A literature review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines ([PRISMA]; Moher et al., 2009). In June 6, 2020, a search was conducted in the Scopus, Web of Science, PsycINFO, and ScienceDirect databases to investigate studies that applied body image software using images of the entire body.

Exclusion criteria were defined to filter out: 1) articles duplicated between databases, 2) reviews and meta-analysis articles, 3) theoretical articles, book chapters, conference presentations, comments, dissertations, and theses, 4) articles unrelated to the topic, 5) articles that were not peer-reviewed, 6) articles without full access to online text, 7) articles published in languages other than English, Spanish and Portuguese, and 8) articles that used measures not considered body image software, such as self-report scales, computer-only measurement versions of software programs not programmed for body image evaluation (e.g., image editing).

The first step of the review consisted of a search in the aforementioned databases, considering a publishing interval from 2000 to 2020. This extended period was defined to capture eventual technological transitions from the beginning of body image software research to recent publications. The search terms and their recombination were: [“body image" OR “body perception” OR “body size” OR “body awareness”] AND ["software" OR "assessment"]. These terms were chosen based on the terminology indexed by the Thesaurus-APA. The term "human body" was also indicated by the Thesaurus; however, it was omitted because results using this term referred to articles with content other than body image assessment.

The results were uploaded to Rayyan QCRI (Ouzzani et al., 2016), a web application that supports the integration of results from systematic searches among different databases. Articles duplicated between databases were excluded, and the resulting list was screened (identification). Two judges independently reviewed the abstracts on this list, following exclusion criteria 2, 3, and 4 (eligibility screening). Next, the full text of the remaining articles was examined, and exclusion criteria 5 to 8 were applied in the analysis (included). In cases of doubt, a third judge carried out an analysis. In addition to the resulting list of articles included, additional records were searched based on cross-reference checking, considering only citations of the original articles that introduced the software programs identified in this review. Two software programs did not disclose their original reference, and one software program referenced a presentation at an annual meeting in Germany. In these cases, all articles that used the software identified by the review were included in the citation analysis. Cross-reference checking for this set of articles was carried out on the Google Scholar platform. Results of this selection process are presented in Figure 1.

Figure 1

The first search in the databases retrieved 2487 records and the Google platform cross-reference search included another 23. Following the exclusion criteria, 35 articles were selected for inclusion in the review. In total, 13 articles referred to software development, two of which only presented the software, and the other 11 presented and tested the body image software. The remaining articles were applications of these software programs in different populations. The majority of the articles reviewed were conducted at universities in the United States (31.4%) and published after 2010 (62.8%). The original field of research of the journals publishing these studies was mainly the broad scope of Psychology. However, the present study did not delimit this area of research as an inclusion criterion. In addition to Psychology, journals encompassed areas such as health sciences, sports, aesthetics, and combined areas of psychology and computer science. Information regarding the software name and articles that applied each program is presented in Table 1.

Table 1

Most software programs allow the simultaneous assessment of more than one dimension of body image, with two programs evaluating only the perceptual dimension of body recognition. Of the 543 citations linked to the 13 software programs identified, the program was applied to 23 studies, representing 4.2% of the total citation index. Figure 2 shows the time frame of software program citation and application between 2000 and 2020. The Body Size Distortion Program was the most applied software, followed by the Anamorphic Micro Software and the Body Image Assessment Software. All the programs used a depictive method to evaluate the body image dimensions. In general, the main technique used among the software programs was widening and thinning the images displayed by means such as the morphing technique, as shown in Figure 3. Only the VR Standardized 3D Avatars software did not have its technique described in the articles reviewed.

Figure 2

Note. Search was conducted in June of 2020

Of the 35 articles, 11 studies made comparisons between the applied software and body image self-report measures. Eight different software tools were used in these comparisons. However, no article compared more than one software program in the same study. These articles used a total of 16 different body image self-report instruments. In addition, two studies evaluated how experimental tasks affected the responses to the body image software.

Gardner and Brown (2010) compared the BSDP to the Body Image Assessment Scale-Body Dimensions scale drawings (Gardner et al., 2009), which evaluates body image perception and dissatisfaction with the body image. According to the authors, the scale was less accurate in evaluating the body image distortion and dissatisfaction of male and female undergraduate students than the software program. The authors found that the participants overestimated body size more robustly in the self-report measure than in the body image software and presented significantly higher levels of dissatisfaction. This software was also used by Hagman et al. (2015), presenting similar results when compared to the Eating Disorder Inventory-2-Body Dissatisfaction scale ([EDI-2]; Garner, 1991). Both measures evidenced that women with anorexia had significantly greater body dissatisfaction than controls, with large effect sizes.

