RESUMEN
PURPOSE: We present a new method for knowledge-based isocenter selection for treatment planning in radiosurgery. Our objective is to develop a prediction model that can learn from past manually designed treatment plans. We leverage recent advances in deep learning to predict isocenter locations in treatment plans in order to provide a decision support tool. METHODS: The proposed method adapts a geometric approach using orthogonal moment expansions as a feature vector for describing the shape of the tumor. Our approach accounts primarily for tumor shape and OAR proximity, the two factors that are known to greatly affect the isocenter placement. We solve the prediction problem by training a residual neural network with skip connections on the formed shape descriptors. Our network was trained on 533 patient cases and was validated on a set of out-of-sample cases. RESULTS: Our method generates heatmap predictions for isocenter locations that are in most cases comparable to the experienced human planners, which shows that the method can be used in treatment planning to guide the users for determining the isocenters. CONCLUSIONS: Our numerical experiments indicate a positive predictive value on an independent validation set when compared against a test dataset that was not seen by the model during training.
Asunto(s)
Radiocirugia , Humanos , Bases del Conocimiento , Dosificación Radioterapéutica , Planificación de la Radioterapia Asistida por ComputadorRESUMEN
Stereotactic radiosurgery is an effective technique to treat brain tumors for which several inverse planning methods may be appropriate. We propose an integer programming model to simultaneous sector duration and isocenter optimization (SDIO) problem for Leksell Gamma Knife® IconTM (Elekta, Stockholm, Sweden) to tractably incorporate treatment time. We devise a Benders decomposition scheme to solve the SDIO problem to optimality. The performances of our approaches are assessed using anonymized data from eight previously treated cases, and obtained treatment plans are compared against each other and against the clinical plans. The plans generated by our SDIO model all meet or exceed clinical guidelines while demonstrating high conformity.
Asunto(s)
Neoplasias Encefálicas/cirugía , Neoplasias del Tronco Encefálico/cirugía , Cóclea/efectos de la radiación , Radiocirugia/métodos , Radiocirugia/normas , Planificación de la Radioterapia Asistida por Computador/normas , Humanos , Planificación de la Radioterapia Asistida por Computador/métodosRESUMEN
Stereotactic radiosurgery (SRS) is an effective technique to treat brain metastasis for which several inverse planning methods may be appropriate. We compare three different optimization models for segment duration optimization in SRS using Leksell Gamma Knife® IconTM (Elekta, Stockholm, Sweden). We investigate (1) a linear programming approach, (2) a piecewise quadratic penalty approach, and (3) an unconstrained convex moment-based penalty approach. We examine the performances of these approaches using anonymized data from 14 previously treated cases. In addition, we investigate the important modeling question of selecting weights for the objective functions where we use a simulated annealing algorithm to determine these weights for each model. The inverse plans obtained via optimization models are compared against each other and against the clinical plans. The three inverse planning models can all yield optimal treatment plans in a reasonable amount of time and the treatment plans obtained by these models meet or exceed clinical guidelines while displaying high conformity.
Asunto(s)
Neoplasias Encefálicas/radioterapia , Radiocirugia/métodos , Planificación de la Radioterapia Asistida por Computador/métodos , Algoritmos , Humanos , Dosificación RadioterapéuticaRESUMEN
Purpose: To evaluate early perfusion changes in normal tissue following stereotactic radiosurgery (SRS). Methods: Nineteen patients harboring twenty-two brain metastases treated with SRS were imaged with dynamic susceptibility magnetic resonance imaging (DSC MRI) at baseline, 1 week and 1 month post SRS. Relative cerebral blood volume and flow (rCBV and rCBF) ratios were evaluated outside of tumor within a combined region of interest (ROI) and separately within gray matter (GM) and white matter (WM) ROIs. Three-dimensional dose distribution from each SRS plan was divided into six regions: (1) <2 Gy; (2) 2-5 Gy; (3) 5-10 Gy; (4) 10-12 Gy; (5) 12-16 Gy; and (6) >16 Gy. rCBV and rCBF ratio differences between baseline, 1 week and 1 month were compared. Best linear fit plots quantified normal tissue dose-dependency. Results: Significant rCBV ratio increases were present between baseline and 1 month for all ROIs and dose ranges except for WM ROI receiving <2 Gy. rCBV ratio for all ROIs was maximally increased from baseline to 1 month with the greatest changes occurring within the 5-10 Gy dose range (53.1%). rCBF ratio was maximally increased from baseline to 1 month for all ROIs within the 5-10 Gy dose range (33.9-45.0%). Both rCBV and rCBF ratios were most elevated within GM ROIs. A weak, positive but not significant association between dose, rCBV and rCBF ratio was demonstrated. Progressive rCBV and rCBF ratio increased with dose up to 10 Gy at 1 month. Conclusion: Normal tissue response following SRS can be characterized by dose, tissue, and time specific increases in rCBV and rCBF ratio.
