Organ-based tube current modulation and bismuth eye shielding in pediatric head computed tomography.

BACKGROUND: Exposure of the eye lens to ionizing radiation results in cataract. Several dose optimization techniques to protect the lens are available for computed tomography (CT). OBJECTIVE: The radiation dose to the eye lens, volume CT dose index (CTDIvol) and image quality of various methods of dose optimization were evaluated for pediatric head CT: automated tube current modulation (ATCM), automated tube voltage selection (ATVS), organ-based tube current modulation (OBTCM) and bismuth shielding. MATERIALS AND METHODS: An anthropomorphic phantom of a 5-year-old child was scanned with nine protocols: no dose optimization technique and then adding different dose optimization techniques alone and in combination. Dose to the eye, thyroid and breast were estimated using metal oxide semiconductor field effect transistor (MOSFET) dosimetry. CTDIvol, influence of timing of shield placement, image noise and attenuation values in 13 regions of interest of the head and subjective image quality were compared. RESULTS: The eye shield significantly reduced the eye lens dose when used alone, to a similar degree as when using all software-based techniques together. When used in combination with software-based techniques, the shield reduced the eye lens dose by up to 45% compared to the no dose optimization technique. Noise was significantly increased by the shield, most pronounced in the anterior portion of the eye. CONCLUSION: The combination of ATCM, ATVS, OBTCM and a bismuth shield, with the shield placed after acquiring the localizer image, should be considered to reduce the radiation dose to the eye lens in pediatric head CT.

Teleradiotherapy Network: Applications and Feasibility for Providing Cost-Effective Comprehensive Radiotherapy Care in Low- and Middle-Income Group Countries for Cancer Patients.

Globally, new cancer cases will rise by 57% within the next two decades, with the majority in the low- and middle-income countries (LMICs). Consequently, a steep increase of about 40% in cancer deaths is expected there, mainly because of lack of treatment facilities, especially radiotherapy. Radiotherapy is required for more than 50% of patients, but the capital cost for equipment often deters establishment of such facilities in LMICs. Presently, of the 139 LMICs, 55 do not even have a radiotherapy facility, whereas the remaining 84 have a deficit of 61.4% of their required radiotherapy units. Networking between centers could enhance the effectiveness and reach of existing radiotherapy in LMICs. A teleradiotherapy network could enable centers to share and optimally utilize their resources, both infrastructure and staffing. This could be in the form of a three-tier radiotherapy service consisting of primary, secondary, and tertiary radiotherapy centers interlinked through a network. The concept has been adopted in some LMICs and could also be used as a "service provider model," thereby reducing the investments to set up such a network. Teleradiotherapy networks could be a part of the multipronged approach to address the enormous gap in radiotherapy services in a cost-effective manner and to support better accessibility to radiotherapy facilities, especially for LMICs.

Probabilistic U-Net model observer for the DDC method in CT scan protocol optimization.

Optimizing complex imaging procedures within Computed Tomography, considering both dose and image quality, presents significant challenges amidst rapid technological advancements and the adoption of machine learning (ML) methods. A crucial metric in this context is the Difference-Detailed Curve, which relies on human observer studies. However, these studies are labor-intensive and prone to both inter- and intra-observer variability. To tackle these issues, a ML-based model observer utilizing the U-Net architecture and a Bayesian methodology is proposed. In order to train a model observer unaffected by the spatial arrangement of low-contrast objects, the image preprocessing incorporates a Gaussian Process-based noise model. Additionally, gradient-weighted class activation mapping is utilized to gain insights into the model observer's decision-making process. By training on data from a diverse group of observers, well-calibrated probabilistic predictions that quantify observer variability are achieved. Leveraging the principles of Beta regression, the Bayesian methodology is used to derive a model observer performance metric, effectively gauging the model observer's strength in terms of an 'effective number of observers'. Ultimately, this framework enables to predict the DDC distribution by applying thresholds to the inferred probabilities (Part of this work has been presented at: Stocker D, Sommer C, Gueng S, Stäuble J, Özden I, Griessinger J, Weyland M S, Lutters G, Scheidegger S (2023). Probabilistic U-Net Model Observer for the DDC Method in CT Scan Protocol Optimization. The 56th SSRMP Annual Meeting 2023, November 30. - December 1., 2023, Luzern, Switzerland).

Anisotropic phantom measurements for quality assured use of diffusion tensor imaging in clinical practice.

