Geometric and dosimetric analysis of CT- and MR-based automatic contouring for the EPTN contouring atlas in neuro-oncology.

PURPOSE: Atlas-based and deep-learning contouring (DLC) are methods for automatic segmentation of organs-at-risk (OARs). The European Particle Therapy Network (EPTN) published a consensus-based atlas for delineation of OARs in neuro-oncology. In this study, geometric and dosimetric evaluation of automatically-segmented neuro-oncological OARs was performed using CT- and MR-models following the EPTN-contouring atlas. METHODS: Image and contouring data from 76 neuro-oncological patients were included. Two atlas-based models (CT-atlas and MR-atlas) and one DLC-model (MR-DLC) were created. Manual contours on registered CT-MR-images were used as ground-truth. Results were analyzed in terms of geometrical (volumetric Dice similarity coefficient (vDSC), surface DSC (sDSC), added path length (APL), and mean slice-wise Hausdorff distance (MSHD)) and dosimetrical accuracy. Distance-to-tumor analysis was performed to analyze to which extent the location of the OAR relative to planning target volume (PTV) has dosimetric impact, using Wilcoxon rank-sum tests. RESULTS: CT-atlas outperformed MR-atlas for 22/26 OARs. MR-DLC outperformed MR-atlas for all OARs. Highest median (95 %CI) vDSC and sDSC were found for the brainstem in MR-DLC: 0.92 (0.88-0.95) and 0.84 (0.77-0.89) respectively, as well as lowest MSHD: 0.27 (0.22-0.39)cm. Median dose differences (ΔD) were within ± 1 Gy for 24/26(92 %) OARs for all three models. Distance-to-tumor showed a significant correlation for ΔDmax,0.03cc-parameters when splitting the data in ≤ 4 cm and > 4 cm OAR-distance (p < 0.001). CONCLUSION: MR-based DLC and CT-based atlas-contouring enable high-quality segmentation. It was shown that a combination of both CT- and MR-autocontouring models results in the best quality.

Real-world analysis of manual editing of deep learning contouring in the thorax region.

Background and purpose: User-adjustments after deep-learning (DL) contouring in radiotherapy were evaluated to get insight in real-world editing during clinical practice. This study assessed the amount, type and spatial regions of editing of auto-contouring for organs-at-risk (OARs) in routine clinical workflow for patients in the thorax region. Materials and methods: A total of 350 lung cancer and 362 breast cancer patients, contoured between March 2020 and March 2021 using a commercial DL-contouring method followed by manual adjustments were retrospectively analyzed. Subsampling was performed for some OARs, using an inter-slice gap of 1-3 slices. Commonly-used whole-organ contouring assessment measures were calculated, and all cases were registered to a common reference shape per OAR to identify regions of manual adjustment. Results were expressed as the median, 10th-90th percentile of adjustment and visualized using 3D renderings. Results: Per OAR, the median amount of editing was below 1 mm. However, large adjustments were found in some locations for most OARs. In general, enlarging of the auto-contours was needed. Subsampling DL-contours showed less adjustments were made in the interpolated slices compared to simulated no-subsampling for these OARs. Conclusion: The real-world performance of automatic DL-contouring software was evaluated and proven useful in clinical practice. Specific regions-of-adjustment were identified per OAR in the thorax region, and separate models were found to be necessary for specific clinical indications different from training data. This analysis showed the need to perform routine clinical analysis especially when procedures or acquisition protocols change to have the best configuration of the workflow.

Fully Automated 3-D Ultrasound Segmentation of the Placenta, Amniotic Fluid, and Fetus for Early Pregnancy Assessment.