RESUMEN
PURPOSE: To investigate the visibility of simulated lesions in digital breast tomosynthesis (BT) image volumes compared with 2D digital mammography (DM). METHODS: Simulated lesions (masses and microcalcifications) were added to images of the same women acquired on a DM system (Mammomat Novation, Siemens) and a BT prototype. The same beam quality was used for the DM and BT acquisitions. The total absorbed dose resulting from a 25-projection BT acquisition and reconstruction (BT(25)) was approximately twice that of a single DM view. By excluding every other projection image from the reconstruction (BT(13)), approximately the same dose as in DM was effected. Simulated microcalcifications were digitally added with varying contrast to the DM and BT images. Simulated masses with 8 mm diameter were also added to BT images. A series of 4-alternative forced choice (4AFC) human observer experiments were conducted. Four medical physicists participated in all experiments, each consisting of 60 trials per experimental condition. The observers interpreted the BT image volumes in cine-mode at a fixed image sequence speed. The required threshold contrast (S(t)) to achieve a detectability index (d') of 2.5 (i.e., 92.5% correct decisions) was determined. RESULTS: The S(t) for mass detection in DM was approximately a factor of 2 higher than required in BT indicating that the detection of masses was improved under BT conditions compared to DM. S(t) for microcalcification detection was higher for BT than for DM at both BT dose levels (BT(25) and BT(13)), with a statistically significant difference in S(t) between DM and BT(13). These results indicate a dose-dependent decrease in detection performance in BT for detection of microcalcifications. CONCLUSIONS: In agreement with previous investigations, masses of size 8 mm can be detected with less contrast in BT than in DM indicating improved detection performance for BT. However, for the investigated microcalcifications, the results of this study indicate potentially worse performance for BT than for DM at the same dose level.
Asunto(s)
Neoplasias de la Mama/diagnóstico por imagen , Calcinosis/diagnóstico por imagen , Procesamiento de Imagen Asistido por Computador/métodos , Mamografía/métodos , Femenino , Humanos , Sensibilidad y EspecificidadRESUMEN
PURPOSE: The purpose of this work is to develop an automated inverse planning approach to generate singe-fraction and fractionated stereotactic radiosurgery (SRS) treatment plans for Gamma Knife Perfexion. METHODS: Our automated approach consists of two steps: 1) a grassfire-based algorithm to carefully determine the isocentre locations; 2) a penalty-based optimization to find the optimal shot shapes and their intensities to minimize the deviation of the delivered dose from the objective dose in all structures. For single-fraction SRS, a margin-less approach was taken: conformity of dose to the gross tumor volume (GTV) with a steep dose fall-off was prioritized. For fractionated radiosurgery, dose homogeneity was given a higher priority since planning target volumes (PTV) were applied to account for daily setup variation, and these PTVs could overlap with organs-at-risk (OARs). The two-step approach was tested on seven clinical cases with PTV sizes of 0.5cmÌ3-56.5cmÌ3. In the tested cases, the PTV had 0%-38% overlap with OARs. RESULTS: For single-fraction SRS, the dose to 1mmÌ3 brainstem was on average 0.24Gy (range: -2.4Gy to +2.0Gy) lower compared to manually-generated plans. Beam-on time varied with the number of isocentres, but on average was 33min longer than manually- generated plans. The optimization algorithm took 215min on average, while isocentre selection performed in <10s.For fractionated SRS, the average PTV coverage was V95=94.9% (range: 92.7%-97.6%) and the mean dose to 1 mmÌ3 brainstem was 87.8% of the prescription dose (range: 35.4%- 108.8%). The mean beam-on time per fraction per dose-per-fraction was 4.8min/Gy (range: 0.9min/Gy-10.3min/Gy). We observed a tradeoff between conformity and OARs-sparing in both plans, and added sensitivity to isocentre locations in fractionated plans. In all the cases, GTV received the full prescription dose. CONCLUSIONS: The results indicated that automated inverse planning yields improved conformity and OAR-sparing for single- fraction SRS and is capable of generating homogeneous fractionated SRS. This work is partially funded by Elekta Instrument, AB, Stockholm, Sweden.
