July 30, 2024 – Researchers reported that infrared retinal scans may have the opportunity to detect multiple sclerosis with “astonishing” accuracy, which could lead on to earlier diagnosis and treatment.
“This approach addresses an unmet need in MS diagnosis,” said study researcher Rahele Kafieh of the Faculty of Engineering at Durham University within the UK.
The researchers used a pc program called a machine learning model that’s trained to acknowledge certain varieties of patterns. The model was designed to search for telltale signs of MS using two various kinds of eye scans. The program had nearly 100% accuracy, the researchers said.
“The improved diagnostic performance with high sensitivity and specificity suggests that this method can better distinguish between MS patients and healthy individuals and meets the need for more accurate and reliable diagnostic tools for MS,” said Kafieh.
The study was published within the July issue of Translational Vision Science and Technology.
Two scans are higher than one
Damage from MS can affect the retina. Optical coherence tomography, a form of scan the model uses, might help detect signs of MS, monitor disease progression, and determine how well therapies are working. But it stays unclear how incorporating infrared scanning laser ophthalmoscopy, the opposite form of scan utilized in the model, can improve automated diagnosis of MS.
To investigate this, the researchers trained computer models to categorise multiple sclerosis using eye scan data from 32 MS patients and 70 healthy individuals.
The results were impressive. But: “Although the results are promising, this approach is not yet ready for clinical use,” said Kafieh.
To determine whether the outcomes are reproducible, further studies with larger and more diverse populations are needed, she said.
Further research needed
Patricia K. Coyle, MD, professor of neurology on the Renaissance School of Medicine in Stony Brook, NY, commented on the research and reiterated that further research is required.
“People are looking for a proven biomarker to diagnose MS, but we don't have one yet,” said Coyle, who was not involved within the study.
This recent research is “preliminary and would require extensive additional validation studies before it could be accepted. The numbers are very small and there is no robust control group with other subjects with neurological diseases,” Coyle said.