Abstract
Background Widespread early dementia detection could drastically increase clinical trial candidates and enable early interventions. Since the Clock Drawing Test (CDT) can be potentially used for diagnosing dementia related diseases, it can be leveraged to devise a computer-aided screening tool.
Objective This work aims to develop an online screening tool by leveraging Artificial Intelligence and the CDT.
Methods Images of an analog clock drawn by 3, 263 cognitively intact and 160 impaired subjects were used. First, we processed the images from the CDT by a deep learning algorithm to obtain dementia scores. Then, individuals were classified as belonging to either category by combining CDT image scores with the participant’s age.
Results We have evaluated the performance of the developed models by applying 5-fold cross validation on 20% of the dataset. The deep learning model generates dementia scores for the CDT images with an Area Under the ROC Curve (AUC) of 81.3% ± 4.3%. A composite logistic regression model using age and the generated dementia scores, yielded an average AUC and average weighted F1 score of 92% ± 0.8% and 94.4% ± 0.7%, respectively.
Discussion CDT images were subjected to distortion consistent with an image drawn on paper and photographed by a cell phone. The model offers a cost-effective and easily deployable mechanism for detecting cognitive impairment online, without the need to visit a clinic.
Competing Interest Statement
The authors have declared no competing interest.
Clinical Trial
NCT00005121
Funding Statement
The research was partially supported by the NSF, by the ONR, by the NIH, by the DOE, by the Framingham Heart Studys National Heart, Lung, and Blood Institute contract, by the NIH National Institute on Aging, by the Alzheimers Association, and by Pfizer.
Author Declarations
I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
The entire dataset for all participants was anonymized prior to analysis. All participants have provided written informed consent and study protocols and consent forms were approved by the Boston University Medical Campus Institutional Review Board.
All necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Footnotes
Conflicts of Interest: Rhoda Au is a scientific advisor to Signant Health, consultant to Biogen, and has received grant support from Pfizer; she states no conflict of interest with the present work. There is no declaration from other authors.
Data Availability
The dataset is coming from Framingham heart study and is not available to public.