Intelligent Thermography for Intraoperative Brain Imaging

 

During my Master’s studies, I designed, optimized, and fabricated a fully novel and intelligence-based thermography system with an application in intraoperative brain imaging and minimally invasive diagnosis in neurosurgery. The device was capable of localizing brain tumors with the measurement of temperature on the brain surface with the fusion of machine learning techniques and advanced heat transfer models. To verify the accuracy and efficacy of the device during intraoperative brain imaging, tissue phantoms with similar properties to brain tissue were fabricated in a collaborative study with a team of medical specialists. The outcome of my research was applauded at both university and national levels, and as a result, my master’s thesis was selected as the best and the most innovative engineering research outcome of the year among all Iranian universities in 2016. From this contribution, we could file a patent (granted) and published the research results in well-known journals in the field of thermal and biomedical engineering Journal of Physics D: Applied Physics, and Applied Thermal Engineering.

Thermography is a new non-invasive developing technique based on the temperature measurement in the soft tissue which has been successfully used in the diagnosis and detection of breast cancers, skin cancers, and eye diseases. Initial studies using intraoperative imaging of brain tumors revealed a significant temperature difference between the tumor tissue and the surrounding neural tissue while depending on the type of the tumor, the temperature difference is varied. Intraoperative Thermal Imaging (ITI) is a novel neuroimaging technique that can potentially localize brain tumors for optimized surgical resection. Estimation of tumor characteristics including tumor temperature, location, depth, and size plays a significant role in safe tumor surgery. This study introduces a new approach to ITI based on artificial tactile sensing (ATS) technology in conjunction with artificial neural networks (ANN) and the feasibility and applicability of this method in the diagnosis and localization of brain tumors is investigated. In order to analyze the validity and reliability of the proposed method, two simulations were performed. (i) An in vitro experimental setup was designed and fabricated using a resistance heater embedded in agar tissue phantom in order to simulate heat generation by a tumor in the brain tissue; and (ii) A case report patient with parafalcine meningioma was presented to simulate ITI in the neurosurgical procedure. In the case report, both brain and tumor geometries were constructed from MRI data and tumor temperature and depth of location were estimated. For experimental tests, a novel assisted surgery robot was developed to palpate the tissue phantom surface to measure temperature variations and ANN was trained to estimate the simulated tumor's power and depth. Results affirm that ITI based ATS is a non-invasive method that can be useful to detect, localize and characterize brain tumors..

 

Analysis
Forward-inverse analysis (Sadeghi-Goughari and Mojra 2016, journal of Physics D: Applied Physics D

 

Geometry
Geometry derivation from real patient model (Sadeghi-Goughari and Mojra 2016, Journal of Physics D: Applied Physics)

 

temperature simulation
thermal simulation of brain tumor ((Sadeghi-Goughari and Mojra 2016, Journal of Physics D: Applied Physics)

 

AI
Tumor features estimation using AI analysis

 

In vitro experimental setup

thermal