Portable MRI and AI improve Alzheimer’s diagnosis with cost-effective precision

Revolutionizing dementia care: Learn how wearable, AI-powered MRI systems are breaking barriers to Alzheimer’s diagnosis, enabling early detection and global accessibility.

Study: Portable low-field magnetic resonance imaging for evaluation of Alzheimer’s disease. Image credits: illustrissima / Shutterstock

A recent one Nature communication The study optimized portable LF-MRI acquisition and developed a machine learning pipeline to estimate brain morphometry and white matter hyperintensities (WMH) for the diagnosis of Alzheimer’s disease.

Alzheimer’s disease (AD): pathology and diagnosis

AD is a progressive neurodegenerative disease that affects memory, thinking and behavior. It is pathologically characterized by the deposition of amyloid-β (Aβ) and the development of neurofibrillary tangles in the brain. Over time, increased accumulation of these proteins leads to an adverse change in brain structure and increased vascular injury, which is determined by quantifiable brain atrophy and WMH, respectively.

Typically, the progressive presymptomatic stage of AD lasts between 10 and 20 years. This could be why 75% of people with dementia go undiagnosed for long periods of time. The availability of anti-amyloid therapies has increased the urgency for early detection of people with AD or mild cognitive impairment (MCI), because early diagnosis increases treatment benefits.

The diagnosis of AD is based on cognitive testing, which assesses Aβ and phosphorylated tau burden using fluid biomarkers, positron emission tomography (PET), and magnetic resonance imaging (MRI). Doctors can determine the changes in brain structure and integrity using multi-contrast MRI. These imaging indicators include generalized atrophy and hippocampal atrophy, allowing physicians to understand the course of disease progression and cognitive decline.

Although neuroimaging greatly aids in the diagnosis and treatment of AD and MCI, its limited local and global accessibility contributes to its underdiagnosis. A previous study has highlighted the development of a portable LF-MRI, which could effectively increase accessibility and improve the diagnosis of various neurodegenerative diseases. This study highlighted the safety profile and low-cost point-of-care scanning potential of LF-MRI. However, the reduced magnetic field strength lowers the signal-to-noise ratio (SNR), affecting image resolution.

About the study

The current study addressed the aforementioned limitation of LF-MRI for AD and MCI diagnosis by developing machine learning tools that can automatically quantify brain morphometry and white matter lesions.

An imaging pipeline was established that helped quantify brain volumes. The refined super-resolution and contrast synthesis technique (LF-SynthSR) was optimized to increase the LF image resolution in subsequent segmentation (SynthSeg). For example, hippocampal volumes derived from LF-MRI showed close agreement with high-field MRI counterparts, with an absolute symmetric percent difference (ASPD) of 2.8% and a Dice similarity coefficient of 0.87. This strategy helped establish the optimal LF acquisition parameters for accurate quantification. It enabled measurement of white matter hyperintensity (WMH) (WMH-SynthSeg) using automated segmentation of WMH lesions from T2 fluid-attenuated inversion recovery (FLAIR) images acquired at LF. This study validated the LF-SynthSR, SynthSeg, and WMH-SynthSeg using a prospective cohort of patients diagnosed with MCI or AD.

To establish an imaging pipeline, participants from three cohorts were enrolled to undergo MRI acquisition on a portable, low-field MRI at 0.064 T with a conventional high-field scan at a field strength of 1.5–3 T. The first cohort consisted of twenty healthy individuals (10 men and 10 women) without a history of neurological disorders or memory complaints.

The second cohort consisted of 23 participants (11 men and 12 women) with at least one vascular risk factor. However, none of the participants had neurological complaints or a history of memory disorders. The third cohort included 54 individuals (32 men and 22 women) diagnosed with MCI or AD. These participants underwent an LF-MRI imaging protocol that included T1w, T2w, and FLAIR sequences.

Findings of the study

Although the LF-MRI images did not have sufficient resolution for automatic segmentation with high-end software analysis tools, they were first super-resolved (SR) to 1 mm isotropic T1-weighted (T1w) magnetization-prepared fast gradient echo (MP-RAGE )-like images. The study found that isotropic voxel sizes of ≤3 mm improved segmentation accuracy, producing ASPD values ​​of less than 5% for hippocampal volumes. Furthermore, the accuracy of the automated segmentation has improved with the refinement of the LF-SynthSR v2 pipeline, allowing greater utility for low-field imaging applications.

In the first cohort, the accuracy of automated segmentation was assessed by comparing AD-relevant hippocampal, lateral ventricle, and whole brain segmentation volumes generated by the original LF-SynthSR and LF-SynthSR v2 with conventional high-field (HF ) MRI acquired at 3 T.

An improvement in lateral ventricular volume accuracy was achieved by comparing LF-SynthSR v2 with LF-SynthSR v1. Image acquisition time varied between 1:53 and 9:48 minutes depending on voxel size and order. The study also found that isotropic voxel sizes of ≤3 mm improved segmentation accuracy, especially in the low SNR regime of LF-MRI. The accuracy of brain morphometry was found to be influenced by voxel size and geometry. Additionally, the LF-SynthSR v2 segmentation pipeline was validated against HF T1w MP-RAGE segmentations derived from the FreeSurfer segmentation tool ASEG.

WMH lesions due to axonal loss or cerebral small vessel disease were common in patients with cognitive impairment and were quantified using WMH-SynthSeg. The use of these findings on FLAIR as T2 hyperintense lesions and the automated quantification of these lesions increased the AD diagnosis and monitoring capacity of LF-MRI.

This study used machine learning to produce WMH lesion volumes (WMHv) from LF-FLAIR images using WMH-SynthSeg. This strategy allowed simultaneous segmentation of WMH T2 FLAIR lesions, in addition to the previous brain morphometry. The WMH volumes correlated strongly with manual annotations and high-quality imaging standards.

Based on WMHv generated by WMH-SynthSeg, the machine learning tool was validated as it could detect patients with MCI, AD and those who were cognitively normal (CN).

Conclusions

The current study demonstrated that LF-MRI with machine learning tools can diagnose patients with AD or MCI. In the future, this device could also be assessed for its ability to detect neurodegenerative tauopathies and vascular dementia. The portability, low cost, and automated analysis pipeline suggest significant potential for addressing diagnostic disparities globally.