Classification of various sources of error in range assessment using proton radiography and neural networks in head and neck cancer patients

Abstract

This study evaluates the suitability of convolutional neural networks (CNNs) to automatically process proton radiography (PR)-based images. CNNs are used to classify PR images impaired by several sources of error affecting the proton range, more precisely setup and calibration curve errors. PR simulations were performed in 40 head and neck cancer patients, at three different anatomical locations (fields A, B and C, centered for head and neck, neck and base of skull coverage). Field sizes were 26 × 26cm 2 for field A and 4.5 × 4.5cm 2 for fields B and C. Range shift maps were obtained by comparing an unperturbed reference PR against a PR where one or more sources of error affected the proton range. CT calibration curve errors in soft, bone and fat tissues and setup errors in the anterior-posterior and inferior-superior directions were simulated individually and in combination. A CNN was trained for each type of PR field, leading to three CNNs trained with a mixture of range shift maps arising from one or more sources of range error. To test the full/partial/wrong agreement between predicted and actual sources of range error in the range shift maps, exact, partial and wrong match percentages were computed for an independent test dataset containing range shift maps arising from isolated or combined errors, retrospectively. The CNN corresponding to field A showed superior capability to detect isolated and combined errors, with exact matches of 92% and 71% respectively. Field B showed exact matches of 80% and 54%, and field C resulted in exact matches of 77% and 41%. The suitability of CNNs to classify PR-based images containing different sources of error affecting the proton range was demonstrated. This procedure enables the detection of setup and calibration curve errors when they appear individually or in combination, providing valuable information for the interpretation of PR images.

Deep learning–based 4D‐synthetic CTs from sparse‐view CBCTs for dose calculations in adaptive proton therapy

“…All PRs were simulated with a gantry angle of 0 degrees (anterior–posterior direction), an energy of 210 MeV, and a spacing between individual pencil beams of 1 mm in left–right and 2 mm in CC direction (similar to the 4D‐CT grid). Range errors between PRs were calculated according to previous studies 37–39 …”

“…Range errors between PRs were calculated according to previous studies. [37][38][39] Range error maps were computed for 4D-sCT-50%, 4D-sCT-average, and the 3D-sCT. The range error maps were analyzed by calculating mean and standard deviation for range probes within the patient outline and reported for each patient individually.…”

Technical note: Flat panel proton radiography with a patient specific imaging field for accurate WEPL assessment

“…Patient specific FP-PR acquisitions with the proposed spot-wise energy selection method provide a WEPL accuracy below 1% in the base of skull, neck, and brain regions, and hold the potential to serve as a quality control tool in online adaptive proton therapy workflows, detecting sources of range error such as patient misalignments, calibration curve errors, or anatomical changes. 10 Future investigations could evaluate the performance of the patient specific FP-PRs when a source of range error is introduced. Furthermore, the simultaneous detection of various sources of range error could be automated by means of artificial intelligence, 10 providing information to assist decisions about plan adaptation.…”

“…[2][3][4][5] Several studies have proven the suitability of PR to detect and/or mitigate multiple sources of range uncertainty such as patient misalignments, CT calibration curve errors, or anatomical variations. 4,[6][7][8][9][10][11] In the context of adaptive proton therapy, PR has a potential role as a quality control tool that provides in vivo range verification measurements. PR could assist decisions upon plan adaptation in combination with other daily x-ray-based imaging modalities.…”

“…PR with a single pencil beam energy and a multi-layer ionization chamber (MLIC-PR), also known as range probing, has proven to be suitable as a quality control tool to detect multiple sources of range errors. 6,7,10,20,21 Furthermore, MLIC-PR has enabled in vivo proton range verification in head and neck (HN) cancer patients. 12,13 Flat panel (FP) detectors suitable for proton radiation offer the most compact integrating PR solution, as well as large readout areas.…”