Related Experiment Video
Updated: Sep 13, 2025
05:19
Author Spotlight: Therapeutic Benefit of Closed-Loop Deep Brain Stimulation in Depression Treatment
Published on: July 7, 2023
2.5K
Predicting Pre- and Post-Diagnostic Depression in Women with Abnormal Pap Screening Tests: A Neural Network Approach
Irena Ilic1, Goran Babic2, Sandra Sipetic Grujicic3
1Faculty of Medicine, University of Belgrade, 11000 Belgrade, Serbia.
Life (Basel, Switzerland)
|July 29, 2025
Summary
This study used neural networks to predict depression in women with abnormal Pap smears. The models accurately identified women at risk, enabling timely psychological support for better adherence to diagnostic procedures.
Area of Science:
- Medical Informatics
- Psychiatry
- Gynecology
Background:
- Abnormal Papanicolaou (Pap) smear results can lead to depression, impacting women's adherence to necessary follow-up diagnostic procedures.
- Early identification of depressive symptoms is crucial for providing timely psychological support and ensuring continuity of care.
Purpose of the Study:
- To develop and validate neural network models for predicting pre- and post-diagnostic depressive symptoms in women with abnormal Pap screening tests.
- To identify key predictors of depression in this patient population.
Main Methods:
- A cohort of 172 women with positive Pap screening results underwent diagnostic procedures (colposcopy/biopsy/endocervical curettage).
- Socio-demographic data and Hospital Anxiety and Depression Scale (HADS) scores were collected before and after procedures.
- Multilayer perceptron neural networks were employed for predictive modeling.
Main Results:
- Pre-diagnostic depression was identified in 37.2% of women, with anxiety, depression (CESD), worry (POSM), and sedative use as key predictors. The predictive model achieved 79.41% accuracy (AUROC 0.842).
- Post-diagnostic depression increased to 48.3%. The model for predicting post-diagnostic depression, using HADS anxiety, place of residence, and CESD score, achieved 88.24% accuracy (AUROC 0.939).
Conclusions:
- Neural network models can effectively predict depressive symptoms in women undergoing diagnostic procedures following an abnormal Pap smear.
- These predictive capabilities can assist healthcare providers in offering targeted psychological interventions, improving patient adherence and outcomes.