Volumetric placental measurement using 3-D ultrasound has proven clinical utility in predicting adverse pregnancy outcomes. However, this metric cannot currently be employed as part of a screening test due to a lack of robust and real-time segmentation tools. We present a multiclass (MC) convolutional neural network (CNN) developed to segment the placenta, amniotic fluid, and fetus. The ground-truth data set consisted of 2093 labeled placental volumes augmented by 300 volumes with placenta, amniotic fluid, and fetus annotated. A two-pathway, hybrid (HB) model using transfer learning, a modified loss function, and exponential average weighting was developed and demonstrated the best performance for placental segmentation (PS), achieving a Dice similarity coefficient (DSC) of 0.84- and 0.38-mm average Hausdorff distances (HDAV). The use of a dual-pathway architecture improved the PS by 0.03 DSC and reduced HDAV by 0.27 mm compared with a naïve MC model. The incorporation of exponential weighting produced a further small improvement in DSC by 0.01 and a reduction of HDAV by 0.44 mm. Per volume inference using the FCNN took 7-8 s. This method should enable clinically relevant morphometric measurements (such as volume and total surface area) to be automatically generated for the placenta, amniotic fluid, and fetus. The ready availability of such metrics makes a population-based screening test for adverse pregnancy outcomes possible.

Assessment of manual adjustment performed in clinical practice following deep learning contouring for head and neck organs at risk in radiotherapy.

BACKGROUND AND PURPOSE: Auto-contouring performance has been widely studied in development and commissioning studies in radiotherapy, and its impact on clinical workflow assessed in that context. This study aimed to evaluate the manual adjustment of auto-contouring in routine clinical practice and to identify improvements regarding the auto-contouring model and clinical user interaction, to improve the efficiency of auto-contouring. MATERIALS AND METHODS: A total of 103 clinical head and neck cancer cases, contoured using a commercial deep-learning contouring system and subsequently checked and edited for clinical use were retrospectively taken from clinical data over a twelve-month period (April 2019-April 2020). The amount of adjustment performed was calculated, and all cases were registered to a common reference frame for assessment purposes. The median, 10th and 90th percentile of adjustment were calculated and displayed using 3D renderings of structures to visually assess systematic and random adjustment. Results were also compared to inter-observer variation reported previously. Assessment was performed for both the whole structures and for regional sub-structures, and according to the radiation therapy technologist (RTT) who edited the contour. RESULTS: The median amount of adjustment was low for all structures (<2 mm), although large local adjustment was observed for some structures. The median was systematically greater or equal to zero, indicating that the auto-contouring tends to under-segment the desired contour. CONCLUSION: Auto-contouring performance assessment in routine clinical practice has identified systematic improvements required technically, but also highlighted the need for continued RTT training to ensure adherence to guidelines.

3D ultrasound file reading and coordinate transformations.

The Kretzfile format is used to store 3D ultrasound data, from GE Voluson ultrasound scanners. The geometry used in these hies is a toroidal coordinate system. Cartesian coordinates are required to allow application of advanced image libraries like ITK and scikit-image. We present ITK transformation and utilities to convert Kretzfiles to cartesian coordinates. Previous work (SlicerHeart, 2017) has enabled the reading of kretz files and approximate coordinate transformations. This work will enable medical imaging researchers to investigate clinically 3D ultrasound.

Three-dimensional US Fractional Moving Blood Volume: Validation of Renal Perfusion Quantification.

Background Three-dimensional (3D) fractional moving blood volume (FMBV) derived from 3D power Doppler US has been proposed for noninvasive approximation of perfusion. However, 3D FMBV has never been applied in animals against a ground truth. Purpose To determine the correlation between 3D FMBV and the reference standard of fluorescent microspheres (FMS) for measurement of renal perfusion in a porcine model. Materials and Methods From February 2017 to September 2017, adult pigs were administered FMS before and after measurement of renal 3D FMBV at baseline (100%) and approximately 75%, 50%, and 25% flow levels by using US machines from two different vendors. The 3D power Doppler US volumes were converted and segmented, and correlations between FMS and 3D FMBV were made with simple linear regression (r2). Similarity and reproducibility of manual segmentation were determined with the Dice similarity coefficient and 3D FMBV reproducibility (intraclass correlation coefficient [ICC]). Results Thirteen pigs were studied with 33 flow measurements. Kidney volume (mean Dice similarity coefficient ± standard deviation, 0.89 ± 0.01) and renal segmentation (coefficient of variation = 12.6%; ICC = 0.86) were consistent. The 3D FMBV calculations had high reproducibility (ICC = 0.97; 95% confidence interval: 0.96, 0.98). The 3D FMBV per-pig correlation showed excellent correlation for US machines from both vendors (mean r2 = 0.96 [range, 0.92-1.0] and 0.93 [range, 0.78-1.0], respectively). The correlation between 3D FMBV and perfusion measured with microspheres was high for both US machines (r2 = 0.80 [P < .001] and 0.70 [P < .001], respectively). Conclusion The strong correlation between three-dimensional (3D) fractional moving blood volume (FMBV) and fluorescent microspheres indicates that 3D FMBV shows excellent correlation to perfusion and good reproducibility. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Morrell et al in this issue.

