Photon-counting computed tomography thermometry via material decomposition and machine learning.
Nathan WangMengzhou LiPetteri HaverinenPublished in: Visual computing for industry, biomedicine, and art (2023)
Thermal ablation procedures, such as high intensity focused ultrasound and radiofrequency ablation, are often used to eliminate tumors by minimally invasively heating a focal region. For this task, real-time 3D temperature visualization is key to target the diseased tissues while minimizing damage to the surroundings. Current computed tomography (CT) thermometry is based on energy-integrated CT, tissue-specific experimental data, and linear relationships between attenuation and temperature. In this paper, we develop a novel approach using photon-counting CT for material decomposition and a neural network to predict temperature based on thermal characteristics of base materials and spectral tomographic measurements of a volume of interest. In our feasibility study, distilled water, 50 mmol/L CaCl 2 , and 600 mmol/L CaCl 2 are chosen as the base materials. Their attenuations are measured in four discrete energy bins at various temperatures. The neural network trained on the experimental data achieves a mean absolute error of 3.97 °C and 1.80 °C on 300 mmol/L CaCl 2 and a milk-based protein shake respectively. These experimental results indicate that our approach is promising for handling non-linear thermal properties for materials that are similar or dissimilar to our base materials.
Keyphrases
- neural network
- computed tomography
- dual energy
- radiofrequency ablation
- image quality
- high intensity
- positron emission tomography
- contrast enhanced
- machine learning
- magnetic resonance imaging
- resistance training
- big data
- electronic health record
- gene expression
- living cells
- oxidative stress
- optical coherence tomography
- artificial intelligence
- monte carlo
- atrial fibrillation
- small molecule
- protein protein