Skip links

Thyroid QUS

A non-invasive ultrasound technique to characterize human thyroid


Date:
12/09/2014

the problem

Thyroid cancer is the fifth most common cancer in women.

My mother was diagnosed with thyroid cancer in 2013 after the doctor made the fine needle aspiration procedure to detect malignant tissue. After the results, my mother’s thyroid was extracted in a surgery and ran tests to validate its malignancy. Unfortunately, they found only benign nodules which are treatable with medicine. Having experiencing this negligence, I was determined to find a solution to this problem and avoid other women like my mother to suffer from this condition.

the solution

Backscatter coefficient is an intrinsic property that quantifies the amount of energy reflected by a material as a function of the frequency of the ultrasound wave.

After finding out the potential benefit of quantitative ultrasound to characterize tissue in a lab presentation at my university, I decided to work on this subject for my undergraduate thesis. My one-year research advised by Dr. Lavarello, finished with the successful implementation of a Matlab program to characterize healthy human thyroid in order to establish a standardization of its backscatter coefficient (BSC) curves. In a further study, the comparison with the BSC curves of thyroid nodules would provide significant differentiation to diagnose cancer.

the results

Part 1

Backscatter coefficient curves were estimated at several location within thyroid tissues for each volunteer. The average of BSC curves for each volunteer is presented. It can be observed that a good agreement was obtained for BSCs estimated from all different thyroids. As well, the mean BSC curve over each average BSC curves is plotted in thick dashed black line. The BSC values ranged between 0.0057 sr-1.cm-1 and 0.13 sr-1.cm-1 at 3MHz and 8 MHz, respectively.

the results

Part 2

BSC-based images of all thyroid samples are presented using the mBSC parameter value at the ROIs locations throughout the selected domain of thyroid tissue. It can be observed that the mBSC values exhibit variations within the selected domain but were relatively homogeneous and spanned similar ranges across all analyzed thyroids.

the conclusions

These results suggest that BSCs from healthy thyroids can be reliably and consistently estimated in vivo.

These encouraging results need to be further analyzed by expanding the limited size of the database used in this study in order to obtain a more consistent statistic representation. Furthermore, higher frequency transducers (i.e., up to 20 MHz) typically employed in high resolution echographic imaging will be studied in order to improve the spatial resolution of the estimates and explore if reliable BSC estimation is achievable at this frequency bands. Moreover, model-based tissue characterization using effective scatterer diameters (ESDs) and effective acoustic concentration (EACs) should be explored in order to determine the variability of other BSC-derived parameters for tissue characterization.