Radiomics Analysis of Deltoid Muscle on Preoperative Shoulder Imaging for Reverse Total Shoulder Arthroplasty

Authors

• Henry Lee BS
• Christina Hills BS
• Connor SakisFollow
• Humza Bhatti BS, Michigan State University College of Human Medicine
• Arya Sharma, Michigan State University
• Perra Razoky MD, Henry Ford Health
• Prateek Chintalapati MD, Henry Ford Health
• Lisa Doan BS
• Adam Alessi PhD, Michigan State University
• Jared Mahylis MD, Henry Ford Health
• Stephanie Muh MD, Henry Ford HealthFollow
• Chad Klochko MD, Henry Ford Health

Research Mentor Name

Dr. Stephanie Muh

Research Mentor Email Address

smuh1@hfhs.org

Institution / Department

Henry Ford Health Department of Orthopedic Surgery, Shoulder and Elbow Division; Henry Ford Health Department of Radiology; Michigan State University Department of Radiology

Document Type

Research Abstract

Research Type

clinicalresearch

Graduate Level Research

no

Abstract

Title: Radiomics Analysis of Deltoid Muscle on Preoperative Shoulder Imaging for Reverse Total Shoulder Arthroplasty

Authors: Henry J. Lee BS, Christina Hill BS, Connor Sakis BS, Humza Bhatti BS, Arya Sharma, Perra Razoky MD, Prateek Chintalapati MD, Lisa Doan BS, Adam Alessio PhD, Jared Mahylis MD, Stephanie Muh MD, Chad Klochko MD

Background: Studies have shown that increased preoperative deltoid volume is associated with improved postoperative outcomes in reverse total shoulder arthroplasty (rTSA). Radiomics is a field which uses mathematics and machine learning for image analysis. It shows promise as a tool for characterizing the deltoid preoperatively. This study aims to determine whether ultrasound (US) can be used to accurately estimate deltoid volume.

Methods: Using a search of the electronic medical record (EMR), this IRB-approved retrospective study identified patients aged 18-89 who underwent US and computed tomography (CT) imaging of the shoulder within 2 years prior to rTSA between 1/1/2014 and 12/31/2022. While preoperative CT images were segmented to allow for volume calculation, US was used to measure deltoid thickness as a proxy for volume.

Results: Search of the EMR yielded 79 subjects. 26 CT images and 20 Us images from these subjects were processed. From CT, mean deltoid volume was 293+/-106 cm³. From US, mean deltoid thickness was 0.91+/-0.23 cm. CT-derived volume and deltoid thickness on ultrasound were correlated with a coefficient of 0.52.

Conclusion: This preliminary analysis suggests that US is imperfect, but viable for estimating deltoid volume. Compared to CT, ultrasound has a low cost and is frequently obtained in patients with a history of shoulder pathology. This positions ultrasound as a useful tool for preoperative, opportunistic screening of deltoid volume to determine patient candidacy for rTSA.

Disciplines

Medicine and Health Sciences

Recommended Citation

Lee, Henry BS; Hills, Christina BS; Sakis, Connor; Bhatti, Humza BS; Sharma, Arya; Razoky, Perra MD; Chintalapati, Prateek MD; Doan, Lisa BS; Alessi, Adam PhD; Mahylis, Jared MD; Muh, Stephanie MD; and Klochko, Chad MD, "Radiomics Analysis of Deltoid Muscle on Preoperative Shoulder Imaging for Reverse Total Shoulder Arthroplasty" (2026). Medical Student Research Symposium. 520.
https://digitalcommons.wayne.edu/som_srs/520