollection 2018. — View Citation

Panayotov V., Chen G., Povey D., Khudanpur S. (2015). Librispeech: an ASR corpus based on public domain audio books, in Proceedings of the ICASSP (South Brisbane, QLD:), 5206-5210

T. Baltrusaitis, A. Zadeh, Y. C. Lim and L. Morency,

Zhou ZR, Wang WW, Li Y, Jin KR, Wang XY, Wang ZW, Chen YS, Wang SJ, Hu J, Zhang HN, Huang P, Zhao GZ, Chen XX, Li B, Zhang TS. In-depth mining of clinical data: the construction of clinical prediction model with R. Ann Transl Med. 2019 Dec;7(23):796. doi: — View Citation

Outcome

Type	Measure	Description	Time frame	Safety issue
Primary	Audiovisual recording of voice exercises to detect patterns and changes in voice and facial symptoms	participants to complete the audio and visual data modules designed to capture patient MG symptoms (especially ocular and voice).; e.g; Vocal e.g.: Say "papapapa" for 4 seconds; Say "tatatatata" for 4 seconds; Say "kakakaka" 4 seconds; Say "mamamama" 4 seconds; Say "papapapa" 4 seconds; Say "buttercup, buttercup, buttercup" 4 seconds; Say "aaaahhh" and hold it as long as you can; Counting e.g.: Look straight at the camera for 4 seconds; Count as precisely as possible from 1 to 25 while looking up; Look straight at the camera for 4 seconds; The recordings will be used to detect change from baseline and any patterns that may occur. This will be used to analyze where and if different features are linked to see if a single or combined effect of the features is connected to flare frequency and/or severity.	After enrollment, 3 months with in-app twice a week audiovisual recording of symptoms.
Secondary	Completion of MG-Quality of Life assessment	Participants complete MG activities of Daily living and MG-Quality of Life assessments weekly. This assessment has been adapted from www.myasthenia.org/healthprofessionals/educationalmaterials.aspx	Approximately 10 minutes each week for 3 months.

External Links

•   "Myasthenia gravis - Genetics Home Reference - NIH."
•   Assessment Instruments for Patients with Myasthenia Gravis (MG)