RESUMEN
PURPOSE: The purpose of this work was to evaluate the visibility of simulated lesions in 2D digita mammography (DM) and breast tomosynthesis (BT) images of patients. METHODS: Images of the same women were acquired on both a DM system (Mammomat Novation, Siemens Healthcare, Erlangen, Germany) and a BT prototype system adapted from the same type of DM system. Using the geometrical properties of the two systems, simulated lesions were projected and added to each DM image as well as to each BT projection image prior to 3D reconstruction. The same beam quality and approximately the same total absorbed dose to the glandular tissue were used for each breast image acquisition on the two systems. A series of four-alternative forced choice human observer experiments was conducted for each of five simulated lesion diameters: 0.2, 1, 3, 8, and 25 mm. An additional experiment was conducted for the 0.2 mm lesion in BT only at twice the dose level (BT2x). Threshold signal was defined as the lesion signal intensity required for a detectability index (d') of 2.5. Four medical physicists participated in all experiments. One experiment, consisting of 60 cases, was conducted per test condition (i.e., lesion size and signal combination). RESULTS: For the smallest lesions (0.2 mm), the threshold signal for DM was 21% lower than for BT at equivalent dose levels, and BT2x was 26% lower than DM. For the lesions larger than 1 mm, the threshold signal increased linearly (in log space) with the lesion diameter for both DM and BT, with DM requiring around twice the signal as BT. The difference in the threshold signal between BT and DM at each lesion size was statistically significant, except for the 0.2 mm lesion between BT2x and DM. CONCLUSIONS: The results of this study indicate that low-signal lesions larger than 1.0 mm may be more visible in BT compared to DM, whereas 0.2 mm lesions may be better visualized with DM compared to BT, when compared at equal dose.
Asunto(s)
Neoplasias de la Mama/diagnóstico , Mama/patología , Mamografía/métodos , Algoritmos , Neoplasias de la Mama/patología , Simulación por Computador , Femenino , Humanos , Procesamiento de Imagen Asistido por Computador/métodos , Imagenología Tridimensional , Modelos Estadísticos , Variaciones Dependientes del Observador , Fantasmas de Imagen , Reproducibilidad de los Resultados , Programas Informáticos , Interfaz Usuario-ComputadorRESUMEN
This study aimed to investigate the effect of dose reduction on diagnostic accuracy and radiation risk in digital mammography. Simulated masses and microcalcifications were positioned in an anthropomorphic breast phantom. Thirty digital images, 14 with lesions, 16 without, were acquired of the phantom using a Mammomat Novation (Siemens, Erlangen, Germany) at each of three dose levels. These corresponded to 100%, 50% and 30% of the normally used average glandular dose (AGD; 1.3 mGy for a standard breast). Eight observers interpreted the 90 unprocessed images in a free response study, and the data were analysed with the jackknife free response receiver operating characteristic (JAFROC) method. Observer performance was assessed using the JAFROC figure of merit (FOM). The benefit of radiation risk reduction was estimated based on several risk models. There was no statistically significant difference in performance, as described by the FOM, between the 100% and the 50% dose levels. However, the FOMs for both the 100% and the 50% dose were significantly different from the corresponding quantity for the 30% dose level (F-statistic = 4.95, p-value = 0.01). A dose reduction of 50% would result in three to nine fewer breast cancer fatalities per 100,000 women undergoing annual screening from the age of 40 to 49 years. The results of the study indicate a possibility of reducing the dose to the breast to half the dose level currently used. This has to be confirmed in clinical studies, and possible differences depending on lesion type should be examined further.
