A team of researchers has discovered initial signs of Parkinson’s disease in eye scans well before the onset of symptoms. This breakthrough paves the way for early screening tools and preventative actions against neurodegenerative diseases, utilizing the burgeoning field of oculomics.
The possibility of early detection of Parkinson’s disease through eye scans opens the door to preventive tactics against neurodegeneration. This advancement is attributed to a significant study employing AI in the analysis of retinal images.
A collaboration between UCL and Moorfields Eye Hospital has led to the identification of indicators of Parkinson’s disease in patients approximately seven years prior to their clinical symptoms.
This marks the first instance of such findings being observed years before a formal diagnosis, made possible by the most extensive study on retinal imaging in Parkinson’s disease to date.
Study Approach and Results
Published in Neurology, the journal of the American Academy of Neurology, the study utilized AI to identify Parkinson’s markers in eye scans. The study’s analysis of the AlzEye dataset was corroborated by data from the larger UK Biobank database, which consists of healthy volunteers, confirming the initial findings. The incorporation of these two extensive datasets allowed the researchers to detect these subtle markers despite Parkinson’s disease having a relatively low prevalence in the population (0.1-0.2%). The creation of the AlzEye dataset was facilitated by INSIGHT, the largest global database of retinal images and related clinical data.
The Diagnostic Role of Eye Examinations
Previously, eye scans have uncovered signs of other neurodegenerative diseases, such as Alzheimer’s, multiple sclerosis, and recently, schizophrenia. This forms part of an innovative and promising research area known as “oculomics.”
Eye scans have also been instrumental in detecting predispositions to conditions like hypertension, cardiovascular diseases including strokes, and diabetes.
Medical professionals have long recognized the eye as a ‘window’ to overall health, offering direct insights into various health aspects. High-resolution retinal images, especially using a type of 3D scan known as ‘optical coherence tomography’ (OCT), have become a routine aspect of eye care. An OCT scan can generate a detailed cross-section of the retina in under a minute, with precision up to a thousandth of a millimeter.
AI in Unveiling Hidden Insights
These images are crucial for monitoring eye health, and their significance extends further. As the only non-invasive method to view cell layers beneath the skin, retinal scans have become invaluable. Recent advancements have seen researchers using powerful computers to analyze vast numbers of OCTs and other eye images rapidly. Employing a type of AI called ‘machine learning’, computers can now extract hidden information about the body from these images. This is the essence of oculomics.
Dr. Siegfried Wagner of UCL Institute of Ophthalmology and Moorfields Eye Hospital, and principal investigator of several AlzEye studies, remarked on the remarkable discoveries possible through eye scans. He noted that while it’s not yet feasible to predict individual Parkinson’s development, this method could soon serve as a pre-screening tool for those at risk.
He emphasized the significance of detecting signs of diseases before symptoms emerge, allowing for lifestyle changes to prevent certain conditions and delaying the onset of life-altering neurodegenerative disorders.
Collaborative Research Efforts
The study involved collaboration between several NIHR Biomedical Research Centres, including those at Moorfields Eye Hospital, University Hospital Birmingham, Great Ormond Street Hospital, Oxford University Hospital, University College Hospital London, and the UCL Great Ormond Street Institute of Child Health. These collaborations enhanced the research’s scope and quality.
Professor Alastair Denniston, consultant ophthalmologist at University Hospitals Birmingham and part of NIHR Moorfields BRC, highlighted the study’s demonstration of eye data’s potential when combined with technology to detect early Parkinson’s signs.
Miss Louisa Wickham, medical director at Moorfields, pointed out the significant public health impact of increasing imaging across a broader population. She emphasized that OCT scans, being more scalable, non-invasive, less expensive, and faster than brain scans, are ideal for predictive analysis.
Technical and Background Details
Parkinson’s disease is marked by a decrease in dopamine. Post-mortem examinations of patients with Parkinson’s have shown differences in the retina’s INL (inner nuclear layer). Previous OCT scan studies have indicated potential morphological abnormalities linked to Parkinson’s, albeit with inconsistencies.
This study confirmed earlier findings of a thinner GCIPL (ganglion cell–inner plexiform layer) and, for the first time, a thinner INL. It also established that reduced thickness in these layers correlates with an increased Parkinson’s risk, beyond other factors or comorbidities.
Future Research Directions
Further studies are needed to ascertain whether GCIPL atrophy progression is influenced
