Diana Elena Vulpe

Dr. Diana Elena Vulpe

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
17 July 2025 in Journal of Orthopaedic Translation
Verified
Journal Article
Open access
Serban DragosloveanuSerban Dragosloveanu
Serban Dragosloveanu
DVDiana Elena Vulpe
Diana Elena Vulpe
CAConstantin Adrian Andrei
Constantin Adrian Andrei
DNDana-Georgiana Nedelea
Dana-Georgiana Nedelea
NGNicolae Dragos Garofil
Nicolae Dragos Garofil
Cristian ScheauCristian Scheau
Cristian Scheau
Periprosthetic joint infection (PJI) is a serious complication that can occur after joint arthroplasty, such as hip or knee replacement surgeries. It involves the invasion of the periprosthetic space by pathogens, leading to severe inflammation and often requiring complex medical intervention. PJI is associated with significant morbidity, increased healthcare costs, and a reduced quality of life for patients. This study aims to evaluate the performance of multiple supervised machine learning models in predicting PJI using clinical and demographic data collected from patients who underwent joint arthroplasty. Eight supervised machine learning models—Logistic Regression, Random Forest, XGBoost, Artificial Neural Network (ANN), k-Nearest Neighbors (KNN), AdaBoost, Gaussian Naive Bayes (GNB), and Stochastic Gradient Descent (SGD)—were trained and tested on a dataset of 27,854 patients. Models were evaluated using accuracy, precision, recall, specificity, F1 score, and area under the ROC curve (AUC). Random Forest and XGBoost showed the best overall performance, with high accuracy and balanced metrics across all evaluation criteria. KNN also performed strongly, particularly in minimizing misclassifications. GNB and SGD yielded weaker results, with higher error rates. Random Forest, XGBoost, and KNN are the most promising models for clinical implementation in PJI prediction. Their robust performance may support earlier diagnosis and improved patient outcomes in orthopedic care. This study demonstrates that machine learning models—particularly Random Forest and XGBoost—can accurately predict periprosthetic joint infection (PJI) using structured electronic health record data. By integrating these models into preoperative assessment workflows, clinicians may be able to identify high-risk patients earlier, personalize prophylactic strategies, and reduce infection-related morbidity. The implementation of these predictive tools has the potential to enhance clinical decision-making, improve surgical outcomes, and optimize the use of healthcare resources in orthopedic practice.

Verified
Journal Article
Open access
PDF file
Serban DragosloveanuSerban Dragosloveanu
Serban Dragosloveanu
BCBogdan Sorin Capitanu
Bogdan Sorin Capitanu
RJRadu Josanu
Radu Josanu
DVDiana Elena Vulpe
Diana Elena Vulpe
Romica CerganRomica Cergan
Romica Cergan
Cristian ScheauCristian Scheau
Cristian Scheau
Background: The Coronal Plane Alignment of the Knee (CPAK) classification system has been developed as a comprehensive framework delineating nine coronal plane phenotypes, based on arithmetic hip–knee angle (aHKA) and joint line obliquity (JLO). Our study aimed to assess the prevalence of knee phenotypes in the Romanian population using the CPAK classification, encompassing both osteoarthritic and healthy cohorts. Methods: We conducted an observational cross-sectional study, analyzing data from 500 knees with osteoarthritis and 500 healthy knees that met the inclusion criteria. Demographic data were collected, and radiological parameters including lateral distal femoral angle (LDFA), medial proximal tibial angle (MPTA), aHKA, and JLO were measured. Knee phenotypes were categorized using the CPAK classification. Results: In the osteoarthritic cohort, the most prevalent CPAK phenotype was type I (42.4%), characterized by varus alignment and an apex distal joint. Conversely, in the healthy population, CPAK type II, indicating neutral alignment and an apex distal joint, was the most prevalent phenotype (39.0%). CPAK types VII, VIII, and IX were rare. Conclusions: Our findings demonstrate similarities in knee phenotypes compared to other populations, with some minor differences and particularities. The CPAK classification proves to be a valuable tool in assessing knee tyalignment.

30 July 2025 in Biomedicines
Verified
Journal Article
Open access
PDF file
DVDiana Elena Vulpe
Diana Elena Vulpe
Catalin AnghelCatalin Anghel
Catalin Anghel
Cristian ScheauCristian Scheau
Cristian Scheau
Serban DragosloveanuSerban Dragosloveanu
Serban Dragosloveanu
Oana SandulescuOana Sandulescu
Oana Sandulescu
Periprosthetic joint infections (PJIs) represent one of the most problematic complications following total joint replacement, with a significant impact on the patient’s quality of life and healthcare costs. The early and accurate diagnosis of a PJI remains the key factor in the management of such cases. However, with traditional diagnostic measures and risk assessment tools, the early identification of a PJI may not always be adequate. Artificial intelligence (AI) algorithms have been integrated in most technological domains, with recent integration into healthcare, providing promising applications due to their capability of analyzing vast and complex datasets. With the development and implementation of AI algorithms, the assessment of risk factors and the prediction of certain complications have become more efficient. This review aims to not only provide an overview of the current use of AI in predicting PJIs, the exploration of the types of algorithms used, and the performance metrics reported, but also the limitations and challenges that come with implementing such tools in clinical practice.

