Corrigendum to “TRIMAGE: A dedicated trimodality (PET/MR/EEG) imaging tool for schizophrenia” [Eur Psychiatry 50 (2018) 7–20]

This Corrigendum for the article ‘TRIMAGE: a dedicated trimodality (PET/MR/EEG) imaging tool for schizophrenia’ by A. Del Guerra and colleagues, concerns the analyses and results of the F-DOPA PET data assessed in Munich. Particularly, it concerns the section ‘Results’, subsection ‘3.1.2. Results of the pilot study: analysis of Munich data’ and corresponding Figure 7, as well as parts of the Discussion.

The authors noted that in the analysis of Munich F-DOPA PET data, influx rate constants (Kc) were wrongly calculated. Specifically, in the calculation of the integral used in the Patlak graphical analysis, PET frames were treated as equidistant, which led to higher than normal Kc values. We recalculated the Kc values with the correct integral (taking into account different frame lengths) and ran several control analyses to assess their reliability.

In detail, the corrections affect the following parts of the article:

1) Results, section ‘3.1.2. Analysis of the Munich data’: in the last 2 paragraphs, the sentences “Schizophrenic patients showed …. or other striatal sub-regions.” have to be replaced by the following: “Schizophrenic patients showed decreased Kc values in the caudate nucleus (0.0103 ± 0.001 min⁻¹) compared to healthy controls (0.0132 ± 0.001 min⁻¹) (p < 0.001) (Figure 7, upper left). Similarly, a decrease of Kc values was found in the functional executive region (Figure 7, lower left), which largely overlaps with the anatomical region of the caudate, for schizophrenic patients (0.0120 ± 0.001 min⁻¹) compared to controls (0.0145 ± 0.001 min⁻¹) (p = 0.002). Additionally, patients showed decreased Kc values in the putamen (Figure 7, center upper part) (0.0131 ± 0.001 min⁻¹) compared to healthy controls (0.0152 ± 0.002 min⁻¹) (p = 0.017), and the functional correspondent in the sensorimotor striatum (Figure 7, center lower part) for patients (0.0119 ± 0.002 min⁻¹) compared to healthy controls (0.0160 ± 0.002 min⁻¹) (p < 0.001). Finally, a decrease in Kc values was found for the whole striatum (Figure 7, right) in patients (0.0120 ± 0.001 min⁻¹) compared to healthy controls (0.0144 ± 0.001 min⁻¹) (p = 0.002). No differences were found for the accumbens nucleus (p = 0.71) and the corresponding functional subdivision, namely the limbic striatum (p = 0.65).”

2) Figure 7 of the article has to be replaced by the following new Figure 7:

3) In the section ‘4. Discussion and conclusions’, second paragraph, the sentence “The evaluation shows that group differences between healthy controls and schizophrenic patients are well replicable.” has to be deleted, since these findings are rather novel and surprising and therefore do not directly replicate previous results.