ºÚ°µ±¬ÁÏÍø

Optional
Meeting ID: 995 6974 0550
Passcode: 837273

Abstract: Prostate cancer is the most common cancer among men and the second most common cancer overall in the United States, with 288,300 new cases in 2023 and 1.4 million globally. Approximately 40.5% of patients undergo radical prostatectomy (RP), where the primary goal is complete tumor removal while preserving neurovascular structures and urethral length to maintain erectile function and continence. Achieving this balance contributes to variable positive surgical margin (PSM) rates, reported between 4–48% (average 21%). PSMs increase recurrence risk and often necessitate costly, noxious adjuvant therapies. Despite this there is no widely adopted method for intraoperative margin assessment in real time. Electrical impedance has shown promise in distinguishing malignant from benign prostate tissue in ex vivo studies.

This dissertation presents the development, optimization, and evaluation of a broadband Electrical Impedance Tomography (EIT) system for intraoperative surgical margin assessment during robotic-assisted RP. A custom electrode array was optimized for tetrapolar measurements across 100 Hz–1 MHz. Integration with signal conditioning circuitry at the probe tip expanded bandwidth from 10 kHz–100 kHz in previous designs to 100 Hz–1 MHz with SNRs >85 dB. In an ex vivo bovine model, complex 3D EIT reconstructions identified an optimal electrode pressure (2–5 kPa) and introduced a best fit factor (β) to improve permittivity reconstructions. Difference EIT at 100 Hz achieved sensitivity of 0.92, specificity of 0.90, and reconstruction time of 1.67 s, supporting real-time feasibility. Clinical studies collected 784 ex vivo and 154 in vivo impedance datasets across 64 patients, with fresh-section prostate data showing strong separation for malignancies >50% of the sampled area. This work demonstrates a compact, laparoscopic-compatible, high-bandwidth EIT probe interfacing with the Da Vinci robotic system, enabling accurate, stable current delivery and rapid reconstructions, forming the foundation for real-time intraoperative margin assessment in RP.

Thesis Committee: Ryan Halter (Chair), Ethan Murphy, Kofi Odame, Lawrence Dagrosa, and Gary Saulnier (University at Albany)