Tumornet 1 Tumornet 1
Tumornet 2 Tumornet 2
The aim of this application is help the neurosurgeon define the margin of a tumor during the surgery via a holographic representation aided by an AI algorithm.


Intraoperative modalities aimed at defining tumor margins suffer from limited accuracy and inefficient workflow. Although Stimulated Raman Histology (SRH) – the only available modality today – can determine pathological diagnosis from a specimen in less than one minute, anatomic information on the location of the specimen is not retained, limiting the ability to define the tumor margin. TUMOR-NET is an intraoperative modality that is accurate and streamlines workflow to facilitate defining tumor margins.

Project Aims

After extensive research and planning, our team has developed the following aims/goals for TUMOR-NET:

  1. Development and optimization of the TUMOR-NET platform to create a 3D holographic tumor margin model based on preoperative imaging and allow biopsies of this margin to be assessed intraoperatively, updating the AR model.

  2. Integration of TUMOR-NET software interface and the NIO pathology system using artificial intelligence to identify tumor margins instantaneously for clinical use.

  3. Development of the TUMOR-NET geotagging digital-pathology-repository to allow for outcomes analysis with a machine learned probability of tumor progression algorithm.

Staff and Student Berk Basarer and Adam McMahon
Faculty Dr. Michael Ivan
College UHealth