🤖 AI Summary
To address tumor displacement caused by respiratory motion and system latency during lung radiotherapy—leading to reduced beam targeting accuracy—this paper proposes a lightweight online respiratory prediction method leveraging 3D time-series infrared marker data. We systematically evaluate the applicability of online recurrent neural networks (RNNs), including UORO, SnAp-1, and DNI, for real-time radiotherapy prediction. Our contributions include: (i) a Jacobian-compressed implementation of SnAp-1 to reduce computational overhead; and (ii) an exact linear-coefficient update mechanism for DNI to enhance adaptation fidelity. Integrating time-series resampling (3.33/10/30 Hz), real-time credit assignment, and low-latency inference optimization, our approach requires only the first minute of data for initialization. Results show SnAp-1 achieves nRMSE of 0.335 (3.33 Hz) and 0.157 (10 Hz); UORO attains 0.0897 at 30 Hz; and DNI achieves single-step inference in just 6.8 ms—outperforming most offline-trained, resource-intensive methods.
📝 Abstract
In lung radiotherapy, infrared cameras can record the location of reflective objects on the chest to infer the position of the tumor moving due to breathing, but treatment system latencies hinder radiation beam precision. Real-time recurrent learning (RTRL), is a potential solution as it can learn patterns within non-stationary respiratory data but has high complexity. This study assesses the capabilities of resource-efficient online RNN algorithms, namely unbiased online recurrent optimization (UORO), sparse-1 step approximation (SnAp-1), and decoupled neural interfaces (DNI) to forecast respiratory motion during radiotherapy treatment accurately. We use time series containing the 3D position of external markers on the chest of healthy subjects. We propose efficient implementations for SnAp-1 and DNI based on compression of the influence and immediate Jacobian matrices and an accurate update of the linear coefficients used in credit assignment estimation, respectively. The original sampling frequency was 10Hz; we performed resampling at 3.33Hz and 30Hz. We use UORO, SnAp-1, and DNI to forecast each marker's 3D position with horizons (the time interval in advance for which the prediction is made) h<=2.1s and compare them with RTRL, least mean squares, and linear regression. RNNs trained online achieved similar or better accuracy than most previous works using larger training databases and deep learning, even though we used only the first minute of each sequence to predict motion within that exact sequence. SnAp-1 had the lowest normalized root mean square errors (nRMSE) averaged over the horizon values considered, equal to 0.335 and 0.157, at 3.33Hz and 10.0Hz, respectively. Similarly, UORO had the highest accuracy at 30Hz, with an nRMSE of 0.0897. DNI's inference time, equal to 6.8ms per time step at 30Hz (Intel Core i7-13700 CPU), was the lowest among the RNN methods examined.