A clinical evaluation of the performance of five commercial artificial intelligence contouring systems for radiotherapy.

Purpose/objectives: Auto-segmentation with artificial intelligence (AI) offers an opportunity to reduce inter- and intra-observer variability in contouring, to improve the quality of contours, as well as to reduce the time taken to conduct this manual task. In this work we benchmark the AI auto-segmentation contours produced by five commercial vendors against a common dataset. Methods and materials: The organ at risk (OAR) contours generated by five commercial AI auto-segmentation solutions (Mirada (Mir), MVision (MV), Radformation (Rad), RayStation (Ray) and TheraPanacea (Ther)) were compared to manually-drawn expert contours from 20 breast, 20 head and neck, 20 lung and 20 prostate patients. Comparisons were made using geometric similarity metrics including volumetric and surface Dice similarity coefficient (vDSC and sDSC), Hausdorff distance (HD) and Added Path Length (APL). To assess the time saved, the time taken to manually draw the expert contours, as well as the time to correct the AI contours, were recorded. Results: There are differences in the number of CT contours offered by each AI auto-segmentation solution at the time of the study (Mir 99; MV 143; Rad 83; Ray 67; Ther 86), with all offering contours of some lymph node levels as well as OARs. Averaged across all structures, the median vDSCs were good for all systems and compared favorably with existing literature: Mir 0.82; MV 0.88; Rad 0.86; Ray 0.87; Ther 0.88. All systems offer substantial time savings, ranging between: breast 14-20 mins; head and neck 74-93 mins; lung 20-26 mins; prostate 35-42 mins. The time saved, averaged across all structures, was similar for all systems: Mir 39.8 mins; MV 43.6 mins; Rad 36.6 min; Ray 43.2 mins; Ther 45.2 mins. Conclusions: All five commercial AI auto-segmentation solutions evaluated in this work offer high quality contours in significantly reduced time compared to manual contouring, and could be used to render the radiotherapy workflow more efficient and standardized.

Computed tomography guided interstitial percutaneous high-dose-rate brachytherapy in the management of lung malignancies. A review of the literature.

A growing number of patients with lung cancer are not amenable to surgery due to their age or comorbidities. For this reason, local ablative techniques have gained increasing interest recently in the management of inoperable lung tumors. High-dose-rate percutaneous interstitial brachytherapy, performed under CT-guidance, is a newer form of brachytherapy and is a highly conformal radiotherapy technique. The aim of this study was to describe this method and review the existing literature. Eight articles comprising 234 patients reported toxicity and clinical outcome. The follow-up ranged from 6 to 28 months. Diverse fractionation schemes were reported, with 20 Gy in a single fraction being the most frequently utilized. Toxicity was limited; major pneumothoraces occurred after only 8% of the interventions. Local control rates at one year ranged between 37% and 91%. In conclusion, high-dose-rate percutaneous interstitial brachytherapy is a safe, fast, and efficient treatment option for inoperable lung tumors.

Impact of varying planning parameters on proton pencil beam scanning dose distributions in four commercial treatment planning systems.

PURPOSE: In pencil beam scanning proton therapy, target coverage is achieved by scanning the pencil beam laterally in the x- and y-directions and delivering spots of dose to positions at a given radiological depth (layer). Dose is delivered to the spots on different layers by pencil beams of different energy until the entire volume has been irradiated. The aim of this study is to investigate the implementation of proton planning parameters (spot spacing, layer spacing and margins) in four commercial proton treatment planning systems (TPSs): Eclipse, Pinnacle3 , RayStation and XiO. MATERIALS AND METHODS: Using identical beam data in each TPS, plans were created on uniform material synthetic phantoms with cubic targets. The following parameters were systematically varied in each TPS to observe their different implementations: spot spacing, layer spacing and margin. Additionally, plans were created in Eclipse to investigate the impact of these parameters on plan delivery and optimal values are suggested. RESULTS: It was found that all systems except Eclipse use a variable layer spacing per beam, based on the Bragg peak width of each energy layer. It is recommended that if this cannot be used, then a constant value of 5 mm will ensure good dose homogeneity. Only RayStation varies the spot spacing according to the variable spot size with depth. If a constant spot spacing is to be used, a value of 5 mm is recommended as a good compromise between dose homogeneity, plan robustness and planning time. It was found that both Pinnacle3 and RayStation position spots outside of the defined volume (target plus margin). CONCLUSIONS: All four systems are capable of delivering uniform dose distributions to simple targets, but their implementation of the various planning parameters is different. In this paper comparisons are made between the four systems and recommendations are made as to the values that will provide the best compromise in dose homogeneity and planning time.

