There is a growing number of studies showing that analysis of macular parameters provides additional information about ganglion cell loss in glaucoma and compliments traditional markers of glaucoma. In this paper, we develop an image processing pipeline for the macular thickness deviation maps generated by Heidelberg Spectralis optical coherence tomography (OCT) to evaluate the information within the macular measurements for diagnosing glaucoma. Logistic regression is applied to analyze features extracted from the deviation maps and the strength of their relationship with the diagnosis of glaucoma. Our experimental results show that the proportion of regions with thickness significantly below the normative range in the nerve fiber layer thickness deviation maps can be used to detect glaucoma with an accuracy of up to 70.16%. Moreover, the ganglion cell layer deviation maps also possess significant diagnostic ability for detection of glaucoma, which can be combined with the traditional assessments of the optic nerve head and peripapillary retinal nerve fiber layer to improve the reliability of glaucoma diagnosis.