Hagman et al. (2011) also used the BSDP to investigate the effect of Risperidone on the body image of anorexic women. The study tested body image using this software along with the EDI-2 and the Color-A-Person-Test ([CAPT]; Wooley & Roll, 1991), a scale of frontal and lateral contours of a female adolescent figure used to assess body image dissatisfaction. The results did not show significant differences between anorexic and placebo groups in the EDI-2, CAPT, and the software outputs. There were also no significant differences in body image distortion assessed by the software. Cornelissen et al. (2013) applied the BSDP to low and average BMI women along with the Body Shape Questionnaire ([BSQ]; Cooper et al., 1987). The results identified significant positive correlations between overestimation of the body size output from the BSDP and eating and body shape concerns evaluated by the BSQ. However, there was no correlation between the BMI and the degree of participants’ overestimation.

Yamamotova et al. (2017) used the AMS along with the Body Attitude Test questionnaire ([BAT]; Probst et al., 1995) to assess body image in women patients diagnosed with anorexia and bulimia and a control group with no diagnosis. The BAT questionnaire represents a mixture of perceptual and emotional attitudes toward one’s body. Results from the two instruments showed that the BAT was, in all groups, more related to body dissatisfaction than to body image perception. More specifically, the BAT and the software perceptual dimension were positively correlated in the patients with bulimia and the controls, but not in those with anorexia.

The BIAS software was applied together with the figure rating scale Body Image Assessment-Revised ([BIAS-R]; Beebe et al., 1999), the EDI-2, and the BSQ for female university students and female patients with an eating disorder (Ferrer-García & Gutiérrez-Maldonado, 2008). Positive and significant correlations were observed between the BIAS and BIA-R in the eating disorder and control groups. The body dissatisfaction scores of the software positively correlated with the self-report scales’ dissatisfaction scores, indicating that all participants would have liked to be thinner. However, the measures could not discriminate between the eating disorder and control groups.

Caspi et al. (2017) used the CABI software along with the EDI-2 and the Contour Drawing Rating Scale ([CDRS]; Thompson & Gray, 1995) with women with eating disorders (restrictive anorexia, purgative anorexia, and bulimia) and healthy women. The patients scored significantly higher than healthy participants in the perceptual and affective dimensions of the CABI and EDI-2. The CDRS discriminated only patients with bulimia from controls in the perceptual and affective dimensions. Only the CABI evidenced differences in the perceptual and affective dimensions between the three groups with eating disorders and the healthy participants. These differences were maintained when controlling for age, years of education, BMI, depression, and anxiety. The other instruments only maintained the differences between the groups when controlling age, years of education, and BMI.

The body image of women with and without eating disorders was investigated by Roy and Forest (2007) using the Q-BID software, the Eating Disorder Examination Questionnaire ([EDE-Q]; Fairburn & Beglin, 1994) and the Physical Appearance State and Trait Anxiety Scale ([PASTAS]; Reed et al., 1991). The control group participants were adolescents. The clinical group consisted of two groups diagnosed with restrictive anorexia, one composed of hospitalized patients and the other of patients undergoing outpatient treatment. The software program was positively correlated with the other body image measures in the anorectic samples; however, not in the control group. Body image distortion was positively correlated with the global EDE-Q score, shape concern, and weight concern subscales and with the global score of the PASTAS only in the anorectic samples.

Stewart et al. (2009) evaluated the BMA 2.0 together with the silhouettes scale Body Image Assessment-Obesity ([BIA-O]; Williamson et al., 2000) and the Body Satisfaction Scale questionnaire ([BSS]; Slade et al., 1990) in Caucasian men and women and African-American men and women. The results showed positive correlations between the dimensions of dissatisfaction and body size estimation of the BMA 2.0 and the corresponding BIA-O subscales. However, concerning the BSS, there was a selective positive and significant correlation between the software dissatisfaction index for women only and a positive correlation trend for men.

The body image of lean, obese, and lean regular exercise participants was evaluated by Johnstone et al. (2008) using the NMT together with the Body Image Ideal Questionnaire (Cash & Szymanski, 1995) and the traditional Figural Stimuli Questionnaire Silhouette Scale (Stunkard et al., 1983). The results evidenced a significant and positive correlation among the body image dissatisfaction instruments in all three groups. The correlation was higher between the NMT and the Figural Stimuli Questionnaire.

Fisher et al. (2020) compared the VR-3D with the paper-based Figure Rating Scales ([FRS]; Sala et al., 2012) in female adolescents with anorexia. The results showed that the participants overestimated their own body size regardless of the assessment tool used. Body perception and body dissatisfaction did not differ significantly between the FRS and the software.

In addition to the correlational analysis, two other studies analyzed the influence of experimental tasks on the body image evaluated by the software program. Arciszewski et al. (2012) used the BSD to examine the effect of a threat aroused by the perception of thin-ideal images combined with beliefs about the malleability of the body-on-body image perception and dissatisfaction in female psychology students. The results showed that the women had greater body self-discrepancies when confronted with threatening thin ideals, regardless of their body mass index. Owen and Spencer (2013) used the BIAS to verify whether the visualization of healthy and lean body models influenced the subjective body ideal in female university students. The results indicated that the subjective body ideals increased after viewing images of healthy weight models.