Osteoarthritis (OA) is one of the most common joint diseases. Thermotherapy, such as that using a microwave diathermy applicator, is widely used for OA. The deep tissue of a knee joint should be heated to 36 ℃-38 ℃ for an effective thermotherapy. However, heating this deep region using a microwave diathermy applicator is challenging. Previously, we proposed a resonant cavity applicator to overcome these problems and confirmed its ability (heating experiments on bovine knees) to heat the deep region of a knee joint without physical contact. Furthermore, we proposed a method of temperature measurement using ultrasound images. In this method, the temperature distribution was measured using noninvasive image analysis. In a previous study, we found temperature measurement accuracy of ≤ 1.0 ℃. In the present paper, we describe a temperature distribution using a 3D-printed knee model for treating OA. First, we created a 3D finite element model (FEM) of the knee and a 3D-printed knee model from 2D medical images. Second, we calculated temperature distributions in the FEM model and performed a heating experiment with a prototype of the heating system. Third, we performed positioning accuracy experiments to investigate the accuracy of our temperature measurement system comprising a robotic arm, 3D-printed knee model, and ultrasound diagnostics. Finally, we measured the temperature distribution inside the 3D-printed knee model from ultrasound images. The heating experiments confirmed that our proposed method could heat deep regions of a knee joint without any undesirable hotspot. Therefore, our results suggest that this method is useful for effective thermotherapy of OA.

　　We designed and prototyped a small rectangular resonant cavity applicator system for thermal treatment of knee osteoarthritis, and discussed its usefulness from both results of computer simulations and heating experiments. The authors had proposed a deep knee joint thermal rehabilitation system using a cylindrical resonant cavity, and demonstrated its usefulness from the results of agar phantom heating experiments and clinical heating experiments by volunteers. However, the area of the leg placed inside the cavity was wide when using the cylindrical cavity, so the healthy calf and thigh, which are not the targeted tissues, were also slightly heated.

　　Therefore, in the present study, we clarified the possibility of deep thermal treatment with higher safety by changing the cylindrical shape of the resonant cavity to a compact rectangular shape. First, the dimensions and resonant frequency band of the rectangular cavity applicator were determined using the finite element method (FEM)． Based on the numerical results, a heating system was prototyped, and heating experiments were conducted on a cylindrical agar phantom and a human leg-shaped agar phantom. Dimensions of the rectangular resonant cavity applicator are 300 mm in height, 350 mm in width and 200 mm in length. Heating experiments were conducted with a heating power of 30 W and a heating time of 10 minutes. As a result, it was confirmed that the central part of the cylindrical agar phantom was locally heated, and the temperature increase value was approximately 8.0 ℃. Furthermore, in the experiment which the leg-shaped agar phantom was heated, it was confirmed that only the knee joint was locally heated and no hot spots were generated in other areas. From these experimental results, it was concluded that the deep part of the knee joint could be safely and locally heated by using the rectangular resonant cavity applicator.