Derailment

Automated derailment classification processes clinical transcripts by evaluating single or multiple spoken or written fragments against the DER/NO-DER taxonomy. It returns structured metadata including severity scores, rationale, and optional toxicity flags for research datasets.

Derailment classification in language thought disorder assessments identifies whether speech fragments exhibit derailment (DER) or no derailment (NO-DER). It enables standardized annotation and analysis of clinical interview transcripts for diagnostic support and research.

Yes, you can annotate multiple speech fragments for derailment simultaneously by submitting them together for evaluation. The classifier processes batch fragments and returns structured results with severity scores and rationale for each DER/NO-DER determination.

Derailment classification provides standardized metadata payloads including severity scores and classification rationale alongside the DER/NO-DER determination. Optional toxicity flags are also delivered to support comprehensive language thought disorder analysis.

Automated DER/NO-DER taxonomy classification is designed for clinical transcripts, research datasets, and QA annotations. It should be used to support diagnostic workflows and standardized analysis rather than serve as a standalone clinical diagnostic tool.