Asunto(s)
Neoplasias de la Mama/diagnóstico , Mama/efectos de la radiación , Mamografía/métodos , Adulto , Neoplasias de la Mama/etiología , Calcinosis/diagnóstico por imagen , Femenino , Humanos , Mamografía/efectos adversos , Persona de Mediana Edad , Variaciones Dependientes del Observador , Fantasmas de Imagen , Curva ROC , Dosis de Radiación , Medición de Riesgo , Sensibilidad y EspecificidadRESUMEN
There is a need for tools that in a simple way can be used for the evaluation of image quality related to clinical requirements in mammography. The aim of this work was to adjust the present European image quality criteria to be relevant also for digital mammography images, and to use as simple and as few criteria as possible. A pilot evaluation of the new set of criteria was made with mammograms of 28 women from a General Electric Senographe 2000D full-field digital mammography system. One breast was exposed using the standard automatic exposure mode, the other using about half of that absorbed dose. Three experienced radiologists evaluated the images using visual grading analysis technique. The results indicate that the new quality criteria can be used for the evaluation of image quality related to clinical requirements in digital mammography in a simple way. The results also suggest that absorbed doses for the mammography system used may be substantially reduced.
Asunto(s)
Mamografía/instrumentación , Mamografía/métodos , Interpretación de Imagen Radiográfica Asistida por Computador/métodos , Anciano , Mama/patología , Europa (Continente) , Femenino , Humanos , Procesamiento de Imagen Asistido por Computador , Mamografía/normas , Persona de Mediana Edad , Proyectos Piloto , Dosis de Radiación , Intensificación de Imagen Radiográfica , Radiometría , Pantallas Intensificadoras de Rayos XRESUMEN
The European Commission (EC) quality criteria for screen-film mammography are used as a tool to assess image quality. A new set of criteria was developed and initially tested in a previous study. In the present study, these criteria are further evaluated using screen-film mammograms that have been digitised, manipulated to simulate different image quality levels and reprinted on film. Expert radiologists have evaluated these manipulated images using both the original (EC) and the new criteria. A comparison of three different simulated dose levels reveals that the new criteria yield a larger separation of image criteria scores than the old ones. These results indicate that the new set of image quality criteria has a higher discriminative power than the old set and thus seems to be more suitable for evaluation of image quality in mammography.
Asunto(s)
Mamografía/instrumentación , Mamografía/métodos , Interpretación de Imagen Radiográfica Asistida por Computador/métodos , Europa (Continente) , Estudios de Evaluación como Asunto , Femenino , Humanos , Procesamiento de Imagen Asistido por Computador , Mamografía/normas , Modelos Estadísticos , Dosis de Radiación , Intensificación de Imagen Radiográfica/métodos , Magnificación Radiográfica , Radiología/instrumentación , Radiología/normas , Tecnología Radiológica , Pantallas Intensificadoras de Rayos XRESUMEN
The effect of pixel size on shape determination in screening digital mammography systems was studied using a shape identification task as the measured outcome. Ten microcalcifications on screen-films were digitised to a range of pixel sizes (2.5-200 microm) and extracted from computed radiography (CR) images (50 microm) acquired under equivalent imaging conditions. Fifteen observers attempted to identify the shape of each microcalcification at each pixel size. The results were collated to provide a fraction of correct responses vs. pixel size curve for each microcalcification. Averaging over all shapes, pixel values >100 microm lead to a significant decrease in shape determination ability (p < 0.01) for digitised screen-film. For CR images, half the shapes were not properly identified. Hence, although 20-100 microm was sufficient for microcalcification shape determination for digitised screen-film images, 50 microm was only borderline sufficient for the CR digital images.
Asunto(s)
Enfermedades de la Mama/diagnóstico , Calcinosis/diagnóstico , Mamografía/instrumentación , Mamografía/métodos , Interpretación de Imagen Radiográfica Asistida por Computador/métodos , Magnificación Radiográfica/métodos , Femenino , Humanos , Proyectos PilotoRESUMEN
In this study a set of structures has been simulated to represent a range of clinically relevant breast cancer mammographic lesions including solid tumours and microcalcifications. All structures have been created using simple random-based mathematical functions and have been inserted into a subset of digital mammography images at appropriate contrast levels into various regions of the breast, including dense fibroglandular and adipose tissue. These structures and their appearance in these clinical images were evaluated in terms of how realistic they looked. They will be used as the input to a large-scale clinical trial designed to examine the effect of significant dose reduction in digital mammography by comparing the detectability of such structures in images acquired at full and quarter automatic exposure control (AEC) dose level and in images with simulated noise levels in between.