A Neural Network Approach to Comparing AMPD and Object Relations Theory for Personality Disorder Assessment.

Area of Science:

• Psychiatry and Psychology
• Personality Disorders Research
• Computational Psychiatry

Background:

• Limited comprehensive comparisons exist between Object Relations Theory (ORT) and the Alternative Model for Personality Disorders (AMPD).
• Investigating their predictive accuracy in diverse clinical populations is crucial for advancing personality disorder assessment.

Purpose of the Study:

• To compare the predictive accuracy of AMPD and ORT in identifying personality psychopathology.
• To evaluate the utility of neural network models in distinguishing clinical and non-clinical groups based on these theoretical frameworks.

Main Methods:

• Utilized a mixed sample of 639 participants (non-clinical and psychiatric inpatients).
• Employed Persian translations of the Level of Personality Functioning Scale-Self-Report (LPFS-SR), Personality Inventory for DSM-5 (PID-5) for AMPD, and Structured Interview of Personality Organization-Revised (STIPO-R) for ORT.
• Applied neural network models and Receiver Operating Characteristic (ROC) analysis for predictive accuracy assessment.

Main Results:

• Significant differences in both AMPD and ORT measures were found between clinical and non-clinical groups.
• Neural network models achieved over 65% accuracy in predicting group membership, with AMPD (66%+) slightly outperforming ORT (65%+).
• ROC analysis indicated high sensitivity for both models, with Area Under the Curve (AUC) values between 0.79 and 0.94.

Conclusions:

• Both AMPD and ORT demonstrate significant utility in the assessment and diagnosis of personality disorders.
• Neural network models show promise for early identification and classification of personality psychopathology using these frameworks.