TOPSY 7T psychosis cohort

TOPSY (Tracking Outcomes in Psychosis) is a longitudinal ultra-high-field MRI study run at Western University and the Lawson and Robarts research institutes in London, Ontario, led by Lena Palaniyappan. It recruits people referred to the local Prevention and Early Intervention Program for Psychoses, most of them scanned within days of first contact and before antipsychotic treatment had taken hold, together with people with established schizophrenia and matched healthy controls. Its aims are to relate glutamate and glutathione, white matter, thought disorder and later outcome. The images have not been released publicly.

Composition

The registered protocol plans first-episode, clinical high-risk and established-illness patients plus controls. The diffusion analysis by Van Dyken et al. (2024) used 112 participants after exclusions: 58 with untreated first-episode schizophrenia spectrum psychosis, 17 with established schizophrenia and 37 controls, aged about 22 to 29 on average by group. Before that, 17 people did not complete the protocol, one was dropped for data quality and four first-episode patients with mood disorders were left out. Dempster et al. (2020) analysed baseline spectroscopy in 26 first-episode patients and 27 controls. First-episode patients were planned to be rescanned during follow-up.

Acquisition

Scans used a head-only Siemens 7 T scanner at the Centre for Functional and Metabolic Mapping, Western University, described in 2020 with an 8-channel transmit, 32-channel receive head coil and in 2024 as a MAGNETOM Plus. The protocol lasts about 75 minutes and includes 0.75 mm MP2RAGE, diffusion MRI at 2 mm with 64 directions in two phase-encoding directions, resting-state fMRI, myelin-sensitive imaging and single-voxel semi-LASER proton spectroscopy (TE 100 ms) of the dorsal anterior cingulate cortex.

Annotations

There are no image labels. Each participant has a consensus DSM-5 diagnosis, PANSS-8 symptom ratings, functioning scores (SOFAS), substance use measures and, for patients, time to treatment response and follow-up outcomes.

Known limitations

• The images are not public. PsychosisBank shares only speech recordings, transcripts and clinical scores from part of the cohort, to approved non-commercial researchers.
• Patient groups are mostly male (48 of 58 first-episode patients in the diffusion sample).
• Some patients had a few days of antipsychotic exposure before the scan.
• Published papers use different subsets, so counts differ between them.