First Patient Study with Visual Biofeedback for Gating Delivery Efficiency Improvements on a 1.5 T MR-Linac.

Intra-fraction motion of lesions in the thorax and abdomen due to respiration is an important source of uncertainty in SBRT. MR-Linacs provide the ability to continuously track the target and gate the radiation beam accordingly. Gating allows for smaller margins and can facilitate hypo-fractionation, but also lengthens the treatment due to the typically reduced delivery efficiency. Offering biofeedback (BF) and providing breathing instructions to the patient potentially helps counteracting the delivery efficiency penalty introduced by gating. In this work we present the first results of a patient study on a 1.5 T MR-Linac, using a prototype visual BF system, compatible with the vendor-provided motion management software. Experiments were conducted on a 1.5 T MR-Linac, equipped with clinical motion monitoring and gating functionality. 12 healthy volunteers (age: 24-79y, female/male: 6/6) were included in the study in phase 1, followed by 6 cancer patients (age: 30-74y, female/male: 2/4), in phase 2. Target motion was estimated in real-time based on 6 Hz sagittal/coronal interleaved 2D cine-MRI data. The resulting target position was streamed to in-house developed BF software and displayed to the study participants on an MRI-compatible monitor positioned at the rear of the MR-Linac bore, visible via an in-bore mirror. End-exhale and end-inhale breath hold (BH) workflows with 6 mm gating windows (GW) were simulated. Each study participant received BF for two consecutive 5-minute sessions. In phase 1, participants were used to select the preferred graphical user interface (GUI) from two GUI concepts (no data shown). In phase 2, patients received instructions and BF using a graph-style GUI to maintain a subject-determined BH (SDBH) as long as comfortable. The duration of each BH was not specified. SDBHs were conducted by three patients in end-exhale and by three patients in end-inhale. For the SDBH workflow, the average estimated delivery efficiency, defined as the ratio between the time spend within the GW and the total session time (5 min), was 49% (62%, 41%, 40%) for end-exhale SDBHs and 69% for end-inhale SDBHs (50%, 74%, 84%) and. Three patients showed a strong learning effect (+8, +6, +18 percentage points) between the subsequent sessions. Questionnaires revealed that all patients fully agreed that it was clear when to perform a BH and when to breathe normally. Two patients, performing an end-exhale SDBH, indicated that they might have found an end-inhale SDBH easier and more intuitive. Visual BF is expected to dramatically improve the delivery efficiency during gating treatments on the MR-Linac. Patient selection for BF systems is important, especially with regards to visual or cognitive impairments. Performance improvements over time in individual patients point towards the value of dedicated training sessions. [ABSTRACT FROM AUTHOR]