Abstract: It is of paramount importance to ensure accurate temperature monitoring in order to guarantee the safety and efficacy of thermal therapy treatments. The temperature increase in biological tissues is directly influenced by the thermal power deposition density (PDD). Given that the distribution of the change in backscattered energy (CBE) in ultrasound imaging reflects the location and geometry of the heated zone in tissue, this paper presents a method for estimating PDD based on the CBE distribution feature. The method employs the Gaussian distribution feature of CBE to reconstruct the temperature field, and the PDD is accurately estimated based on bio-heat transfer model. Initially, computer experiments simulating the heating process of muscle tissues were conducted to confirm the viability of the presented method, with the estimated results showing an average error of less than 5% compared to preset values. Then, the method was applied in a focused ultrasound heating experiment of ex vivo porcine muscle tissues. The temperature rise was predicted based on the estimated PDD, showing an error of about 0.3 °C compared to the actual temperature rise measured by thermocouples.