Additional investigations involving larger sample size of and more diverse participants (e.g. ones addictive to other substance and in different genders) are needed for generalizing our results, examining the relationship between behavioral and imaging data, and clarifying the effects of different treatment stages of addiction on functional connectivity. In particular, the present study did not dissociate the effects of methadone treatment on resting-state functional connectivity from those of heroin addiction, so further studies in this regard is needed; such a study may be of potentially clinical significance. Meanwhile, methods that could address potential noise due to low-frequency fluctuations from possible thermal noise, scanner instability, and alias of cardiac and/or respiratory cycling, more advanced analysis, such as independent component analysis (Greicius et al., 2007) and effective connectivity analysis (Fox and Raichle, 2007), and combination with diffusion tensor imaging (Greicius et al., 2009) may be used in further studies.
