Integration of thermal imaging and checklist to differentiate cellulitis from pseudocellulitis

Summary: Abstract Body: Cellulitis is a skin and soft tissue infection that poses diagnostic challenges leading to substantial healthcare costs and unnecessary hospitalizations. It is the most misdiagnosed disorder in dermatology, highlighting the need for improved diagnostics. This study investigates using thermal imaging (TI) and a clinical checklist to improve accuracy in distinguishing cellulitis from mimickers (pseudocellulitis). TI was used to analyze temperature differentials between affected and unaffected sites in patients with cellulitis symptoms and mimickers. We collected 1165 thermal images from 105 patients with dermatologist-confirmed cellulitis and pseudocellulitis. Patients were enrolled from outpatient clinics, the emergency department, and inpatient unit. Pertinent characteristics were analyzed and a diagnostic checklist was developed from the clinical assessments and TI data. Cellulitis cases showed higher prevalence of tenderness over pruritus 92.2%, confluent erythema over patchy 81.0%, and warmth 90.5% compared to psuedocellultis (tenderness 73.2%, confluence over patchy 43.9%, warmth 50.0%). The distribution of patchy or confluent erythema (p=0.001), tenderness or pruritic (p=0.015), and warmth (p<0.0001) also differed between cellulits and pseudocellulitis. TI showed a temperature difference of 3.1°C between cellulitis affected and unaffected areas. Cellulitis patients also showed higher mean heart rate (81.2±16.5 vs 79.5±13.3 BPM; p=0.62) and white blood cell counts (9.5±5.5 vs 7.1±3.4 K/uL; p=0.03). This highlights distinct differences in presentation between cellulitis and psuedocellulitis. Integrating TI and a clinical checklist as a point-of-care diagnostic tool shows promise in improving diagnostic precision and lowering cellulitis misdiagnoses, mitigating healthcare costs, reducing unnecessary hospitalizations, and subsequently optimizing patient outcomes. Nora Bensellam^{1, 2}, Ursula Biba^{1, 2}, Allison Yan^{1, 2}, Mingye Gao³, Ahana Gaurav^{1, 2}, Eric Xia^{1, 2}, James Choe^{1, 2}, Dev Sahni^{1, 2}, David Li^{1, 2}, Brian Anthony³, Arash Mostaghimi^{1, 2} 1. Brigham and Women's Hospital, Boston, MA, United States. 2. Harvard Medical School, Boston, MA, United States. 3. Massachusetts Institute of Technology, Cambridge, MA, United States. Clinical Research: Epidemiology and Observational Research