Deep Learning in Predicting Aging-Related Diseases: History
Please note this is an old version of this entry, which may differ significantly from the current revision.

Aging refers to progressive physiological changes in a cell, an organ, or the whole body of an individual, over time. Aging-related diseases are highly prevalent and could impact an individual’s physical health. Recently, artificial intelligence (AI) methods have been used to predict aging-related diseases and issues, aiding clinical providers in decision-making based on patient’s medical records. Deep learning (DL), as one of the most recent generations of AI technologies, has embraced rapid progress in the early prediction and classification of aging-related issues.

  • aging
  • deep learning

1. Introduction

Aging refers to the persistent decline in the age-specific fitness due to internal physiological changes, anatomical, and immunological changes in living beings [1]. Physiological changes are usually associated with a wide range of disorders, including neurodegenerative, cardiovascular, respiratory, and eye-related diseases [2]. Efforts to early detection of diseases in aging population and related therapeutics have now become a hot topic of research. Cutting edge modern computing technologies in Artificial Intelligence (AI) such as Deep Learning (DL) are being recently applied to improve understanding of aging-related diseased conditions and have been engaged to assist clinicians and healthcare professionals for improved decision-making [3]. DL-based algorithms indicate great potential in extracting features and learning patterns from complex and heterogeneous medical data pertaining to an individual’s health status. Such data may involve medical images, such as scans from imaging devices; genomic data relating human genes to diseases; smart sensor data to detect medical conditions and their effects; data from electronic health records (EHRs); and the time series data from electrograms [4]. DL methods aid in learning these data representations to predict diseased states related to aging [4]. DL transforms the data through layers of nonlinear computational processing units, providing knowledge discovery from the complex data. In recent years, DL algorithms have indicated superior performance in many data-rich application scenarios relating to the healthcare of the aging population.

2. Deep Learning in Predicting Aging-Related Diseases

2.1. Age-related macular degeneration (AMD)

Age-related macular degeneration (AMD) is one of the leading causes of visual loss in the aging population (60–90 years). Methods of DL are useful in better predicting the severity of AMD using imaging methods. DL methods could be used for the public screening or monitoring of AMD worldwide and could further assist in referring the aging population, susceptible to AMD, to a health care provider.
In an interesting study by Qi Yan et al. [15] examined genotype and fundus images of AMD patients which were used as inputs to DL models to dynamically predict an eye that is progressing to the late AMD state, providing disease-severity-related phenotypes. The study indicated that AMD is associated with age, smoking status and a number of genetic variants. As late AMD is irreversible, such a prediction in the early stages would aid patients to adopt preventative care, slowing disease progression. DL methods served as efficient decision support systems (averaged area under the curve (AUC) value of 0.85 (95%CI: 0.83–0.86)), thereby providing various eye services by reducing assessment time and finances via automated analysis. The authors provided a web-based application available online: (accessed on 17 October 2021) to predict AMD state, using both fundus images and genotypes. This could better aid clinicians in predicting year-wise progression of AMD and providing preventive care for patients. The study by Chuan and Yeung et al. [67] served as the first study to use multimodal DL–based architecture for detecting multiple retinal vascular vision threatening eye diseases using multiple image modalities, including retinal fundus photography, optical coherence tomography (OCT), and fluorescein angiography with or without indocyanine green angiography (FA/ICGA). AUCs of 0.987 and 0.969 were attained for predicting retinal vascular diseases and for predicting eye-treatment-requiring diseases. Multimodal imaging is similar to real-world ophthalmology practice, helping in the early screening of the eye diseases and treatment requirements, saving time and making it easy for ophthalmologist on reviewing the images.
With the advent of more advanced-feature extraction and classification, using DL seems to be more supportive of the clinical assessment in early intervention studies to identify aging patients with high risk for progression to advanced AMD. Careful monitoring and detecting preferred practice patterns, identifying individuals at the intermediate AMD stage in a timely manner, can reduce the risk of vision loss due to AMD in aging persons. DL aids in the fine categorization of clinically relevant features of AMD to guide patients who need an ophthalmologist’s opinion. 

2.2. Cardiovascular and Respiratory Disorders in Aging People

The prevalence of cardiovascular and respiratory diseases also increases with age and could be the cause of morbidity and mortality in older patients. Table A2 presents a summary of studies relating to cardiovascular and respiratory diseases in the scope of the current review. Heart disease is the first cause of death after age 65 [70]. Henceforth it becomes important to deal with such issues with advanced methods.
In a study by Zhang et al. [70], DL techniques were applied to predict complications of coronary heart disease in aging patients with an accuracy rate of 87.50%, which further provided a guiding nursing plan. High-risk patients with coronary heart disease related to old age, medical history and lifestyle patterns to predict complications for implementing a better care plan.
The authors in [71] trained DL models on heart MRI videos, ECGs and heart health indicators obtained from UK Biobank participants to identify biomarkers and clinical phenotypes, associated with accelerated heart aging. Heart aging is a measure of the changes that have accumulated in the individual’s heart over their life span having two main heart facets of heart anatomical (MRI-based) and electrical (ECG-based) aging. These facets contain valuable signals and information which needs to be processed at a pace surpassing “traditional” analytical methods. Using DL methods provided a platform to integrate, analyse and make predictions based on the heterogeneous data of MRI scans, ECG signals and environmental phenotypes of age, smoking and hereditary status. DL was used in [72] to analyse chest radiographs for obstructive lung disease in aging people. The results in [72] indicate that a DL Image Model, improves the detection of obstructive lung disease compared to current practices. The results can be used to direct patients to the medical care of pulmonary diseases and lung cancer screening augmenting radiology clinical reports.

