Model appears to outperform ChatGPT and six previously published large text-embedding models<\/p>\n<\/h3>\n
By Elana Gotkine HealthDay Reporter<\/b><\/p>\n
<\/b><\/p>\n
TUESDAY, April 16, 2024 (HealthDay News) — A text-embedding-ada-002 (ADA) machine learning model can identify posttraumatic stress disorder following childbirth (CB-PTSD) from maternal childbirth narratives, according to a study published online April 11 in Scientific Reports<\/em>.<\/p>\n Alon Bartal, Ph.D., from Bar-Ilan University in Ramat Gan, Israel, and colleagues examined the effectiveness of ChatGPT and the ADA model for detecting CB-PTSD in a sample of 1,295 women who gave birth in the last six months and were 18 years or older by analyzing maternal childbirth narratives. CB-PTSD was assessed using the PTSD Checklist for the Diagnostic and Statistical Manual of Mental Disorders<\/em>, Fifth Edition.<\/p>\n The researchers found that CB-PTSD was identified via narrative classification by developing a machine learning model that utilizes numerical vector representation of the ADA model. The model had an F1 score of 0.81 and outperformed ChatGPT and six previously published large text-embedding models trained on mental health or clinical domain data.<\/p>\n “This textual personal narrative-based assessment strategy employing natural language processing analysis has the potential to become an accurate, efficient, low-cost, and patient-friendly strategy for identifying CB-PTSD in the clinic, and facilitating timely interventions to mitigate this maternal mental health disorder,” the authors write.<\/p>\n