A comparison of the dose distributions from three proton treatment planning systems in the planning of meningioma patients with single-field uniform dose pencil beam scanning.

With the number of new proton centers increasing rapidly, there is a need for an assessment of the available proton treatment planning systems (TPSs). This study compares the dose distributions of complex meningioma plans produced by three proton TPSs: Eclipse, Pinnacle3, and XiO. All three systems were commissioned with the same beam data and, as best as possible, matched configuration settings. Proton treatment plans for ten patients were produced on each system with a pencil beam scanning, single-field uniform dose approach, using a fixed horizontal beamline. All 30 plans were subjected to identical dose constraints, both for the target coverage and organ at risk (OAR) sparing, with a consistent order of priority. Beam geometry, lateral field margins, and lateral spot resolutions were made consistent across all systems. Few statistically significant differences were found between the target coverage and OAR sparing of each system, with all optimizers managing to produce plans within clinical tolerances (D2 < 107% of prescribed dose, D5 < 105%, D95 > 95%, D99 > 90%, and OAR maximum doses) despite strict constraints and overlapping structures.

Review of backscatter measurement in kilovoltage radiotherapy using novel detectors and reduction from lack of underlying scattering material.

Lack of underlying material can lead to dose reduction in kilovoltage radiotherapy treatments because of backscatter reduction. Conversely, the use of lead shielding can lead to large dose enhancement close to the lead interface. GAFCHROMIC film has been shown to be of use in verification of local backscatter factors compared to reference data in codes of practice, but careful handling and multiple readings are required to reduce systematic uncertainties to between 3% and 4%. Monte Carlo modeling of the specific treatment unit should be performed in cases which are found to differ from reference values before alternative values are adopted clinically, but these cases are expected to be few. GAFCHROMIC film may also be used to estimate backscatter reduction more readily than customized ionization chambers, for a range of beam qualities, applicator sizes and depth, with and without lead shielding. Differences were found between different studies, and it is not clear to what extent these are due to variation in equipment and/or technique. However, a layer of wax around lead shielding of 1 mm thickness should be sufficient to eliminate lead enhancement effects for all kilovoltage energies from 40 kV to 300 kV.

Evaluation of quality control strategies in Scutellaria herbal medicines.

The statutory regulation of herbal medicines is under review within the United Kingdom (UK) and by 2011 all herbal medicines will require either a Product Licence or a Traditional Herbal Registration. The species Scutellaria baicalensis has been shown to possess anti-inflammatory, anti-viral and anti-tumor properties and is one of the most widely used Chinese herbal extracts in Eastern and Western medicines. The bioactivity of this herbal medicine is due to the radical scavenging activities of the flavone components of which there are more than 60. This research has characterised 5 key flavones in 18 extracts of Scutellaria using a combination of HPLC with DAD and MS detection. Employing an internal standard approach, the validated HPLC method afforded good sensitivity and excellent assay precision. Assays for the ferric reducing antioxidant power (FRAP) and total phenol determinations enabled determination of the antioxidant coefficient (PAC) of each Scutellaria extract. The potential usefulness of employing multivariate statistical analysis using a combination of the key parameters collected namely, FRAP activity, total phenol content, levels of 5 flavone biomarkers and the PAC as a means of quality evaluation of the Scutellaria herbal extracts was investigated. The PAC value was predicted by soft independent modelling of class analogy (SIMCA) as being the most discriminatory parameter and applying this ranking the herbal extracts were grouped into 3 clusters. The second most influential parameter in determining the clustering of the samples was the level of baicalin in each extract. It is proposed that the PAC value alone or in combination with a chromatographic fingerprint of key biomarkers [e.g. baicalin or (baicalin+baicalein)] may be useful indicators to adopt for the quality control of S. baicalensis.