Pathways to chronic disease detection and prediction: Mapping the potential of machine learning to the pathophysiological processes while navigating ethical challenges

Abstract

Chronic diseases such as heart disease, cancer, and diabetes are leading drivers of mortality worldwide, underscoring the need for improved efforts around early detection and prediction. The pathophysiology and management of chronic diseases have benefitted from emerging fields in molecular biology like genomics, transcriptomics, proteomics, glycomics, and lipidomics. The complex biomarker and mechanistic data from these "omics" studies present analytical and interpretive challenges, especially for traditional statistical methods. Machine learning (ML) techniques offer considerable promise in unlocking new pathways for data-driven chronic disease risk assessment and prognosis. This review provides a comprehensive overview of state-of-the-art applications of ML algorithms for chronic disease detection and prediction across datasets, including medical imaging, genomics, wearables, and electronic health records. Specifically, we review and synthesize key studies leveraging major ML approaches ranging from traditional techniques such as logistic regression and random forests to modern deep learning neural network architectures. We consolidate existing literature to date around ML for chronic disease prediction to synthesize major trends and trajectories that may inform both future research and clinical translation efforts in this growing field. While highlighting the critical innovations and successes emerging in this space, we identify the key challenges and limitations that remain to be addressed. Finally, we discuss pathways forward toward scalable, equitable, and clinically implementable ML solutions for transforming chronic disease screening and prevention.

Keywords: OMICs data; big data; chronic diseases; disease prediction; machine learning algorithms.

Figure 1

Laboratory methods and application of machine learning for disease prediction and diagnosis. Diagnosis of diseases begins with screening a population, sample collection, processing, and quantitation with molecular/analytical methods (e.g., polymerase chain reaction [PCR], liquid chromatography mass spectrophotometry [LC‐MS], high‐performance liquid chromatography [HPLC], sodium dodecyl sulfate polyacrylamide gel electrophoresis [SDS‐PAGE]), and immunofluorescent macroscopy among others. Machine learning algorithms can discover patterns in the data generated from analytical methods and make predictions.

Figure 2

Pathophysiology of Type II diabetes. Alpha and beta cells of the pancreas secret insulin and glucagon, respectively. These two hormones interact with the liver to regulate blood sugar levels. While glucagon stimulates glycogenolysis and gluconeogenesis, leading to a rise in sugar levels in the blood, insulin promotes glycogenesis and glucose uptake in skeletal muscles and other tissues. In diabetes, there is either impaired insulin secretion or insulin resistance, resulting in elevated plasma concentration of sugar.

Figure 3

Pathophysiology of chronic kidney disease. Kidney injury results in loss of nephrons and increased levels of angiotensin II. Angiotensin II is a vasoconstrictor that triggers increased blood pressure. Alongside hypertension, there is an increase in glomerular permeability, tubular protein reabsorption, tubular/interstitial inflammation, and eventually renal scarring.

Figure 4

Different types of cardiovascular diseases.

Figure 5

Principal component analysis (PCA) (with 95% confidence ellipses) showcases the clustering patterns and reasonable discriminatory power for samples with congenital heart defects (case) and control using the top 20 expressive genes identified via RF based on orthogonal linear combinations of the features.

Figure 6

The metastatic cascade. The spread of a tumor is characterized by a sequence of events. These events are local invasion, intravasation, circulation through the vasculature, extravasation, formation of micrometastasis, and colonization.

Figure 7

Response to inflammation. The components of inflammation include vascular changes―vasodilation that results in increased blood flow and increased vascular permeability that allows fluids to reach the infected site. Immune cells, including macrophages and leukocytes, are also activated to phagocytose infectious agents. Monocytes are also recruited, which become macrophages. These macrophages can differentiate into M1 and M2 when stimulated by cytokines. M1 releases the anti‐bactericidal compound NO, whereas M2 macrophages promote wound healing and tissue repair. Protein systems, including the complement, coagulation, and kinin, are all activated inflammatory responses. Depending on the stimulus, T‐cells may be differentiated into TH1, TH2, and TH17, and all these are involved in various stages of inflammation.