2.3. Aging People and Arthritis

The studies have shown that DL can be used effectively to prognosticate joint pain or arthritis outcomes. Such diseases could otherwise trigger inflammation that could lead to irreversible damage in aging people. The study in [75] demonstrated a comprehensive classification and regression analysis using a novel DL on rheumatoid arthritis to determine concrete numerical predictions of disease activity instead of just classifying high or low risk patients, henceforth making treating-to-(predicted)-target strategies better. It was observed that female patients face a higher risk of clinical progression in rheumatoid arthritis. Potentially, lifestyle, sleep or nutrition also contribute to disease prediction. The DL model developed serves as a potential tool for clinical decision support for patients suffering from rheumatoid arthritis. Leung et al. [76], predicted the risk of osteoarthritis and the likelihood of the patient undergoing the total knee replacement, using DL models. These models accurately predicted osteoarthritis progression in patients requiring a total knee replacement within a nine-year time span than traditional grading systems [76]. In this prognostic study [77], electronic health records were monitored for medications, patient demographics, laboratories visited, and of disease activity measures using DL models to prognosticate future patient outcomes for rheumatoid arthritis. This study forecasted RA disease activity for future clinic visits to better guide specialized treatment on an individualized basis. DL methods measured RA disease activity scores across two healthcare systems and suggested that the disease activity, laboratory values, and medications combined together are the strongest predictor of RA at every clinical visit. DL models trained on the large and diverse patient populations proved to be robust and provided useful insights for patient care.

2.4. Alzheimer’s Disease (AD): A Common Disease in Aging People

Alzheimer’s disease (AD) is a progressive brain disorder that gradually destroys brain memory, it is a common disease in aging people, which is caused by dementia. DL approaches have shown promising results for automated diagnosis and the multi-class classification of AD using resonance imaging and tomographic images. Table A4 highlights the review of studies applying DL over AD subjects. The accurate diagnosis of AD is important, especially at the disease’s early stages, so that patients undergo preventive measures even before the occurrence of irreversible brain damage. Deep Learning (DL) has become a common technique for the early diagnosis of AD. Brain imaging techniques are used to visualize the structure and function of the human brain. The most commonly used imaging technique of MRI helps in measuring brain volumes indicating any kind of degeneration due to AD. For the functional connectivity studies of the human brain, independent components analysis (ICA) has been widely used for analysing neuroimaging data [79]. In the study by Qiao et al. [80], a DL-based method was developed to distinguish AD from controls by fusing the functional connectivity. The study detected the underlying biomarkers of AD by analysing functional MRI. Intrinsic functional connectivity in AD patients was noted to be significantly reduced in subcortical brain regions of the hippocampus, amygdala, insula and putam [80].

2.5. Prediction over Spectrum of Age-Related Issues

The studies [17,19,86,87,88,89,90,91,92,93,94,95], highlighted DL-based prediction of other age-related other issues such as Type -2 diabetics, COVID-19 in older patients, coronary blockage in arteries, age-related eye diseases, brain age with old age, age-related disease gene associations, and heart stroke.

In general, overall high performance was achieved (>84% AUC) by adopting DL models in predicting aging-related diseases (Figure 9). Remarkable performance (high AUC of 98%) was achieved in studies involving DL [83] to predict AD in aging people.

Figure 9. Predictive performance of DL methods applied to aging-related diseases.

3. Summary

DL has been playing an important role in providing healthcare professionals with insights, which aids in the detection of health issues early on aging, leading to better patient care. DL has been used in various spheres involving medical image analysis of critical aging diseases, genomics to link genes with diseases, EHR data for personalized care, analysing medical history and providing drugs based on it, cell scope recording medical data on devices, and decreasing frequent visits to consult clinicians. Different types of un-structured and complex data emerging in today’s medical world around aging people are converted into useful formats using DL models. DL models further aid clinicians in the medical classification of diseases, medical resilience, segmentation, cellular senescence, and various other tasks. DL is recommended in dealing with the health of elderly people due to its following benefits:
  • DL learns the important patterns or relationships in large amounts of healthcare data and allows clinicians to perform model-based analysis integrated with their observations; leading to smart care achievable from such big data.
  • Remarkably, DL has achieved human-level performance in disease classification, learning over patterns/objects contained in medical images.
  • When DL is applied over the training data, it becomes more precise with multi-stream architecture and subsequently provides more accurate insights into care processes and diagnostics of aging diseases.
  • DL helps in the detection of clinically relevant features by learning patterns in medical imaging data beyond as perceived by a human observer/clinician.
  • DL approaches are now leading to lower costs and improved and faster outcomes in monitoring the health of aging people.
  • DL provides end-to-end learning models for heterogenous, uncertain and complex medical data.
  • DL provides clinicians with the support they need to understand medical environments.
The development of novel methods making use of DL is acceptable as an objective, complementing the prediction of diseases in the aging population. To facilitate clinicians, researchers could objectively and accurately classify the diseased states using DL. The review presented valuable insights and informed the research in DL, related to the healthcare of elderly people. DL methods were used for pre-processing of medical data and in the analysis, visualization and optimization of deep neural networks in studying aging-related issues.
Research using DL is still evolving to achieve better performance. As medical data grow rapidly, research on the diagnostic classification of aging diseases is shifting towards DL models or their ensembles, integrating completely different formats of data in a DL framework. In the future, new DL modelling architectures could be designed and explored further to provide better clinical presentations for routine care of aging people.

This entry is adapted from the peer-reviewed paper 10.3390/cells10112924

This entry is offline, you can click here to edit this entry!
Video Production Service