On behalf of the Scholarship Committee and Officers of the International College of Surgeons - United States Section we are pleased to announce the winners of the 2025 Research Scholarship Competition.
First Place
Mert Marcel Dagli, MD
Developing AI Infrastructure Through the Utilization of NLP with LLM to Advance Clinical Data Extraction and Postoperative Billing
Second Place
Nicole Marker, BA
Surgical Treatment of Cardiac Tumors: A Single Center Experience
Third Place
Lauren Cox, BA
A Comprehensive Review of Global Surgery Programs Across the United States
We congratulate all the participants in our competition for the work and time they invested to participate and hope they will continue to remain active with the International College of Surgeons - United States Section throughout their training and beyond.
Keep scrolling to see their abstracts...
We also wish to congratulate the following winners of our International Scholarship Competition.
First Place
Adiveeth Deb, MBBS, MS
Prognostic significance of Inhibitors of Apoptosis Proteins in Gallbladder Cancer
Second Place
Akash Agrawal, MS
Heterodimerization of cholecystokinin 1 and cholecystokinin 2 receptors in gallbladder cancer: a new mechanism for carcinogenesis
Honorable Mention
Omar Rahal, Medical Student
The Emerging Field of Neural Implants: A Review for the Neurosurgeon
And
Elaf Mohamed Hashim Abdelraheem
Assessment of Preoperative Anxiety among Patients Undergoing Elective Major Surgery in White Nile State during Armed Conflict, Sudan, 2023
Mert Marcel Dagli, MD
Sharpe Postdoctoral Research Fellow, Department of Neurosurgery, Perelman School of Medicine University of Pennsylvania, Philadelphia, PA
Manual chart review (MCR) for extracting surgical data from Electronic Health Records (EHRs) is time-consuming, prone to error, and a significant bottleneck in clinical research and quality control. This study aimed to develop and validate a novel artificial intelligence (AI) framework that integrates Natural Language Processing (NLP) with a Large Language Model (LLM) to automate the extraction of relevant clinical data from spinal surgery EHRs and automate postoperative billing.
This study was supported by the TRIPOD+AI guidelines. We utilized three institutional databases comprising thoracolumbar adult spinal deformity cases (N=646), lumbar endoscopic spinal surgery cases (N=182), and lumbar decompression cases (N=5,998). The AI framework was replicated ten times to address hallucinations.
The primary outcome was the accurate identification of surgical details, including surgery type, levels operated, number of disks removed, levels fused, incidental durotomies, and postoperative billing. Secondary objectives explored time efficiency and costs. Performance metrics such as accuracy, sensitivity, AUC-ROC, F1-score, and positive predictive value were calculated with 95% confidence intervals using bootstrapping.
The NLP+LLM framework achieved a sensitivity of 0.999 and an AUC-ROC of 0.997 for clinical data extraction, demonstrating similar performance in billing automation, outperforming the human control. The use of a majority vote, utilizing data from the deduplicated (ten replications) run, eliminated all errors from singular runs.
Tokenization and cost analyses indicated substantial time savings (38.8 seconds overall) and cost savings ($9.04 overall) compared to manual chart reviews.
We demonstrated that the integration of NLP and LLM within an AI framework can significantly improve the accuracy, time, and cost efficiency of clinical data extraction and postoperative billing. These results suggest the potential for widespread adoption in healthcare. Further research will focus on enhancing the sensitivity and validating the model in broader clinical settings to further optimize billing automation and clinical documentation processes.
Nicole Marker, BA
Medical Student, University of Kentucky College of Medicine, Lexington, KY
Cardiac tumors affect fewer than 1 in 2,000 people. 75% of primary tumors are benign and the most common primary tumor in adults is myxoma. As a regional referral center, UK healthcare provides ‘standard of care’ treatment for these patients. This study is a retrospective review of diagnostic procedures, surgical management, and outcomes in patients treated for tumors of the heart in our institution. We compare our management approaches, clinical and surgical outcomes with those reported in the literature.
The study population includes patients 7-79 years old that presented to University of Kentucky Healthcare for tumors of the heart from July 2004 - January 2023. With IRB approval, subjects for this study were identified by searching the University of Kentucky database on the CPT codes for tumors of the heart (benign neoplasm of the heart or malignant tumor). CCTS Data Warehouse has provided data based on CPT codes. All data was stored on REDCap.
