C Hsieh^1,2*, K Lu^2,3, G Lin^2,3, C Wu⁴, F Chen^4,5, (1) Institute for Radiological Research, Chang Gung University, Taoyuan, TW, ROC, (2) Imaging Core Lab, Chang Gung Memorial Hospital, Linkou, Taoyuan, TW, ROC (3) Department of Medical Imaging and Intervention, Chang Gung Memorial Hospital, Linkou, Taoyuan, TW, ROC (4) Radiation Biology Core Lab., Chang Gung Memorial Hospital, Linkou, Taoyuan, TW, ROC (5) Department of Medical Imaging and Radiological Sciences, Chang Gung University, Taoyuan, TW, ROC

• 
  C Hsieh

Warning: Undefined variable $sessionDesc in D:\inetpub\aapmnfe\meetings\2020AM\programInfo\programAbs.php on line 495
(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: The activity of anaerobic glycolysis reaction may be detected in-vitro or in-vivo by ¹³C MRS of hyperpolarized (HP) substrates. The dynamic signals from HP substrates and products have been fitted to a number of mathematical models to determine chemical reaction rates. However, in our study, the traditional model with chemical reaction rate constants has shown a large difference between the model and our results. Here, we have introduced a correction term, the chemical reaction rate in anaerobic glycolysis as a function of time, into the traditional model to minimize the discrepancy.


Methods: Our kinetic model for the detection of ¹³C-enriched lactate after the administration of labeled pyruvate is to consider bidirectional three pools: intracellular pyruvate, intracellular lactate, and extracellular pyruvate. In addition, with consideration of the effects of T1, radiofrequency (RF) excitation, and the dynamic chemical reaction rate, this three-pool model becomes rather complex. For our in vitro experiment, THP1 6x107 cells were in 9 ml medium. The HP ¹³C pyruvate, 1 ml, was poured into this medium. HP ¹³C signal in this medium was acquired by generic ¹³C MRS in 3T. The mixing and delivery of substrate and cells were at different time points and recorded. These time records were reference points in our model.


Results: For the pyruvate comparison between two kinetic models and experimental data, the maximum difference was about 35%! However, for the dynamic lactate comparison, the difference between the real data and the traditional model (51%) was larger than that between the real data and our model (37%) when SNR=4. Although the correction term did not cause any change in substrate, it has provided ¹³C lactate better fitting in our data.


Conclusion: The correction term in our three-pool model was necessary to explain and fit the experiential data. The further validation is still needed.

Keywords

MR, Cell Kinetics

Taxonomy

TH- Dataset Analysis/Biomathematics: Informatics

Contact Email