Machine learning AI Facial and voice Analysis applicaTion to appraIse fatiGUe in chronic liver disease patiEnts (mAI FATIGUE)
试验速览
- 阶段
- 不适用
- 状态
- 招募中
- 入组人数
- 100
- 试验地点
- 3
- 主要终点
- The primary analysis will be the evaluation of the correlation between each of the parameters assessed for fatigue score by the AI interactive tool and each PRO tool, the overall study visits in which both PROs and AI tool were assessed. The correlation will be calculated for all these visits combined as well as over both groups. In addition, both groups will be analyzed separately. The Pearson correlation coefficient will be used to assess correlation. Further details will be defined in the Statistical Analysis Plan (SAP). Primary endpoint outcomes will be assessed based on the PPS population.
研究概览
简要总结
Despite the potential utility of machine learning models in objectively identifying and assessing fatigue severity, few studies evaluating the benefits of machine learning technology in assessing fatigue in patients with chronic diseases have been published. This study is expected to add to the body of knowledge the insights of using a novel technology that may improve the identification of fatigue in patients with CLD, leading to better symptom management.
A total of 100 participants who meet the eligibility criteria and provide voluntary informed consent will be enrolled in this study. The study population comprises two cohorts as below:
Cohort A: 50 adults without CLD and self-reporting not being fatigued, as assessed by the Patient Global Impression of Severity (PGI-S) questionnaire with a response of “None†at Screening (Visit 1).
Cohort B: 50 adults with CLD and self-reported moderate to severe fatigue, as assessed by the PGI-S questionnaire with a response of “moderateâ€, “severeâ€, or “very severe†at Screening (Visit 1).
研究设计
- 研究类型
- Interventional
- 分配方式
- Na
- 盲法
- None
入排标准
- 年龄范围
- 18.00 Year(s) 至 65.00 Year(s)（—）
- 性别
- All
入选标准
- •For Cohort A and Cohort B
- •Adult participants aged 18 to 65 years (both inclusive) who are willing to provide written informed consent, including consent for AV recording via a mobile application, and who agree to adhere to all study procedures.
- •Only For Cohort A
- •Adult subjects who do not have CLD, as confirmed by a clinician, and who report no fatigue, as assessed by the Patient Global Impression of Severity (PGI-S) with a response of “none.â€.
- •Only for Cohort B:
- •Participants diagnosed with CLD who are receiving standard care for CLD and self-reported moderate to severe fatigue, as assessed by the Patient Global Impression of Severity (PGI-S) with responses of “moderate,†“severe,†or “very severeâ€.
排除标准
- •Participants will be excluded if they do not have access to a mobile device or if their mobile device does not meet the specifications required to use the Blueskeye AI application, as outlined in the user manual.
- •This will be verified by trained site staff.
- •Participants receiving treatment for fatigue from the last 3 weeks prior to screening.
- •Participants with a known history of Parkinson’s disease.
- •Participants with a history of alcohol or drug abuse within the last three months prior to screening or those currently abusing alcohol (greaterthan 7 drinks per week for females and graterthan 14 drinks per week for males) or drugs, as determined by self-report or medical records.
- •Women of childbearing potential will be excluded if they are currently pregnant, planning to become pregnant during the study, or breastfeeding.
- •Participants must agree to use effective contraception methods during the study.
- •At screening, urine pregnancy tests will be conducted for Women of childbearing potential to confirm the eligibility.
- •Participants who are unable to speak or read English.
- •Participants who have undergone surgery within the past three months or have planned surgery within the next month from the screening date.
- •Participants with any comorbidity or concurrent medical condition (other than CLD or its related comorbidities), including but not limited to depression, that, at the discretion of the Investigator, might prevent adherence to the trial procedures.
结局指标
主要结局
The primary analysis will be the evaluation of the correlation between each of the parameters assessed for fatigue score by the AI interactive tool and each PRO tool, the overall study visits in which both PROs and AI tool were assessed. The correlation will be calculated for all these visits combined as well as over both groups. In addition, both groups will be analyzed separately. The Pearson correlation coefficient will be used to assess correlation. Further details will be defined in the Statistical Analysis Plan (SAP). Primary endpoint outcomes will be assessed based on the PPS population.
时间窗: Visit 1(day -2 to day 1) to Visit 10 (Day 22)
As a pre-processing step, we will normalize the raw fatigue scores from PROs to the scale 0-1 using the min-max normalization technique. The maximum value of each PRO is the highest value that can be scored (1), and the minimum value of each PRO is the lowest value that can be scored (0).
时间窗: Visit 1(day -2 to day 1) to Visit 10 (Day 22)
次要结局
- Secondary endpoint outcomes will be evaluated based on the FAS population.(Out of N equal to metrics, those that will have the highest correlation with PRO fatigue scores, the RMSE ranges will be estimated from the linear correlation analysis of combined cohorts from the Primary Efficacy outcomes.)
研究者
Dr Kinjalkumar Shah
Abbott EPD