Introduction
Dementia, or Major Neurocognitive Disorder (MND), is a general term for any disease that causes a substantial decline in at least one cognitive domain including memory, learning, executive function, and additionally, impairs an individual’s ability to perform daily tasks. Alzheimer’s disease (AD) can be divided into three stages: early-stage (mild), middle-stage (moderate), and late-stage (severe). Each phase serves to classify the development of mental decline. In the mild stage, the person may start to experience memory loss, poor judgement, the repetition of questions, misplacing items, and difficulty planning or organizing. Signs of moderate stage diseases involve increased memory loss and confusion, difficulty with language, difficulty organizing thoughts, paranoia, hallucinations, and mood changes such as anxiety and aggression. As the disease progresses to the severe stage, the patient will be completely dependent on others for care and will display symptoms such as an inability to communicate, loss of awareness of their surroundings, weight loss, increased propensity to infections like pneumonia, and changes in physical abilities including walking, bowel control, sitting, and eventually swallowing. Each person will experience different symptoms of the disease and may progress through the stages at a variable rate. The progression pace is influenced by age, genetics, and other factors.
There is presently no cure for AD, leaving healthcare professionals to focus on slowing the progression of the disease to improve the quality of life of the patient. Although there is no cure, detecting AD in a timely and accurate manner is important because it allows for the development of an earlier treatment plan and care plan that may preserve cognitive function and prevent irreversible damage through intervention and lifestyle modifications. A definitive diagnosis can be made postmortem with the identification of neurofibrillary tangles (NFT) or diffuse amyloid depositions known to be closely linked to the disease. Improvements in blood biomarkers are also promising, but still underutilized. It is widely believed that the onset of neuropathological hallmarks of AD, such as NFT and abnormal amyloid plaques, begin to form years prior to the appearance of clinical symptoms. Neuroimaging has served as a critical tool in establishing diagnostic indicators of AD during the preclinical stage allowing for early diagnosis and intervention. The diagnostic imaging modalities most widely used in the diagnosis of neurodegenerative diseases is magnetic resonance imaging (MRI), computed tomography (CT) and positron emission tomography (PET). However, as neuroimaging technology quickly advances, there is a challenge in analyzing and interpretating vast amounts of brain imaging data. Given these limitations, there is great interest in using computer-aided algorithms for integrative analysis, namely artificial intelligence.
Artificial intelligence (AI) is a field of developing computer programs that simulate human functioning. There are two subsets that have been used significantly in AD research – machine learning (ML) and deep learning (DL). Illustrating a simple definition, machine learning uses algorithms to recognize patterns from data and applies that knowledge to reach solutions and make predictions for new information. The commonly used learning processes are supervised and unsupervised learning. Deep learning, a more complex subset of ML, uses a convolutional neuronal network architecture to analyze data in a logical form similar to how the human brain functions. There are various artificial intelligence techniques in Alzheimer’s disease detection, for example, ensemble classifiers, support vector machines (SVMs), and random forest are one of many techniques in recent studies.
Current diagnostic imaging and its limitations
Magnetic Resonance Imaging
Magnetic resonance imaging uses powerful magnets to align protons along a magnetic field. A radiofrequency current stimulates and spins the protons out of equilibrium in the patient. MRI has been largely used in the clinical identification of AD due to its ability to provide detailed information about brain structure in vivo.
The diagnostic guidelines of AD created by the National Institute on Aging and the Alzheimer’s association recommend the use of structural MRI (sMRI) in its criteria, highlighting its integral role in the clinical assessment of patients with suspected AD. sMRI assesses brain atrophy and tissue changes with its capability to differentiate between grey matter and white matter. Structural MRI studies of patients with AD have revealed atrophy in medial temporal lobe structures including the hippocampus, amygdala, entorhinal cortex (ERC) and parahippocampal gyrus. Medial temporal lobe atrophy (MTA) is associated with lower executive function, general cognitive function, and episodic memory performance. Studies have found ventricular enlargement, whole brain atrophy and cortical thickness reduction in patients with AD. Other more advanced MR techniques that are not used in routine clinical settings but serve an important role in AD research include functional magnetic resonance imaging (fMRI) and diffusion tensor imaging (DTI). fMRI is based on blood-oxygen level dependent (BOLD) changes in the brain that occur during specific tasks and has been widely used to study pathophysiologic changes seen in memory loss in AD. A meta-analysis of fMRI activation during episodic memory in AD and MCI showed hypoactivation of the medial temporal lobe structures in AD and hyperactivation in MCI.
Computed tomography
Computed tomography is a computerized x-ray imaging procedure that generates cross-sectional images or “slices”. A narrow beam of X-rays are aimed at a patient and quickly rotate around the body to produce signals that are processed by the computer. CT is not recommended for first-line imaging as it is less sensitive in detecting changes associated with cognitive impairment compared to MRI. However, there are still many advantages such as lower cost, wider availability, and shorter acquisition time.
