The depiction of migraine visual aura on the internet

Abstract

Introduction

The most common and multifaceted migraine aura symptoms are visual disturbances. Health information is one of the most popular topics on the internet but the quality and reliability of publicized information is unknown. The aim of this study was to analyze images of migraine aura on Google to determine the frequency of correct presentations of visual aura and distribution of visual aura phenotypes.

Methods

Two authors screened the 100 highest indexed migraine aura related images on Google. The content of the images was categorized into elementary visual symptoms.

Results

Forty out of 100 images were accurate representations of visual migraine aura. Such images included 31 different visual aura phenotypes. The majority had more than one elementary visual symptom (median 2, IQR 1–3), most commonly “bean-like” forms (45%), zigzag lines (40%), and foggy/blurred vision (33%).

Discussion

Forty percent of images were accurate portrayals of visual migraine aura symptoms, but these presented limited phenotypes. The information derived from the internet photos may hinder the effective recognition of aura symptoms. Thus, there is a need to provide a more comprehensive representation of visual migraine aura symptoms on the internet.

Introduction

Visual disturbances are the most common migraine aura symptoms, while also being the most multifaceted (1–3). Health information is one of the most frequently searched topics on the internet (4–6) and many patients view the internet as a valuable and reliable source of health-related information (7). While there is a high prevalence of migraine, the general migraine literacy appears to be poor and patients who are seeking information may not be trained to evaluate this medical information (8). Google is the most popular website and search engine in the world. However, the content is not peer-reviewed, and the reliability of the information is uncertain.

The aim of this study was to analyze the highest indexed images of migraine aura on Google to determine the frequency of correct presentations of visual aura and distribution of visual aura phenotypes.

Materials and methods

Search strategy

We searched for the most popular migraine aura-related images on Google Images (http://images.google.com). We conducted the search on 11 September 2019 using the Mozilla Firefox browser (Mozilla Foundation) with all tracking cookies deleted and the “Private Window” function enabled. These two factors make the browser appear as a naive user. We used the English search term “migraine aura” with the “SafeSearch” filter off. Two migraine researchers (MV, TPD) screened the 100 highest indexed images independently. The rationale for a cutoff of 100 images is that most people doing an online search will look no further than the first three pages generated (9) (Figure 1). Any disagreement during the screening process was resolved by consensus between the authors and involved a third rater (AH) when necessary.

Figure 1.

Flow diagram of the image selection process.

This study did not require the approval of a scientific ethics committee as the data is publicly available.

Image analysis

The content of the images was categorized into elementary visual symptoms (EVSs: e.g. “zigzag or jagged lines”, “foggy/blurred vision”) according to a previous systematic review of the literature (3). Disagreement during EVS categorization was resolved by consensus in the author group.

Statistical analysis

We collected and analyzed data (frequency, median, IQR, range) in Microsoft Excel (Microsoft).

Results

We extracted 100 images from Google Images and seven of these were duplicates. Of the remaining 93 images, we excluded 53 images as they were not relevant and covered, but were not limited to, photographic images of people with acute pain behavior, text images, figures, and diagrams. A total of 40 images were included in our final analysis. The 40 included images presented 31 different visual aura phenotypes (Table 1) with a total of 87 EVSs (Table 2).

Table 1.

Frequency of the occurrence of the overall visual phenomena in order by highest frequency to lowest.

Overall visual aura phenotype	Number of occurrences (total, n = 40)	Percentage of images (% = n/40)
‘Bean-like’ forms like a crescent or C-shaped	5	12.5%
Flashes of bright light/unformed flashes of light/star-shaped figures	2	5%
‘Foggy’/blurred vision or “dimness”	2	5%
‘Foggy’/blurred vision or “dimness” ANDZigzag or jagged lines	2	5%
‘Foggy’/blurred vision or “dimness” AND‘Bean-like’ forms like a crescent or C-shaped	2	5%
Zigzag or jagged lines ANDScotoma	2	5%
Flashes of bright light/unformed flashes of light/star-shaped figures AND‘Foggy’/blurred vision or “dimness” ANDZigzag or jagged lines	1	2.5%
Flashes of bright light/unformed flashes of light/star-shaped figures ANDZigzag or jagged lines	1	2.5%
Flashes of bright light/unformed flashes of light/star-shaped figures ANDZigzag or jagged lines ANDGeometrical shapes	1	2.5%
Flashes of bright light/unformed flashes of light/star-shaped figures ANDGeometrical shapes AND‘Bean-like’ forms like a crescent or C-shaped	1	2.5%
Flashes of bright light/unformed flashes of light/star-shaped figures ANDCurved or circular lines	1	2.5%
‘Foggy’/blurred vision or “dimness” ANDZigzag or jagged lines ANDWhite spots ANDCurved or circular lines	1	2.5%
‘Foggy’/blurred vision or “dimness” ANDZigzag or jagged lines AND‘Bean-like’ forms like a crescent or C-shaped	1	2.5%
‘Foggy’/blurred vision or “dimness” ANDZigzag or jagged lines AND‘Tunnel’ vision ANDLike a mosaic	1	2.5%
‘Foggy’/blurred vision or “dimness” ANDScotoma AND‘Bean-like’ forms like a crescent or C-shaped	1	2.5%
Foggy’/blurred vision or “dimness” ANDColored dots	1	2.5%
‘Foggy’/blurred vision or “dimness” ANDGeometrical shapes	1	2.5%
‘Foggy’/blurred vision or “dimness” AND‘Bean-like’ forms like a crescent or C-shaped	1	2.5%
‘Foggy’/blurred vision or “dimness” AND“Like a negative of film”	1	2.5%
Zigzag or jagged lines ANDWhite spots	1	2.5%
Zigzag or jagged lines AND‘Like looking through heatwaves or water’ ANDGeometrical shapes AND‘Bean-like’ forms like a crescent or C-shaped	1	2.5%
Zigzag or jagged lines ANDGeometrical shapes AND‘Bean-like’ forms like a crescent or C-shaped	1	2.5%
Zigzag or jagged lines ANDBean-like’ forms like a crescent or C-shaped ANDHemianopsia	1	2.5%
Scotoma AND Phosphenes (small bright dots) ANDColored dots ANDBean-like’ forms like a crescent or C-shaped	1	2.5%
Scotoma ANDGeometrical shapes ANDHemianopsia	1	2.5%
Scotomata	1	2.5%
‘Like looking through heatwaves or water’ ANDCurved or circular lines	1	2.5%
Colored dots	1	2.5%
Colored dots ANDLike a mosaic	1	2.5%
Geometrical shapes ANDBean-like’ forms like a crescent or C-shaped	1	2.5%
Geometrical shapes ANDBean-like’ forms like a crescent or C-shaped ANDCurved or circular lines	1	2.5%