Operative treatment was offered to 52 people; 47 had resection and 5 had biopsy. The patient population consisted of 29 females and 23 males. The average patient age was 54 years old with the oldest being 79 and the youngest being 7. Most patients were Caucasian (49) followed by unknown (2) and black/African American (1). The most common presentation symptoms were shortness of breath, dyspnea, and fatigue. The most common diagnosis methods were transthoracic echocardiogram, transesophageal echocardiogram, cardiac MRI, and CT scan. Surgical treatments included 47 resections and 5 biopsies; a certain number of people required closure of septum with or without a patch. The most common postoperative complication was respiratory insufficiency (22) and sepsis (2). 48 patients were diagnosed with a benign neoplasm of the heart and 4 diagnosed with a malignant tumor. There were 49 primary tumors and 3 secondary tumors. The most common location of the tumor was the left atrium. The largest size number was 337.5 cm3 and the smallest was 0.042 cm3. The most common diagnosed cell type for the cardiac tumors was myxoma (32), followed by papillary fibroelastoma (11), fibroma (2), hemangioma (2), fatty infiltration, fibrosis and myocyte hypertrophy (1), metastatic hepatocellular carcinoma (1), metastatic neuroendocrine tumor (1), metastatic squamous cell carcinoma (1), and primary intimal sarcoma (1). The average length of stay in the hospital was 12.5 days with the longest being 59 days and shortest being 0.75 days. 39 patients were discharged home in a stable condition. The study population had one operative mortality (death within 30 days after surgery). 43 of the 52 patients treated are alive after 2 years.
Clinical outcomes such as discharge status, post-operative condition, and length of survival after procedures are non-inferior to those from other referral centers for such conditions. 83% of patients in this study surgically treated for cardiac tumors are alive after 2 years. This study shows surgical removal offers the best chance of cure for cardiac tumors.
Lauren Cox, BA
Medical Student, Medical University of South Carolina, Charleston, SC
Purpose: An increase in global surgical needs has been observed over the recent years by various world organizations including the Lancet Commission on Global Surgery (LCoGS). However, with this growing demand comes the reality of the current lack of exposure to global surgical training amongst medical trainees. This study aimed to assess global surgical involvement of all medical schools, including both allopathic and osteopathic programs, across the United States to determine the projected versus actual global engagement of these institutions through Global Surgery Student Alliance (GSSA) affiliation and several parameters- academics, research, partnerships, bidirectionality and service. We hope to provide a thorough overview of these programs, as well as transparency on the legitimacy and reach of efforts, and hopefully a foundation for future streamlining of programs in this field.
Methods: A total of 195 MD and DO schools were used in our review. Schools who had global surgery Involvement or a GSSA affiliation, but did not have a global health center, institute, or program were excluded from the study. A team of second year medical students conducted a mixed-methods study to evaluate and compare key fields outlined in GSSA’s mission including advocacy, education, research, international engagement and physician mentorship across programs. A secondary analysis was performed for included schools listed as GSSA chapters to evaluate the contributions of the respective GSSA organization. Data for each program was collected through online investigations into the chapter/institution website and social media pages and was input into an Excel matrix categorically according to the aforementioned parameters for statistical analyses.
Results: Of the 195 MD and DO medical school institutions across the United States, 41 had a comprehensive global surgery program. Of the 155 institutions without global surgery programs, 46 had some global surgery partnership, mission trip, international surgical elective or rotation, residency program, exchange or GSSA affiliation. 14 of the 41 institutions with global surgery programs had GSSA affiliations. 40 of the programs were MD institutions, and 20 were public institutions. Regionally, 9 of the 41 programs were located Central US, 10 Mid-Atlantic, 7 Northeast, 6 South, and 8 West.
Conclusion: In response to the worldwide call for a global surgery workforce, medical institutions across the United States have curated student alliances to minimize gaps in pivotal areas such as medical training, research and international partnerships. Through our review of institutions, we have found a discrepancy between the number of reported GSSA chapters versus the number of chapters producing tangible output aligning with GSSA's core values. Takeaways from this study can be used to address longitudinal decreases in GSSA membership activity while targeting areas for global surgery growth in the nation.