Under current recommendation and guidelines, structural imaging (MRI or CT) is required for evaluation of patients presenting with cognitive symptoms in the clinical setting. CT reveals the anatomic structure of the brain to detect brain atrophy and rule out other abnormalities that can be mistaken as AD such as tumors, hydrocephalus, and chronic subdural hematoma. Serial CT imaging has been used to track and observe changes as the disease progresses. There are various limitations to CT scans when compared to MRI. MRI is more sensitive at detecting focal atrophic changes in the nuclei, more sensitive to white matter changes and does not require ionizing radiation. MRI remains the preferred first-line modality; however, both play a key role in ruling out structural lesions of the brain in individuals with dementia.
Positron emission tomography
Positron emission tomography (PET) scans work at the molecular level to produce three-dimensional images that illustrate biochemical and molecular processes. A PET radiotracer attaches to the molecular target which then allows for the measurement of various processes such as metabolism, blood flow, and regional chemical composition in the body. PET radiotracers have been developed in the field of AD to meet the increasing need for early detection and treatment monitoring of the disease. The following section will provide a brief overview of current PET scans available for AD imaging, namely FDG-PET, amyloid-PET, and tau-PET.
Fluorodeoxyglucose (FDG), an analog of glucose, is introduced to the patient intravenously to measure brain metabolism. The main energy source for the human brain is glucose. Any changes to neural activity in neurodegeneration will be reflected by glucose consumption. The term FDG uptake refers to the amount of radiotracer uptake. Areas of low radiotracer uptake are associated with lower brain activity and produce darker spots (hypometabolism) on images. The standardized uptake value (SUV) is a commonly used method to assess the degree of FDG uptake in a region of interest in PET imaging. SUV is calculated as the ratio of tissue activity concentration and administration dose, divided by body weight. The ratio of SUV data from two different regions within the same PET image is referred to as the SUV ratio (SUVr). SUVr also serve an important role in quantifying tracer uptake in amyloid and tau imaging. The characteristic indication of AD on FDG-PET is hypometabolism in the posterior cingulate cortex (PCC), precuneus (PrC), parietotemporal cortex, and in the frontal cortex in advanced stages. FDG-PET has provided clinical value in detecting distinct patterns of cortical hypometabolism in AD, differentiating between other neurological diseases, and in predicting MCI conversion to AD.
Amyloid-PET enables in vivo detection of amyloid deposits in the brain, one of the neuropathological hallmarks of AD. Currently, three amyloid PET tracers are approved by the FDA for clinical use: 18F-Florbetaben (Neuraceq), 18F-Florbetapir (Amyvid), and 18F-Flutemetamol. Amyloid accumulation is commonly assessed with SUVr quantification though this technique has been shown to overestimate amyloid burden in cognitively normal individuals. A major limitation is the fact that a positive amyloid-PET scan is not sufficient to diagnose AD. It serves merely as a specific and sensitive tool that can assess the likelihood of a diagnosis. Therefore, amyloid imaging may serve a limited role in the future for assessing cognitive decline in patients.
Tau is a protein that accumulates in the brain of individuals with AD and other forms of dementias. Recently, the FDA has approved Tauvid (flortaucipir F18) for PET imaging of the brain to assess the distribution of aggregated tau neurofibrillary tables (NFT), another neuropathological hallmark of AD. The distribution of tau proteins deposits has been shown to be more closely associated to cognitive decline when compared to amyloid. Tau PET has also been used to differentiate AD dementia from other neurodegenerative diseases such as frontotemporal lobar degeneration (FTLD) disorders based on the location of tau protein in the brain. Despite its utility, limitations of tau imaging include reports of in vivo off-target binding, variability of thresholds for tau positivity rates between studies, and similarly as with amyloid, a positive tau marker alone is not sufficient for an AD diagnosis.
Current use of AI in AD research
Brain imaging modalities like MRI, CT, PET, as well as molecular biomarkers, such as amyloid plaques and tau in CSF, are used in clinical settings to identify a patient’s cognitive status. The use of AI with neuroimaging for the diagnosis of AD is a rapidly emerging field and has the potential to solve these problems. AI has the ability to integrate complex multimodal data, improve the accuracy of biomarker-based testing, and has a promising future of providing accurate and widely accessible early diagnosis of AD. Machine learning has been used in the disease classification of AD. Machine learning can create models from a combination of AD biomarkers to make a more accurate diagnosis. AI with neuroimaging is a rapidly emerging field and continues to be extensively tested in AD research. However, these models are not currently available in routine clinical practice, but the progression towards use in clinical settings is expected as technology advances.
