Variability in Electrocochleography Imaging Protocol Training Among Medical Trainees

Tuesday, October 20, 2026

8:51 AM - 8:57 AM PT

Location: West Exhibit Hall A, Learning Zone 1

Presenting Author(s)

Anika S. Walia, BA

Research Fellow/ Medical Student
Thomas Jefferson University
Boston, Massachusetts, United States

Co-Author(s)

LH

Lia Herzig, BS

Medical Student
Thomas Jefferson University Hospital, United States
KL

Kalena Liu, BS

Medical Student
Thomas Jefferson University Hospital, United States
AP

Anil Patnala, MS

Director Implant Technologies Research at Advanced Bionics
Advanced Bionics, United States
TW

Thomas O. Willcox, Jr., MD

Associate Professor
Thomas Jefferson University Hospital
Philadelphia, PA, United States
RC

Rebecca C. Chiffer, MD

Associate Professor
Thomas Jefferson University Hospital
Philadelphia, PA, United States

Senior Author(s)

Jacob B. Hunter, MD

Associate Professor
Thomas Jefferson University Hospital
Philadelphia, PA, United States

Disclosure(s):

Jacob B. Hunter, MD:

faculty for this accredited education activity has no relationship(s) with ineligible companies to disclose.

Introduction: Accurate preoperative cochlear measurements are essential for cochlear implant electrode selection and surgical planning, yet the reliability of trainee-derived measurements using emerging planning software remains incompletely characterized.

Methods: In this prospective study, 20 medical trainees measured cochlear length, cochlear width, cochlear duct length (CDL), and C1 insertion angle (C1IA) using CImago (Advanced Bionics) on five preoperative CT scans (slice thickness 0.6–1.0 mm) from five patients. Trainees completed a standardized 1-hour training session and were supervised during all analyses. Measurements were compared with software developer–defined gold standards, and inter-rater reliability was assessed using Spearman correlation and intraclass correlation coefficients (ICC).

Results: Across scans, mean cochlear length ranged from 8.54–9.12 mm (CV 1.77–4.45%), cochlear width 6.49–7.19 mm (CV 1.75–3.36%), CDL 26.0–29.0 mm (CV 1.17–3.08%), and C1IA 412.1 –489.1 ° (CV 1.92–6.53%). Spearman correlation against the gold standard revealed cochlear width (ρ=0.626, p< 0.005) and C1IA (ρ=0.374, p< 0.05) to have the highest reliability, followed by cochlear length (ρ=0.242, p< 0.05). Single-rater ICCs were highest for CDL (0.87) and C1IA (0.72), moderate for cochlear width (0.66), and lowest for cochlear length (0.40; all p< 0.005). Averaging across raters markedly improved reliability for all metrics (ICC 0.92–0.99). Notably, CT scan 2, which a utilized larger slice thickness (1 mm), demonstrated the greatest variability across measures including cochlear length, C1IA, and CDL than all other scans with finer resolution (0.6–0.7 mm).

Conclusions: CImago enables reliable trainee-derived measurements of CDL, cochlear width, and C1 insertion angle, particularly when results are averaged across raters, while cochlear length shows greater variability. Measurement reliability appears influenced by CT slice thickness, showing the importance of image quality and multi-rater averaging in preoperative cochlear planning.