Table 1.
Profile of sample groups.
Fig 1.
Schematic of functional magnetic resonance imaging (fMRI) data acquisition using an event-related design.
We used an event-related design for the fMRI. Scans of 22 slices generated equal numbers of scan noises during a single repetition time (TR) of 2992 ms, and the same number of tones with a stimulus onset asynchrony (SOA) of 136 ms (TR/slice number = 2992/22) were presented. The presentation of tones was completely synchronized to scans by triggers generated by the MRI system in order to minimize the effect of scan noise on fMRI and electroencephalogram (EEG) raw data, which were also used to segregate and average EEG epochs in the ERP analysis. The SOA for event scans including the omission events and tones in the regular positions represent 7 or 8 scans (20,944 or 23,936 ms). Fifty-three event scans were assorted at random among the 247 baseline scans including baseline tones only, with no omission. Total scan time = 2992 ms x 300 scans = 897,600 ms. Event: scan including omission 53 times, indicated by red crossbars. Baseline: scan including only frequent stimuli 247 times, indicated by black crossbars. Frequent stimuli: 6547 trials, 3000 Hz, duration 35 ms, rise/fall 0/5 ms. Rare (omission) events: 53 trials (probability: 53/6600 = 0.00803).
Fig 2.
Grand average waveform of controls.
Omission events elicited an event-related potential (ERP) component recognized as mismatch negativity (MMN) around latency ranges of 120–185 ms in controls.
Fig 3.
Grand average waveforms of patients.
Omission events elicited mismatch negativity (MMN) around latency ranges of 120–185 ms, in patients as well as in controls.
Fig 4.
Grand average of functional magnetic resonance imaging (fMRI) in controls.
Controls exhibited significant activation responding to omission in the left superior temporal gyrus (Brodmann area/BA 41).
Table 2.
Areas significantly responded to sound omission or correlated to other measures.
Fig 5.
Comparison of functional magnetic resonance imaging (fMRI) between patients and controls.
Patients with schizophrenia had reduced blood oxygenation level-dependent (BOLD) activity in the left middle temporal gyrus (Brodmann area/BA 39) compared to controls in a t-test with age and education as covariates.
Fig 6.
Correlation of functional magnetic resonance imaging (fMRI) with mismatch negativity (MMN) in patients.
Blood oxygenation level-dependent (BOLD) activity in the right superior temporal gyrus significantly correlated with MMN amplitudes at the Fz site.
Fig 7.
Correlation of functional magnetic resonance imaging (fMRI) with the PANSS in patients.
Blood oxygenation level-dependent (BOLD) activity in the right parahippocampal gyrus correlated with the Positive and Negative Syndrome Scale (PANSS) positive symptom subscale scores.
Fig 8.
Correlation of functional magnetic resonance imaging (fMRI) with Bell’s PANSS classification.
(a) Blood oxygenation level-dependent (BOLD) activity in the right insula and (b) right parahippocampal gyrus significantly correlated with the Positive and Negative Syndrome Scale (PANSS) hostility and positive symptom subtotal scores, respectively, as classified by Bell’s classification.
Fig 9.
ROI (region of interest) analysis of the temporal gyrus in the middle temporal gyrus.
Patients demonstrated reduced signal change (t-values) in the bilateral middle temporal gyrus compared to controls.
Fig 10.
Correlation analysis of BOLD signal changes in the left Heschl’s Gyrus (HSC).
In a correlation analysis using blood oxygenation level-dependent (BOLD) signal change (%) in the HSC as a covariate in ROI analysis, BOLD activity in the left insula encompassing the superior temporal gyrus (not shown), right insula (indicated by the white circle in (a)), right postcentral gyrus (indicated by the white circle in (b)), and left inferior parietal lobule (indicated by the white circle in (c)) significantly correlated with BOLD signal changes in the left HSC.
Fig 11.
Correlation analysis of BOLD signal changes in the right Heschl’s Gyrus (HSC).
(a) Blood oxygenation level-dependent (BOLD) activity in the right insula, (b) left posterior cingulate, and (c) left middle frontal gyrus significantly correlated with that of the right HSC of patients.
Fig 12.
Regions of significant fractional anisotropy (FA) reduction in patients with schizophrenia relative to controls (red-yellow, p < 0.05, corrected for family-wise error).
(a) Axial slices from Z = 24 to Z = 38 mm and (b) a coronal slice show the bilateral anterior cingulum region along with the right posterior cingulum. The left-right orientation is in accordance with the radiological convention.
Fig 13.
Fractional anisotropy (FA) values in the anterior cingulate gyrus.
Patients had reduced FA values in the left anterior cingulate gyrus.