More 2026 Capstone Innovations
Carle Illinois College of Medicine's 2026 Capstone and Data Science projects bridge patient-centered medicine, engineering, and new technologies that leverage machine learning and data science to generate solutions to a wide range of health problems.
New Devices
Tractura: Dual Surgical Retraction Device
Tractura is a new medical device designed to improve surgeons’ efficiency, positioning, and comfort during long procedures. Retractors are instruments used to hold incisions open during surgery, enabling the surgical team to see clearly and operate within the surgical site. For some larger surgeries, retractors are needed at two points on each side of longer incisions to reduce tissue strain and ensure that surgeons can see the entire surgical site.
Tractura is designed to simultaneously hold two off-the-shelf retractors securely in place while also allowing for adjustments on the fly. The ability to hold two retractors simultaneously with a single hand frees the physician’s other hand for secondary tasks, resulting in more efficient use of surgical personnel. Tractura’s improved ergonomic design promotes better wrist and forearm alignment, reduces repetitive repositioning, and more evenly distributes force in maintaining access to the surgical site.
The Tractura team includes: CI MED students Neddie Byron, Alan Fullenkamp, Joline Marcon Hoffner; Mechanical Science & Engineering undergraduate students: Tanner Amy, Justin Gordon, Eben Lee, Lorenzo Marasigan; and Health Innovation Professor Michael Oelze, the team’s advisor.
Shelbow: Shoulder and Elbow Bracing System to Support Stroke Recovery
Carle Illinois College of Medicine students have designed a new shoulder and elbow brace to help patients paralyzed on one side by stroke as they recover normal movement.
One-sided paralysis or weakness is common in stroke patients. Traditional slings are used now to immobilize the patient’s arm to prevent injury, but traditional slings can also restrict circulation and limit natural arm movement during walking and post-stroke therapy.
The Shelbow device combines a shoulder brace with a spring and damper mechanism at the elbow to allow a controlled arm swing rather than full immobilization. An extension spring helps support the weight of the affected arm and assists with lifting the arm during motion. Overall, the system keeps the arm in a more natural position and allows a controlled arm swing during walking.
In testing, Shelbow restored a more rhythmic and forward arm movement pattern, suggesting improved biomechanical stability during rehabilitation.
Shelbow team members include CI MED students Kevin Henry Zhang, Jose Beltran, and Duane Macatangay; Engineering partners Maria Attaallah, Leonardo Delgado, Omkar Gandhi, and Jorge Jimenezl Narayan; and MBA partners Sonali Gopinath Chetty and Peiyu Wang.
Data Science
Loaded Language in Orthopedic Trauma: Can It Predict Clinical Complexity?
A study by a Carle Illinois College of Medicine team is the first to reveal that language tucked away in radiological reports can be a tip-off to diagnostic challenges, case complexity, and even length of hospitalization in orthopedic trauma cases.
When a patient experiences orthopedic trauma, medical imaging – including X-rays, CT scans, and ultrasound images – plays a key and early role in both diagnosis and treatment. Those images are reviewed and interpreted by radiologists, who describe and report what they see. Their reports guide surgical teams in both diagnosing and treating patients.
The CI MED team analyzed the language in pre-op imaging reports of thousands of orthopedic trauma patients to determine if there is a concrete correlation between the language used in radiology reports and downstream clinical complexity in these patients.
“Our study demonstrated the first empirical evidence that latent linguistic characteristics in radiology reports (i.e. implicit sentiment and uncertainty) correlate with downstream clinical complexity in orthopedic surgery patients,” team member Samuel Blake said.
Blake says the study findings suggest that radiologists (and artificial intelligence models) drafting radiology reports should be trained to avoid the use of potentially biasing language that could create care delays, increase patient and hospital costs, and adversely impact outcomes.
The data science study, “Linguistic Features of Preoperative Imaging Reports as a Predictor of Clinical Complexity in Orthopedic Trauma,” was conducted by CI MED students Samuel Blake and Anthony Bosshardt, in collaboration with Paul Camacho, PhD, of the Beckman Institute at the University of Illinois Urbana-Champaign.