A new algorithm that examines neurological and oculomotor signs may make it much easier to diagnose certain “often overlooked” rare neurological conditions, including Gaucher disease type 3 (GD3), the research team that developed it said.
The algorithm is described in the study “An algorithm as a diagnostic tool for central ocular motor disorders, also to diagnose rare disorders,” published in the Orphanet Journal of Rare Diseases.
Central ocular motor disorders are neurological conditions that affect eye movement; they can be present in diseases including multiple sclerosis, GD3, Tay-Sachs disease, and Niemann-Pick disease type C. Gaucher type 3, specifically, is marked by problems in controlled eye movement.
These conditions can be difficult to diagnose, even for experienced neurologists. But there’s actually quite a lot known about how these conditions affect different parts of the brain, and what symptoms are likely to result.
As such, researchers behind this study wanted to make a tool that would take the findings of a bedside neurological evaluation — without the need for additional laboratory tests — and determine the most likely diagnosis as a way of guiding neurologists and ensuring that particularly rare conditions are not overlooked.
They created a “simple” algorithm in which a neurologist would answer “yes,” “no,” or “unknown” to a list of 60 patient signs and symptoms. The algorithm would then use this data to match each patient to one of 13 neurological disorders, including those listed above, or to a 14th “none-of-the-above” category.
The researchers first tested out their algorithm using data from 102 patients (55% male, average age of 48); all had been previously diagnosed at a hospital and had neurological examination data available. The team next compared what the algorithm determined to each patient’s actual diagnosis, and tweaked the algorithm based on discrepancies.
Researchers then validated their tool using a new group of 104 patients (57% male, average age of 46).
In this validation, the algorithm’s sensitivity ranged from 60% to 100% depending on the disease, and its specificity ranged from 66% to 95%. For GD3, the sensitivity was 80.0%, and the specificity was 91.5%.
Of note, sensitivity is the proportion of patients with a condition who are correctly diagnosed using the test (an algorithm here), while specificity is the proportion of people without the condition who also test negative for that condition.
These high values support the utility of this algorithm as a tool that aids in diagnosing central ocular motor disorders, particularly rare conditions that are treatable — one, Wernicke’s encephalopathy, is even curable — but are often under-diagnosed.
The research team also created a version of the algorithm that linked patient symptoms to different affected regions of the brain, instead of to a particular disease. This algorithm didn’t perform as well: sensitivity ranged from 0% to 100% depending on the brain region, and specificity ranged from 52% to 99%.
Although this aspect of the algorithm is far from fool-proof, it “can still give an indication of where to look for pathologies in imaging,” the researchers wrote, meaning it still has some utility in a clinical setting.
The team concluded that its new algorithm “is a useful tool for diagnosing diseases, in particular rare ones, which present with central ocular motor disorders.”