Subscribe to Tendercast, Data Science Statistics, Data Preparation, Machine Learning with Python courses to receive complimentary 4-hour 1:1 use case support

USING ARTIFICIAL INTELLIGENCE TO BETTER UNDERSTAND PATIENT RISK FOR ALZHEIMER'S

Alzheimer's disease prevalence, incidence and affected patients and families continues to increase.

 

Historical research suggests patients at risk for Alzheimer's may occur and be known earlier in the patient journey. 

Patient level data exists for many years which can be minded using machine learning and artificial intelligence tools.  Types of data include electronic medical records, chart data, patient level data, real world data and public available incidence and diagnosis rates.  Longitudinal, ten-year plus (over 10 years) data from USA and Japan is explored to understand the Risk Factors and generate a risk score for patients.  

  

What is known?

USA

Japan

Risk factors:
Age
Family history
Genetics
Head injury
Heart-Head
Healthy aging
Cost
Features:
>65yo, 2x/5year
Family ill
Gene influence
Hx injury
Hx vascular disease
Diet
Cost/patient

5.7MM

8-12%

$50k/year

4.6MM

12-23%

$120K/year

What is not known?

• Cause is unknown
• When and why symptoms begin
• Why amyloid, tau proteins build up
• What other risk factors?
• Any unique features of patients?

How can Machine Learning and Artificial Intelligence help patients?

• Risk factors/features:  Identify features of diagnosed Alzheimer’s and dementia patients with unstructured real world data, electronic medical records and chart data.  Determine and validate the model to predict if patients developed the disease.
• Risk score:  Apply the risk scoring model to patient level data where data exists or DRG coding is licensed and applied. 
• Patients:  Enroll patients identified in clinical trials and/or medication treatment programs.

 

 

(Illustrative, presented at Global Big Data Artificial Intelligence Conference, May 2021)

Availability