BACKGROUND: Diffusion-weighted magnetic resonance imaging (MRI) is being increasingly applied in clinical practice, for example in neuronavigation and in modern radiation treatment planning. Quality assurance (QA) is therefore important to avoid clinical errors. PURPOSE: To compare four analytical programs and a neuronavigation tool to evaluate our in-house diffusion-weighted imaging protocol in order to be able to implement diffusion tensor imaging (DTI) into clinical practice. MATERIAL AND METHODS: A phantom containing crossing fibers was used for the QA. Fiber tracking and fractional anisotropy (FA) analyses were performed, and the geometrical resolution was verified using the phantom. RESULTS: FA results were reproducible within each program and no significant differences between programs were observed. Also, no significant differences in FA values were found when comparing the results between the four software programs. Geometrical resolution of the anatomical data-set was satisfactory; however the crossing of the fibers was not accurately represented by three of the four programs. CONCLUSION: Phantom QA is necessary before using DTI for novel procedures to identify the uncertainties associated with DTI data. It is important to remember that the results are software-dependent and that representation of the tracts may vary between software products. We therefore recommend caution with regard to the application of fiber tracking results intraoperatively when dealing with abnormal fiber tract anatomy.

Feasibility of a method for low contrast CT image quality assessment using difference detail curves for abdominal scans.

This work describes a measurement method for assessing dose-related image-quality of CT scans based on the difference detail curve (DDC) method, and showcases its use in a low contrast setting. The method is based on a phantom consisting of elliptical slices of different sizes into which contrast object modules can be inserted. These modules contain contrast objects based on (synthetic) resin mixtures with sucrose (native) or sodium iodine (contrast medium). Mixing ratios are provided to achieve a range of clinically relevant CT-numbers with these materials. The phantom is characterized in terms of contrast accuracy, energy dependency and long-term drift with satisfying results. Contrast accuracy and energy dependency are similar to that of water or soft tissue. Image quality of 655 scans of the phantom acquired at 30 different clinical institutions and with 16 different CT scanner models from 4 manufacturers was assessed by calculating a difference detail curve (DDC) from evaluation of up to 5 human observers using a custom-made software (RadiVates) described in this work. Based on these measurements, inter-observer variability was quantified using a bootstrap method and was shown to be a large contributor to the overall variability. This work demonstrates that assessment of CT image quality is feasible with the aforementioned phantom and DDC method.

A LQ-based kinetic model formulation for exploring dynamics of treatment response of tumours in patients.

A kinetic bio-mathematical, linear-quadratic (LQ) based model description for clonogenic survival is presented. In contrast to widely used formulations of models, a dynamic approach based on ordinary differential equations for coupling a repair model with a tumour growth model is used to allow analysis of intercellular process dynamics and submodel interference. The purpose of the model formulation is to find a quantitative framework for investigation of tumour response to radiotherapy in vivo. It is not the intention of the proposed model formulation to give a mechanistic explanation for cellular repair processes. This article addresses bio-mathematical aspects of the simplistic kinetic approach used for description of repair. The model formulation includes processes for cellular death, repopulation and cellular repair. The explicit use of the population size in the model facilitates the coupling of the sub-models including aspects of tissue dynamics (competition, oxygenation). The cellular repair is summarized by using a kinetic model for a dose equivalent Γ describing production and elimination of sublethal lesions. This dose equivalent replaces the absorbed dose used in the common LQ- model. Therefore, this approach is called the Γ- LQ- formulation. A comparison with two kinetic radiobiological models (the LPL model of Curtis and the compartmental model of Carlone) is carried out. The resulting differential equations are solved by numerical integration using a Runge-Kutta algorithm. The comparison reveals a good agreement between the Γ- LQ- formulation and the models of Curtis and Carlone under certain, defined conditions: The proposed formulation leads to results which are identical to the model of Carlone over a wide range of investigated biological parameters and different fractionation schemes when using first order repair kinetics. The comparison with experimental data and the LPL- model of Curtis shows a good agreement of the Γ- LQ- formulation using second order repair kinetics over a wide range of dose rate. Over a limited range, the use of second order repair in the Γ- LQ- formulation approximates the same dose rate dependency of clonogenic survival using only one additional parameter to those of the common LQ model. Within the investigated range of parameters, the presented Γ-LQ- formulation may be used to describe the in-vivo tumour response to radiation. The influence of repopulation, oxygenation and other aspects of tissue dynamics may override the differences between the intrinsic radiosensitivity yielded by each of the models. The proposed model formulation can be extended with additional static and dynamic tissue behaviours. This may be useful for the understanding of the reaction of tissues to heat (hyperthermia) or combined anti-cancer treatments (chemo-radiotherapy).