Table 2.

Frequency of the occurrence of single elementary visual disturbances.

Elementary visual disturbance	Number of occurrences of a single elementary visual disturbance	Percentage of images with this elementary visual disturbance (% = n/40)
‘Bean-like’ forms like a crescent or C-shaped	18	45%
Zigzag or jagged lines	16	40%
‘Foggy’/blurred vision or “dimness”	13	33%
Geometrical shapes	8	20%
Flashes of bright light/unformed flashes of light/star-shaped figures	7	17%
Scotoma	5	13%
Colored dots	4	10%
Curved or circular lines	4	10%
“Like looking through heatwaves or water”	2	5%
White spots	2	5%
Hemianopsia	2	5%
Like a mosaic	2	5%
Scotomata	1	3%
Phosphenes (small bright dots)	1	3%
“Tunnel” vision	1	3%
“Like a negative of film”	1	3%
Visual snow	0	0%
Deformed images/metamorphopsia (alteration of lines/angles)	0	0%
Oscillopsia/autokinesis (movement of stationary objects)	0	0%
Fractured vision	0	0%
Corona phenomena	0	0%
Anopia	0	0%
Macropsia (objects look larger than they really are)	0	0%
Micropsia (objects look smaller than they really are)	0	0%
Complex hallucinations	0	0%

Most visual aura images presented more than one EVS (76 out of 87, 87% of cases), median 2, IQR 1–3, range 1–4.

Discussion

This study systematically reviewed how visual aura is depicted on the most popular website in the world.

Approximately 40% of the images were correct representations of visual migraine aura. Of those, the majority presented at least two EVSs. The different visual auras were quite heterogenous, with 31 variations. There was also a heterogeneity of EVSs included in such images with 16 of them represented. The most common ones were “bean-like’ forms like a crescent or C-shaped (45%)”, “Zigzag or jagged lines” (40%), and “‘Foggy’/blurred vision or ‘dimness’”(33%). Nine EVSs from our previously published list of 25 EVSs (3) were completely absent in the representations (Table 2), which is consistent with these being less frequently reported by patients in clinical studies (3). Nonetheless, the highest-ranked images of visual aura on Google Images provides a lacking representation of the diversity of visual aura. The clinical consequence is that patients who search for visual aura online may be at risk of not recognizing their own symptoms if they have a rare phenotype and thus, ultimately, receive a wrong diagnosis. These findings are not surprising as Google Images ranks digital images by numerous factors, such as the relevance of the file name and the text surrounding the images. In other words, Google Images may represent the overall cultural understanding of what a visual migraine aura should look like rather than what it potentially can be. However, as the algorithm is not publicly available and the search terms were only in English, these speculations cannot be confirmed.

The lack of diversity may also be attributed to the fact that there is no official overview of visual aura provided by a professional society such as the International Headache Society. A potential solution could be to create a comprehensive collection of images representing the whole spectrum of scenarios of visual aura. First and foremost, this requires a study conducted on a large population of patients with migraine with aura to identify representative images. Following that, this set of images should be distributed by a professional society on a web-based platform for widespread use.

Limitations

This is a cross-sectional study of images on the internet, while new content is uploaded over time and the index ranking on Google is revised accordingly, which is an inherent limitation in studies of online content. Furthermore, we cannot exclude regional differences in a Google search; however, we believe the use of English search terms will limit potential changes.

Conclusion

The current study showed that many images related to the search term “migraine aura” are accurate portrayals of visual migraine aura symptoms; however, the images represented a limited number of phenotypes, which may contribute to underdiagnosis of rarer subtypes. The information derived from the internet photos may hinder the effective recognition of aura symptoms, thus there is a need to provide a more comprehensive representation of visual migraine aura symptoms on the internet.

Public health relevance

Many depictions of migraine aura on the internet are accurate portrayals of visual migraine aura symptoms; however, not all aura phenotypes are represented, which may contribute to underdiagnosis of rarer subtypes.

The information derived from the internet photos may hinder the effective recognition of aura symptoms, thus there is a need to provide a more comprehensive representation of visual migraine aura symptoms on the internet.