A T1‐based correction method for proton resonance frequency shift thermometry in breast tissue

Abstract

Purpose Develop and evaluate the effectiveness of a T1‐based correction method for errors in proton resonant frequency shift thermometry due to non‐local field effects caused by heating in fatty breast tissues. Methods Computational models of human breast tissue were created by segmenting MRI data from a healthy human volunteer. MR‐guided focused ultrasound (MRgFUS) heating and MR thermometry measurements were simulated in several locations in the heterogeneous segmented breast models. A T1‐based correction method for PRF thermometry errors was applied and the maximum positive and negative errors and the root mean squared error (RMSE) in a region around each heating location was evaluated with and without correction. The method uses T1 measurements to estimate the temperature change in fatty tissues and correct for their influence. Experimental data from a heating study in cadaver breast tissue were analyzed, and the expected PRFS error computed. Results The simulated MR thermometry had maximum single voxel errors ranging between 10% and 18% when no correction was applied. Applying the correction led to a considerable improvement, lowering the maximum error range to 2%–5%. The 5th to 95th percentile interval of the temperature error distribution was also lowered with correction, from approximately 3.5 to 1°C. This correction worked even when T1 times were uniformly raised or lowered by 5%–10%. The experimental data showed predicted errors of 15%. Conclusions This simulation study demonstrates that the T1‐based correction method reduces MR thermometry errors due to non‐local effects from heating in fatty tissues, potentially improving the accuracy of thermometry measurements during MRgFUS treatments. The presented correction method is reliant on having a patient‐specific 3D model of the breast, and may be limited by the accuracy of the fat temperatures which in turn may be limited by noise or bias present in the T1 measurements.

Characterization of susceptibility artifacts in magnetic resonance thermometry images during laser interstitial thermal therapy: dimension analysis and temperature error estimation

“…Seeing the temperature error entity and the consequent miscalculation of thermal damage produced, methods to correct susceptibility artifacts should be included in the clinical routine. Several studies proposed simulation-based or T 1 -based measurement correction for PRFS thermometry error (Viallon et al 2010, McLean et al 2021, (Baron et al 2014). Despite the potential of such strategies, they are far from being a standard practice.…”

“…On the other hand, artifacts in the images induced by local susceptibility changes may occur during thermal therapies thus interfering with the thermal damage map estimated by the MRTI. Susceptibilityrelated artifacts were observed during high-intensity focused ultrasound (HIFU) thermal therapy (McLean et al 2021), (Sprinkhuizen et al 2010). In these studies, susceptibility changes induced by temperature variation in adipose tissues cause dipole-structured artifacts leading to temperature errors of 15%.…”

Magnetic resonance imaging-guided radiofrequency ablation of breast cancer: a current state of the art review

“…Cheng et al 32 proposed a dual-step iterative temperature estimation of a fat-referenced PRFS method to improve both the accuracy and precision of temperature estimations in fat-containing tissues. T1- and T2-based 33 methods have been implemented recently to correct thermometry errors brought on by the non-local effects of heating in fatty tissues 34 or as an adjunct to PRFS thermometry in aqueous tissues (glandular, tumor, etc.). 35 …”

“…In addition, multi-echo MR thermometry using iterative water–fat separation techniques 36 has been validated in simulations, phantom heating experiments, and in vivo breast and liver experiments. This method could help monitor ablative therapies in fatty tissues 34 such as the breast or liver. In contrast, US guidance cannot quantify the heat deposited in ablation target tissue.…”

Methods for MRI-guided treatment of painful bone metastases with Focused Ultrasound

“…In addition, the T 1 information of fat is beneficial also to correct PRFS estimates in water for errors caused by heating-induced field distortions [190,191].…”