Pros of AI
Improving diagnostic accuracy
AI has the potential of improving diagnostic accuracy in the clinical setting. The results of a large-scale study demonstrated that the ML algorithm had an accuracy of 92% in predicting a 2 years incidence of dementia, presenting far more accuracy than other existing dementia risk prediction models such as Cardiovascular Risk Factors, Aging, and Incidence of Dementia (CAIDE) Risk Score and Brief Dementia Screening Indicator (BDSI). Therefore, not only can the ML algorithm accurately tell clinicians who will go on to develop dementia, but the algorithm could also help reduce the number of people who have been falsely diagnosed. The study demonstrates ML algorithms can be helpful in the decision-making process and be used as a potential diagnostic and validation tool in clinical settings.
Efficient analysis of data
As the number of radiographic images generated rise globally, AI has the potential to help radiologists efficiently analyze data more quickly. By adopting this model in the clinical setting, a worklist prioritization would allow critical patients to receive quicker diagnosis and treatment while granting more time to radiologists for other duties. With more time available, AI would allow radiologists to oversee complex imaging studies, consult with physicians, and spend more direct patient contact. Additionally, AI technology can improve image acquisition and quality while reducing radiation dose to patients. AI technology can also help clinicians detect hard to see lesions. The AI tool can also accurately recognize subtle changes in volume and can assess whether lesions are changing in size or number compared to prior scans. Additionally, AI technology can help identify artifacts in images and decrease false-positive interpretations. Overall, AI would greatly improve the productivity and efficiency in clinical settings.
Reduce physician burnout
Risk factors that may contribute to physician burnout include work factors such as excessive workloads, long working hours, frequent call duties; personal characteristics associated with burnout include being self-critical, sleep deprivation, engaging in unhelpful coping strategies; and organizational factors such as negative leadership behaviors, limited interpersonal collaboration, and limited social support for physicians. Burnout is not only dangerous to the physician, but also to the patients. New technology could serve as a potential solution to help combat these issues and reduce burnout. AI could help radiologists by taking over tedious tasks such as scheduling patients, speeding workflows, and triaging images. Additionally, with improvement in the quality of imaging, diagnostic accuracy and efficiency will improve as well. AI has the potential to relieve stress with improved efficiency and productivity.
Challenges of AI
Generalization and data shortage
A major limitation of AI use in the clinical setting is generalization. There are a series of steps that are implemented when creating an algorithm. The model must be trained, validated, and tested to evaluate its performance. The goal is to develop a model that is able to detect seen data and can generalize to make appropriate predictions to unseen data. The challenge with generalization arises when the trained algorithm loses its performance when applied to different datasets. A major cause of poor generalization is overfitting. Overfitting occurs when a model is too dependent on training dataset and is therefore not able to generalize well to data it has never seen before. A possible strategy to overcome overfitting is to collect more training data. However, there is a lack of sufficient training data in the field of Alzheimer’s disease.
Lack of an in vivo gold standard for diagnosis
Currently, there is no definitive gold standard to diagnosis AD in vivo. Neuropathologic evidence of extracellular amyloid plaques and intracellular neurofibrillary tangles in a post-mortem analysis remains the only gold standard. Therefore, it is not enough to identify individuals at highest risk for AD with simply the presence of amyloid alone but rather, there needs to be evidence of both amyloid-β and biomarkers indicating neurodegeneration.
Conclusion
Alzheimer’s disease is a progressive, neurodegenerative disease that greatly affects the life of the patient and the family. Signs of mental deterioration are often confused for old age and by the time the patient decides to visit the physician; a diagnosis may be too late. Neuroimaging, although costly, has served a vital role in determining markers for the diagnosis of AD. MRI, CT, and PET are modalities used in clinical settings to identify a patient’s cognitive status, each with their own benefits and limits. As noted, limitations exist when using neuroimaging alone to identify AD. With advancing technology, there is a challenge in analyzing and interpretating vast amounts of brain imaging data. The use of AI with neuroimaging for the diagnosis of AD is a rapidly emerging field and has the potential to solve these problems. Research involving AI in neuroimaging has resulted in remarkable outcomes including the ability to classify, track the progression, and diagnose AD in early stages. These models are not currently available in routine clinical practice, but the progression towards use in clinical settings is expected as technology advances.
AI introduces limitless possibilities in the future diagnosis of AD, yet there is still resistance from the healthcare community to incorporate AI in the clinical setting. The main advantages of AI considered are it’s potential to improve diagnostic accuracy, improve the efficiency in analyzing radiographic data, and reduce physician burnout. The challenges include generalization and data shortage, and lack of in vivo gold standard. Not only will patients greatly benefit from AI, but so too will current and future radiologists. No other field of medicine is as technologically dependent as radiology and together with AI, patient health will be greatly improved.