【Lancet】Fudan Brain-like Research Institute and Huashan Hospital Team: A new model accurately predicts the risk of dementia in individuals in the next 5-10 years

This article is the original of the translational medicine network, please indicate the source when reprinting

Author: Lily

Dementia is a neurodegenerative disease, the syndrome involves a variety of advanced cortical disorders, seriously affecting people's thinking, emotions and behavior, has become one of the main diseases of death and disability in
the elderly population.

Currently, the disease affects about 55 million people worldwide
.

It is predicted that the number of people with dementia worldwide will increase to 153 million in 2050, about 2.
7 times
the 2019 value.

To make matters worse, the onset and development of dementia is more insidious, and its course can last for decades, showing long-term progressive development
.

In recent years, the incidence of dementia has become younger, and related pathological changes can occur 20 years
before the onset of the disease.

Often, patients have missed the optimal window of treatment when they reach a clinical diagnosis
.

Therefore, we urgently need to establish the right way to identify people at high risk of dementia so that they can effectively predict, diagnose and intervene
in the early stage of dementia.

On September 23rd, Professor Feng Jianfeng of the Institute of Brain-like Intelligence Science and Technology of Fudan University / Cheng Wei Young Researcher Team, together with the clinical research team of Professor Yu Jintai of Huashan Hospital affiliated to Fudan University, published the latest research results
in the sub-journal eClinicalMedicine of The Lancet.

Based on artificial intelligence algorithms and biomedical big data, the research team developed a new dementia risk prediction model (called UKB-DRP); The model provides prospective predictions of the risk of all-cause dementia and its primary subtype (Alzheimer's disease) – enabling accurate predictions of an individual's risk of developing dementia over the next five, ten or even longer years
.

population sample data for this study were from the UK Biobank (UKB) and involved the follow-up of 425,000 people aged 40-69 years with non-dementia (median follow-up time of 11.

The ten predictors include: age, apolipoprotein E (ApoE) gene, length of cognitive pairing test, percentage of leg fat, number of medications, length of cognitive response test, peak expiratory flow, maternal age of death, chronic disease, and mean red blood cell volume
.

These accessibility predictors can make relatively accurate predictions of the risk of developing dementia and Alzheimer's disease (AD) over a period of five years, ten years, or even longer to identify high-risk individuals
in the general population.

Predictor importance ranking and model incorporating predictors' predictive efficacy for dementia

Among them, the model evaluation index AUC (area under the curve) of all-cause dementia is 0.

UKB-DRP predictive models for dementia are for five, ten years and longer

All due to the predictive efficacy of dementia and Alzheimer's disease

UKB-DRP model advantages

 01

Technically, the study had the advantage that the UKB-DRP model was established based on a large prospective cohort containing more than 500,000 participants with a clinical record
of at least 10 years of follow-up.

The determination of predictors was carefully and carefully selected from the comprehensive clinical feature space in a data-driven manner – only 10 predictors
were selected in the end.

The UKB-DRP model was built using a highly efficient gradient boosting decision tree, one of the most powerful machine learning (ML) techniques: it works with an integrated strategy and is ideal for large data sets
with large sample sizes and high feature space.

Clinically, traditional predictive models are mostly based on complex neuropsychological testing, expensive whole genome sequencing (WGS), invasive lumbar puncture, or positron emission tomography (PET) imaging
.

The UKB-DRP model is based entirely on easily accessible predictors; These factors can be collected from rapid questionnaires, physical measurements and simple blood tests, and the UKB-DRP dementia prediction model can be widely used at all levels of early screening in
healthcare.

About the research team

 02

Professor Feng Jianfeng, President of the Institute of Brain-like Intelligence Science and Technology, said: "Based on data-driven thinking, this research processes and mines a large amount of health information of middle-aged and elderly people through artificial intelligence algorithms, and constructs an early prediction model of dementia driven by medical big data and intelligent models, which provides a new theoretical basis
for improving the early risk discrimination ability of neurodegenerative diseases and preventing disease occurrence.

In the next step, we will continue to focus on the major health issue of brain health in the elderly, and conduct research on a variety of senile brain health killer diseases such as cardiovascular and cerebrovascular diseases, Parkinson's, etc.
, so as to provide scientific and accurate identification methods and practical guidance for the early identification of people at risk of different diseases
.
"

Professor Feng Jianfeng, Cheng Wei Young Researcher of Fudan University Institute of Brain-like Intelligence Science and Technology, and Professor Yu Jintai of Huashan Hospital Affiliated to Fudan University are the corresponding authors of the article, and You Jia, a postdoctoral fellow of the Institute of Brain-like Diseases of Fudan University, and Zhang Yaru, a doctoral candidate of Huashan Hospital Affiliated to Fudan University, are the first authors
.

The research has been supported by the National Natural Science Foundation of China, the National Key Research and Development Program, and the Shanghai Major Special Project
.

Resources:

This article is intended to introduce medical research advances and cannot be used as a reference for
treatment options.

Recommended Live/Events

09/27 14:00-17:30 Shanghai

Innovative technology to facilitate the summit forum on accurate diagnosis and treatment of tumors

Scan the code to participate for free

November 01-02 09:00-17:30 Chongqing

The 1st Southwest Single Cell Omics Technology Application Forum

Scan the code to participate for free

November 25-27 09:00-17:30 Shanghai

The 4th Shanghai International Cancer Congress

Scan the code to participate