Rotational IMRT techniques compared to fixed gantry IMRT and tomotherapy: multi-institutional planning study for head-and-neck cases.

BACKGROUND: Recent developments enable to deliver rotational IMRT with standard C-arm gantry based linear accelerators. This upcoming treatment technique was benchmarked in a multi-center treatment planning study against static gantry IMRT and rotational IMRT based on a ring gantry for a complex parotid gland sparing head-and-neck technique. METHODS: Treatment plans were created for 10 patients with head-and-neck tumours (oropharynx, hypopharynx, larynx) using the following treatment planning systems (TPS) for rotational IMRT: Monaco (ELEKTA VMAT solution), Eclipse (Varian RapidArc solution) and HiArt for the helical tomotherapy (Tomotherapy). Planning of static gantry IMRT was performed with KonRad, Pinnacle and Panther DAO based on step&shoot IMRT delivery and Eclipse for sliding window IMRT. The prescribed doses for the high dose PTVs were 65.1Gy or 60.9Gy and for the low dose PTVs 55.8Gy or 52.5Gy dependend on resection status. Plan evaluation was based on target coverage, conformity and homogeneity, DVHs of OARs and the volume of normal tissue receiving more than 5Gy (V5Gy). Additionally, the cumulative monitor units (MUs) and treatment times of the different technologies were compared. All evaluation parameters were averaged over all 10 patients for each technique and planning modality. RESULTS: Depending on IMRT technique and TPS, the mean CI values of all patients ranged from 1.17 to 2.82; and mean HI values varied from 0.05 to 0.10. The mean values of the median doses of the spared parotid were 26.5Gy for RapidArc and 23Gy for VMAT, 14.1Gy for Tomo. For fixed gantry techniques 21Gy was achieved for step&shoot+KonRad, 17.0Gy for step&shoot+Panther DAO, 23.3Gy for step&shoot+Pinnacle and 18.6Gy for sliding window.V5Gy values were lowest for the sliding window IMRT technique (3499 ccm) and largest for RapidArc (5480 ccm). The lowest mean MU value of 408 was achieved by Panther DAO, compared to 1140 for sliding window IMRT. CONCLUSIONS: All IMRT delivery technologies with their associated TPS provide plans with satisfying target coverage while at the same time respecting the defined OAR criteria. Sliding window IMRT, RapidArc and Tomo techniques resulted in better target dose homogeneity compared to VMAT and step&shoot IMRT. Rotational IMRT based on C-arm linacs and Tomotherapy seem to be advantageous with respect to OAR sparing and treatment delivery efficiency, at the cost of higher dose delivered to normal tissues. The overall treatment plan quality using Tomo seems to be better than the other TPS technology combinations.

Postoperative proton radiotherapy for localized and locoregional breast cancer: potential for clinically relevant improvements?

PURPOSE: To study the potential reduction of dose to organs at risk (OARs) with intensity-modulated proton radiotherapy (IMPT) compared with intensity-modulated radiotherapy (IMRT) and three-dimensional conformal radiotherapy (3D-CRT) photon radiotherapy for left-sided breast cancer patients. METHODS AND MATERIALS: Comparative treatment-planning was performed using planning computed tomography scans of 20 left-sided breast cancer patients. For each patient, three increasingly complex locoregional volumes (planning target volumes [PTVs]) were defined: whole breast (WB) or chest wall (CW) = (PTV1), WB/CW plus medial-supraclavicular (MSC), lateral-supraclavicular (LSC), and level III axillary (AxIII) nodes = (PTV2) and WB/CW+MSC+LSC+AxIII plus internal mammary chain = (PTV3). For each patient, 3D-CRT, IMRT, and IMPT plans were optimized for PTV coverage. Dose to OARs was compared while maintaining target coverage. RESULTS: All the techniques met the required PTV coverage except the 3D-CRT plans for PTV3-scenario. All 3D-CRT plans for PTV3 exceeded left-lung V20. IMPT vs. 3D-CRT: significant dose reductions were observed for all OARs using IMPT for all PTVs. IMPT vs. IMRT: For PTV2 and PTV3, low (V5) left lung and cardiac doses were reduced by a factor >2.5, and cardiac doses (V22.5) were by a factor of >20 lower with IMPT compared with IMRT. CONCLUSIONS: When complex-target irradiation is needed, 3D-CRT often compromises the target coverage and increases the dose to OARs; IMRT can provide better results but will increase the integral dose. The benefit of IMPT is based on improved target coverage and reduction of low doses to OARs, potentially reducing the risk of late-toxicity. These results indicate a potential role of proton-radiotherapy for extended locoregional irradiation in left breast cancer.