Using Famous Paintings to Detect Early Signs of Parkinson’s and Alzheimer’s

A research team found that a mathematical technique, when applied to artworks, could detect signs of neurodegeneration in the artist.

Salvador Dalí, The Persistence of Memory, 1931, 9.5 in × 13 in (image via Mike Steele Flickr)

When neurodegeneration occurs, the structure of neurons slowly deteriorates, often resulting in the death of the neurons themselves. As parts of the brain atrophy, language can become impoverished and motor functions are compromised. In the world of neurological disorders, diseases like Parkinson’s and Alzheimer’s might be some of the most obvious, their effects becoming immediately visible in the bodies and voices of their sufferers.

Dr. Alex Forsythe wondered if there were other ways of detecting neurodegenerative disorders, beyond one’s body or language. She led a research team at the University of Liverpool’s School of Psychology to determine if fractal analysis, a mathematical technique, can be applied to artworks as a way to find signs of neurodegeneration in the artist. The team’s study, recently published in Neuropsychology, proves that it can.

Fractal analysis can identify and reveal repeating patterns, known as “fractals,” in shapes and images that might appear to be random, such as those found in nature. Applied to an artwork, fractals become personalized fingerprints, distinctive hallmarks of the artist’s brushstroke. Richard Taylor, a physicist and art historian at the University of Oregon, previously utilized fractal analysis to find patterns in the seemingly arbitrary drips of Jackson Pollock’s work, ultimately helping to distinguish the original paintings from a series of fakes.

In kind, Forsythe’s research team examined and studied 2,092 paintings by famous artists known to have suffered from Alzheimer’s and Parkinson’s (Salvador Dalí, Willem de Kooning, Norval Morrisseau, and James Brooks), as well as works by Pablo Picasso, Claude Monet, and Marc Chagall, none of whom had any known neurodegenerative disorders. The researchers found that while these fractal patterns — that is, the mathematically distinct “hallmark” of the artist’s work, not necessarily visible to the naked eye — increased in the work of healthy artists over time, they decreased in those with neurodegenerative disorders. These changes are not attributed to the style nor the complexity of the artists’ work — it is simply that they began to utilize subtly different ways of working as their relationship to their own brain began to shift.

The implications are vast: Could the changes in an artwork detect early symptoms of neurological decline? How, then, would that change what we know about such symptoms in the first place? To gain insight, I spoke with Dr. Forsythe about the study’s inception and significance.

*   *   *

Monica Uszerowicz: How did you decide to conduct this study?

Alex Forsythe: Several events aligned somewhat by serendipity but also by tragedy. In my early career I was interested in language analysis and was aware of the changes in writing that had been found in individuals who went on to develop Alzheimer’s Disease. I shifted away from that particular research interest, because I felt the field was quite crowded, into image measurement and the psychology of art for my PhD and subsequent academic career.

I came across Richard Taylor’s Nature paper on the use of fractals to identify fraud in the art world at a time when my father-in-law’s health was deteriorating rapidly from Parkinson’s Disease. The thought just occurred to me: I wonder if we can use fractal analysis to identify changes in artists with neurological deterioration, in the way in which we use micrographia — a diagnostic tool in Parkinson’s Disease. Like the language studies in Dementia, are there changes in art?

MU: Can you explain fractal analysis in laymen’s terms?

AF: Fractal patterns offer one way to represent the rough complexity of the natural world. The mathematician Benoit Mandelbrot developed them to measure what was previously unmeasurable. He called it measuring “the art of roughness” and self-similarity in nature. Fractals are patterns that repeat at different magnitudes: for example, the trunk of a tree breaking into its branches, then breaking into finer and finer branches. Simply placing a fractal helps us measure something that cannot be measured through traditional geometry — a ruler, for example.

In art, the individual brushstrokes self-replicate throughout the painting, creating form, space, and pattern — rather like an artist’s handwriting, but a handwriting which can also require fractal movements from the fingers, hands, arms, and in some cases entire body. We have found that the rhythm of these fractal patterns are captured for eternity in the painting, almost like a DNA footprint. We have found that this ‘code’ stays with an artist even if they change genre. Picasso was a great experimenter and regularly changed his artistic style, but the fractal dimension of his works remained stable over his lifespan.

We are finding higher quantities of fractal content in the early and midlife works of artists who later developed neurological problems. You can see, for example, obvious areas of high fractal content in much of the early works of Dalí, where the artist has used his brushstrokes to create water, sky, and mountains, but also in the curvature of the melting clocks. If we were to zoom into the image, we could see fractal content through the repeating nature of the brushstrokes. Dalí worked within what we call a “normal range” for his pieces until midlife, when something starts to change, first subtly, then drastically.

Later works, such as “Portrait of My Dead Brother,” which he painted more than 30 years later in 1963, has areas of repeating patterns, but much of this content is no longer fractal — for example, the areas of circular dots which are not fractal but adhering to geometric principles. Overall, the artists significantly decreased the amount of fractal content that they normally used to create their work. We found similar patterns in the work of the Canadian artist, Norval Morrisseau. I suspect with late Parkinson’s Disease, the artist may be finding it harder to control the brush and create the delicacy which is necessary to create highly fractal content. However, we can see changes taking place from the mid-careers of both artists, possibly before they were aware they had any neurological deterioration.

The pattern in Dementia is somewhat different, and quite dramatic. De Kooning, famous for his series of Woman paintings, continued to paint after being diagnosed with Alzheimer’s, but the fractal content in his pictures decreased quite rapidly from middle age. A similar pattern is found with James Brooks. These patterns are not seen in artists who aged normally.

Controls were critical to rule out possible explanations, such as changes in artistic style. Once I had determined there was an effect, the natural step for me was to determine if there were patterns in other neurological problems.

MU:  Did their work get “simpler” over time?

AF: Simpler is one way of thinking about the art works, but actually the artists are still able to produce complex works in later life. They rely less on fractal content to depict their message. So as I explained earlier, Dalí worked quite complexly throughout his life, but later complex paintings often had more geometric, rather than fractal content.

MU: What are the practical applications of the study? Can it be used for preventative care and early detection? Could visual art “examinations” be supplementary to other forms of testing?

AF: The sample is large enough to be somewhat representative, and could be built on with the addition of further artists. If we could find a way of speeding up the analysis, it has all sorts of potential, but my hope is that this study opens up routes of examination and exploration that we have not previously considered, new conversations about what symptoms of neurological decline could potentially be.

MU: What are the larger implications here?

There are, for example, some commonly accepted ideas that one of the first signs of dementia is memory loss, or with Parkinson’s Disease there will be a tremor. This study suggest that other subtle changes might be taking place before before we are perhaps even aware that something has changed. We know that aspects of our personality, such as openness to experience, changes across lifespan. I am hoping to explore the extent to which changes in preference might be triggered through neurological deterioration. The motivation for this direction is that patients with dementia often successfully take up artistic activities. De Kooning, for example, actually won more accolades for the work that he created as his illness progressed.

This is about a new perspective: looking at problems from a different direction. If we identified different processes, it could have implications for the development of new treatments and drugs.

comments (0)