21 February 2025 in Symmetry
Verified
Journal Article
Open access
PDF file
Serban DragosloveanuSerban Dragosloveanu
Serban Dragosloveanu
BCBogdan Sorin Capitanu
Bogdan Sorin Capitanu
MMMihnea Moise
Mihnea Moise
DVDiana Elena Vulpe
Diana Elena Vulpe
RJRadu Josanu
Radu Josanu
Cristian ScheauCristian Scheau
Cristian Scheau
Total hip arthroplasty (THA) is commonly performed for conditions like osteoarthritis, avascular necrosis, hip fractures, and hip dysplasia. Complications following THA can result in major proximal femoral defects (MPFD), which are challenging to manage. Megaprostheses provide a solution, offering stability and early weight-bearing, though they present their own challenges. However, maintaining hip symmetry with megaprostheses is challenging in accurately restoring femoral offset, acetabular orientation, and soft tissue tension. These factors can contribute to instability, altered biomechanics, and functional impairment. Therefore, achieving hip symmetry through femoral and acetabular offset, limb length discrepancy (LLD), and center of rotation (COR) is an important consideration. Achieving proper symmetry may improve gait, quality of life, and overall outcomes. This case series includes 10 patients with non-tumoral pathology treated with megaprostheses for complications after THA between 2014 and 2024. Megaprosthesis conversion was performed for sepsis (n = 2), aseptic loosening (n = 4), and periprosthetic fracture (n = 4). Dislocation occurred in 20%, with one case showing restored hip symmetry but dislocating due to trauma and noncompliance. Another case had an LLD of 3 cm, contributing to dislocation. Periprosthetic infection was observed in 20%, managed with a two-stage approach. Although hip symmetry was not fully restored in all patients, strict follow-up and rehabilitation were essential for improving functionality and quality of life. Megaprosthesis is an effective solution for MPFD in non-oncological patients when performed correctly. Restoring hip symmetry is challenging and depends on patient care and compliance, though it is an important factor influencing outcomes.

29 May 2026 in AI
Verified
Journal Article
Open access
PDF file
Catalin AnghelCatalin Anghel
Catalin Anghel
AAAndreea Alexandra Anghel
Andreea Alexandra Anghel
MCMarian Viorel Crăciun
Marian Viorel Crăciun
ACAdina Cocu
Adina Cocu
DVDiana Elena Vulpe
Diana Elena Vulpe
Romica CerganRomica Cergan
Romica Cergan
Large Language Models (LLMs) are increasingly used for rubric-based assessment, but reliable automated grading requires more than a single prompt-response step. This study presents GradeAgentOps, a verification-first framework for evidence-anchored LLM exam grading that combines strict grading contracts, deterministic verification and canonicalization, bounded semantic repair, optional memory modules, and provenance-aware logging. The framework is evaluated on a university-level dataset of 1000 short open-ended exam responses from 100 students across 10 questions, with annotations from two independent expert human graders. A controlled ablation protocol compares six configurations, including a rubric-only baseline and progressively stronger variants with repair and memory augmentation. Human–human agreement provides the reference context, with overall ICC(2,1) = 0.678 and QWK = 0.678 between the two graders. Using one expert grader as the operational reference, the FULL configuration achieves the strongest model-human agreement (MAE = 1.935, RMSE = 2.500, QWK = 0.652, Within ±2 = 0.667), with the consistency memory configuration with repair (C1) emerging as the closest alternative. The gains are not uniform: improvements are more pronounced on argumentative items than on technical ones, and pairwise comparisons show that consistency memory contributes more strongly than rubric memory alone. The stronger configurations also produce cleaner accepted outputs, with lower rates of evidence-related postprocess issues, although at a moderate operational cost. Overall, within this controlled single-course evaluation setting, the results show that reliable automated grading benefits not only from model capability, but also from pipeline design choices that promote verification, evidential coherence, and stable grading behavior.

12 October 2024 in Infectious Disease Reports
Verified
Journal Article
Open access
PDF file
DNDana-Georgiana Nedelea
Dana-Georgiana Nedelea
DVDiana Elena Vulpe
Diana Elena Vulpe
GVGeorge Viscopoleanu
George Viscopoleanu
ARAlexandru Constantin Radulescu
Alexandru Constantin Radulescu
AMAlexandra Ana Mihăilescu
Alexandra Ana Mihăilescu
Serban DragosloveanuSerban Dragosloveanu
Serban Dragosloveanu
Objective: We present the case of a 26-year-old male with severe spinal tuberculosis of the thoracolumbar region. The patient suffered from worsening back pain over five years, initially responding to over-the-counter analgesics. Despite being proposed surgery in 2019, the patient refused the intervention and subsequently experienced significant disease progression. Methods: Upon re-presentation in 2022, mild involvement of the T12-L1 vertebrae was recorded by imaging, leading to a percutaneous needle biopsy which confirmed tuberculosis. Despite undergoing anti-tuberculous therapy for one year, the follow-up in 2024 revealed extensive infection from T10 to S1, with large psoas abscesses and a pseudo-tumoral mass of the right thigh. The patient was ultimately submitted to a two-stage surgical intervention: anterior resection and reconstruction of T11-L1 with an expandable cage, followed by posterior stabilization from T8-S1. Results: Postoperative recovery was uneventful, with significant pain relief and no neurological deficits. The patient was discharged on a continued anti-tuberculous regimen and remains under close surveillance. Conclusions: This paper presents details on the challenges of diagnosis and management of severe spinal tuberculosis, with emphasis on the importance of timely intervention and multidisciplinary care.