The aim of the present study was to investigate the relationships between the automated bone scan index (aBSI) and skeletal-related events (SRE) in breast cancer patients with bone metastasis. A computer-aided software (BONENAVI™) that was developed using an Artificial Neural Network (Artificial Neural Network) was used for the present analysis. Forty-five patients diagnosed with bone metastasis due to breast cancer from April 2005 through March 2013 were retrospectively analyzed. Before and after the time of initial treatment, aBSI, Artificial Neural Network score, and hotspot number were calculated, and the relationships between these scores and SRE were analyzed. Twenty cases showed decreased (improved) aBSI values after initial treatment (Group A), and 25 cases showed unchanged/increased (worsened) aBSI values (Group B). Chi-square analysis revealed a significant difference in incident numbers of SRE between the two groups--one case in Group A and 12 in Group B (P<0.001). Event-free survival was significantly shorter in Group B (hazard ratio: 8.31, 95% CI: 1.33-12.14, log-rank test; P<0.05). The groups were also divided by the results of 2 radiologists' visual scan interpretations, and no significant differences were shown in the number of SRE (P=0.82, P=0.10). When correlation analyses were performed between aBSI and bone metabolic or tumor markers, alkaline phosphatase was significantly correlated with aBSI at the time of initial treatment (R=0.69, P<0.05). In conclusion, aBSI is proposed as a useful and objective imaging biomarker in the detection of breast-cancer patients with bone metastasis at high risk of SRE.