Anterpreet Dua

Dr. Anterpreet Dua

orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 January 2023 in Cureus
Verified
Journal Article
Open access
JOJungmin On
Jungmin On
JGJavier J. Polania Gutierrez
Javier J. Polania Gutierrez
MPMiguel Plaza-Lloret
Miguel Plaza-Lloret
ADAnterpreet Dua
Anterpreet Dua
ZSZhuo Sun
Zhuo Sun
Tarlov cysts are extradural meningeal cysts with a collection of cerebrospinal fluid that most often affects sacral nerve roots, causing chronic low back pain and radiculopathy. Still, there is no consensus regarding the best treatment for symptomatic cysts. We describe a patient who developed worsening lower back pain and radiculopathy after interventional drainage and surgical management. Medication and various procedures failed to relieve the pain. Subsequently, the patient received significant pain relief from spinal cord stimulation (SCS). This case provides evidence that SCS could be used to manage refractory pain from Tarlov cysts that have failed to respond to other treatment modalities.

1 January 2017 in International Journal of Anesthesia and Relaxation
Verified
Journal Article
Open access
ADAnterpreet Dua
Anterpreet Dua

21 November 2026 in Journal of Personalized Medicine
Verified
Journal Article
Open access
PDF file
AMAHMAD MIRZA
AHMAD MIRZA
ADAnterpreet Dua
Anterpreet Dua
PSParamvir Singh
Paramvir Singh
ZSZhuo Sun
Zhuo Sun
Introduction: Pain is a leading cause of disability because of its subjective, multidimensional and heterogeneous nature. Artificial intelligence (AI), encompassing machine learning (ML), deep learning (DL) and natural language processing (NLP) can integrate multimodal data to support personalized pain care. Personalized pain medicine is defined as the use of patient-specific characteristics to guide individualized assessment, risk stratification and treatment selection. Methods: A structured literature review was conducted in PubMed/MEDLINE, Scopus, Web of Science and Google Scholar (January 2020 to 30 June 2026), supplemented by citation searching for foundational studies. Findings were evaluated descriptively because of heterogeneity in populations, algorithms, outcomes and validation methods. A five-stage evidence-maturity framework was applied to classify applications. Results: Most published work remains at Stages 1–2 (model development and independent validation). Automated facial-pain assessment can achieve high internal performance (F1 91–96%) that declines markedly with subject-independent validation (F1 40–67%). A postoperative external test set of 120 images, exact 0–10 score prediction reached 53%, with a reported mean error of 2.4 points. A Random Forest model predicted postoperative opioid-requirement categories with 72% accuracy, compared with 71% for multinomial logistic regression; sensitivity was 89% for lower and 43% for higher requirements. The small accuracy difference, incomplete calibration and class-distribution reporting limit claims of incremental value. Discussion and Conclusions: AI shows technical promise for automated assessment and risk prediction, but high predictive metrics do not equate to clinical utility. Translation requires external multicenter validation and calibration assessment under clinician supervision and demonstration of improved patient outcomes.