Plate 1: Overlay of the most liked images on Instagram mentioning cancer between 8 and 15 April 2019.

The "like" is an intuitive measure of the visibility for an image in social media. Images in the study went from a couple likes to over 5,000. Those with the most (above 1,000) included group pictures, motivational posters, pictures of patients at their last chemotherapy session, or portraits at the hospital. Most subjects are smiling, and the colours are bright and calm. Of the 15 images studied that had more than 1,000 likes, 5 of them were labelled with negative feelings, while 8 were perceived positively, and 2 were mixed. Overall, respondents felt represented by the images; they identified with the situations pictured and expressed being proud of the people in them.

Plate 2: Overlay of some of the least liked images on Instagram mentioning cancer between 8 and 15 April 2019.

Of the least "liked" images (those with fewer than 200 likes in the set), only 3 do not include a person in frame. The images show greater variance in the facial expression of the subject, with the least liked images featuring portraits of patients visibly fatigued. Only two images in this group are perceived negatively by respondents, who otherwise mention feelings of calmness, strength and courage when looking at them. The colour blue makes a stronger appearance through the tints in the hospital rooms.

Plate 3: Overlay of the images that generated the most positive emotional response.

For the 112 images in the data set, participants were asked to rate 56 according to opposite pairs of emotions: sad vs happy, lonely vs accompanied, etc. This panel reflects the images that respondents reacted to with mostly positive feelings (or, to put it more clearly: the most positive images). 70% of the images surveyed inspired mostly positive feelings in the respondents. Among the most common positive feelings are those of pride, bravery, energy and strength.

Plate 7: Overlay of patients pictured at a hospital room on Instagram.

Images that feature hospitals or medical equipment are common in the data set, naturally. While many of these were marked as having a negative impact, some of the most positively perceived photographs were also taken in hospital, suggesting that viewers are not immediately affected by the presence of the hospital in the image. Instead, the attitude and perceived situation of the patient seems to play a bigger role than the context of the picture.

Plate 5: Overlay of portraits and selfies of cancer patients on Instagram.

Of the 112 images surveyed, only 15 of them did not contain a person in frame. Portraits, selfies and group pictures dominate the set, with certain similarities in colour and composition. Most of those without people are memes or informational posters. The psychological perception of images with people varies based on the context and the expression of the person pictured, but also depending on the elements that surround them. Four variables stand out: the smile and the posture of the patient, whether they are in company, and external elements such as plush toys.

Plate 9: Overlay of images taken outdoors by cancer patients on Instagram.

The outdoors is often pictured indirectly, simply as the scene of the photograph. In some cases, however, elements such as the forest or the sea are used as metaphor of freedom or an expression of calmness, especially by patients who have recently completed their last chemotherapy session. The results are effective: all the pictures taken outdoors were labelled as emotionally positive. Respondents indicated a feeling of freedom (very clear, for instance, in an image picturing a woman at the beach), and felt energised, proud and courageous.

Plate 11: Overlay of images from Instagram that mentioned cancer and with which respondents felt identified.

Respondents identified the most with portraits and group images. Where they identify with the picture, they also tend to indicate a feeling of pride and courage, suggesting they feel connected to the patients through their experience accompanying their family or participating in awareness-raising campaigns. These are also images that are perceived positively, suggesting an alignment between representation and positive emotional impact.

Plate 12: Images patients don't align with.

The final group of images are those that respondents could not align with. This is the most diverse group in terms of content, but it presents some clear traits: there are fewer portraits and the images contained are more explicit about the illness. A man holding a sign reading "cancer free", a woman undergoing treatment at home, a woman showing her evolution from diagnosis to regaining her strength, or patients in bed at the hospital. Results suggest that respondents had not been through that experience themselves and thus could not identify with those pictures. These images also show the male and female body more clearly. A young woman who'd suffered a double mastectomy, two muscular men showing their evolution, or a cake with the shape of breasts are amongst the images labelled as least "representative", most likely due to the respondent being of the opposite gender, but perhaps also as a reaction to the use of the body as part of the discourse of illness. Of the 16 images labelled as non-representative, 8 were perceived negatively, 2 with mixed feelings, and 6 positively. Of those perceived negatively, images of uncertainty (showing patients waiting and alone) and demotivational posters generated the most anger in respondents.