New research suggests that an emerging predictive biomarker, derived from a combination of magnetic resonance imaging (MRI) brain scans and a machine learning algorithm, has significantly greater accuracy than previously established measurements for diagnosing Alzheimer’s disease.
Estimates suggest that more than six million people in the United States have Alzheimer’s disease, and the COVID-19 pandemic may have triggered a 17 percent increase in the number of deaths related to Alzheimer’s disease and dementia in 2020.1 However, emerging research suggests the combination of a machine learning model and a single T1-weighted magnetic resonance imaging (MRI) scan could facilitate enhanced detection of Alzheimer’s disease and possibly earlier intervention.
The study, which was recently published in Communications Medicine, examined a predictive machine learning model that ascertains mesoscopic traits from T1-weighted MRI scans of the brain and facilitates subsequent use of a predictive biomarker to help diagnose Alzheimer’s disease.2
The researchers found that the biomarker had a 98 percent accuracy in detecting Alzheimer’s disease in comparison to 62 percent accuracy with the measurement of cerebrospinal fluid beta amyloid and 26 percent accuracy with the measurement of hippocampal atrophy.2
“Currently, no other simple and widely available methods can predict Alzheimer’s disease with this level of accuracy so our research is an important step forward,” noted Eric Aboagye, FMedSci, a professor of cancer pharmacology and molecular imaging at Imperial College London in the United Kingdom. “ … Our new approach could also identify early-stage patients for clinical trials of new drug treatments or lifestyle changes, which is currently very hard to do.”
Employing established software for the segmentation of the brain and radiomics analysis, the MRI-based predictive biomarker “does not require a subject matter expert,” according to the study authors. They noted that the biomarker is based upon the weighted sum of 20 extracted features derived from 14 out of 115 regions of the brain.
“The algorithm computes manually engineered features, allowing an easy interpretation of the (biomarker) and facilitating clinical translation. To avoid overfitting, the dimensionality of the model is reduced with the ‘least absolute shrinkage and selection operator (LASSO),’ which selects the most informative and less redundant features corresponding to specific brain regions,” wrote Aboagye and colleagues.
Noting that the machine learning model identified changes in the cerebellum and ventral diencephalon that had not been associated with Alzheimer’s disease in the past, the study authors suggest that the new algorithm could enhance neuroradiologist assessment of MRI brain scans in this patient population.
“Although neuroradiologists already interpret MRI scans to help diagnose Alzheimer’s (disease), there are likely to be features of the scans that aren’t visible, even to specialists,” suggested Paresh Malhotra, MD, a co-author of the study and a consultant neurologist at the Imperial College Healthcare NHS Trust. “Using an algorithm able to select texture and subtle structural features in the brain that are affected by Alzheimer’s could really enhance the information we gain from standard imaging techniques.”
References
1. Alzheimer’s Association. Alzheimer’s disease facts and figures. Available at: https://www.alz.org/alzheimers-dementia/facts-figures?utm_source=google&utm_medium=paidsearch&utm_campaign=google_grants&utm_content=alzheimers&gclid=CjwKCAjw-8qVBhANEiwAfjXLrqaFewnIi2l9vCjpgb4bXws3TAjpkUTSH2QD6VZw4hVbP_xU_6YjfRoC2sQQAvD_BwE . Accessed June 22, 2022.
2. Inglese M, Patel N, Linton-Reid K, et al. A predictive model using the mesoscopic architecture of the living brain to detect Alzheimer’s disease. Commun Med. 2022. Available at: https://doi.org/10.1038/s43856-022-00133-4 . Published June 20, 2022. Accessed June 22, 2022.
AI Algorithm Comparable to Radiologists in Differentiating Small Renal Masses on CT
May 14th 2024An emerging deep learning algorithm had a lower AUC and sensitivity than urological radiologists for differentiating between small renal masses on computed tomography (CT) scans but had a 21 percent higher sensitivity rate than non-urological radiologists, according to new research.
Current Insights and Emerging Roles for Contrast-Enhanced Mammography
May 10th 2024In a recent lecture at the 2024 ARRS Annual Meeting, Jordana Phillips, MD, discussed the role of contrast-enhanced mammography in staging breast cancer, evaluating response to neoadjuvant chemotherapy and recalls from screening.
MRI-Based Deep Learning Algorithm Shows Comparable Detection of csPCa to Radiologists
May 8th 2024In a study involving over 1,000 visible prostate lesions on biparametric MRI, a deep learning algorithm detected 96 percent of clinically significant prostate cancer (csPCa) in comparison to a 98 percent detection rate for an expert genitourinary radiologist.