Fully automated, real-time 3D ultrasound segmentation to estimate first trimester placental volume using deep learning.

We present a new technique to fully automate the segmentation of an organ from 3D ultrasound (3D-US) volumes, using the placenta as the target organ. Image analysis tools to estimate organ volume do exist but are too time consuming and operator dependant. Fully automating the segmentation process would potentially allow the use of placental volume to screen for increased risk of pregnancy complications. The placenta was segmented from 2,393 first trimester 3D-US volumes using a semiautomated technique. This was quality controlled by three operators to produce the "ground-truth" data set. A fully convolutional neural network (OxNNet) was trained using this ground-truth data set to automatically segment the placenta. OxNNet delivered state-of-the-art automatic segmentation. The effect of training set size on the performance of OxNNet demonstrated the need for large data sets. The clinical utility of placental volume was tested by looking at predictions of small-for-gestational-age babies at term. The receiver-operating characteristics curves demonstrated almost identical results between OxNNet and the ground-truth). Our results demonstrated good similarity to the ground-truth and almost identical clinical results for the prediction of SGA.

The effect of system geometry and dose on the threshold detectable calcification diameter in 2D-mammography and digital breast tomosynthesis.

Digital breast tomosynthesis (DBT) is under consideration to replace or to be used in combination with 2D-mammography in breast screening. The aim of this study was the comparison of the detection of microcalcification clusters by human observers in simulated breast images using 2D-mammography, narrow angle (15°/15 projections) and wide angle (50°/25 projections) DBT. The effects of the cluster height in the breast and the dose to the breast on calcification detection were also tested. Simulated images of 6 cm thick compressed breasts were produced with and without microcalcification clusters inserted, using a set of image modelling tools for 2D-mammography and DBT. Image processing and reconstruction were performed using commercial software. A series of 4-alternative forced choice (4AFC) experiments was conducted for signal detection with the microcalcification clusters as targets. Threshold detectable calcification diameter was found for each imaging modality with standard dose: 2D-mammography: 2D-mammography (165 ± 9 µm), narrow angle DBT (211 ± 11 µm) and wide angle DBT (257 ± 14 µm). Statistically significant differences were found when using different doses, but different geometries had a greater effect. No differences were found between the threshold detectable calcification diameters at different heights in the breast. Calcification clusters may have a lower detectability using DBT than 2D imaging.

The relationship between cancer detection in mammography and image quality measurements.

PURPOSE: To investigate the relationship between image quality measurements and the clinical performance of digital mammographic systems. METHODS: Mammograms containing subtle malignant non-calcification lesions and simulated malignant calcification clusters were adapted to appear as if acquired by four types of detector. Observers searched for suspicious lesions and gave these a malignancy score. Analysis was undertaken using jackknife alternative free-response receiver operating characteristics weighted figure of merit (FoM). Images of a CDMAM contrast-detail phantom were adapted to appear as if acquired using the same four detectors as the clinical images. The resultant threshold gold thicknesses were compared to the FoMs using a linear regression model and an F-test was used to find if the gradient of the relationship was significantly non-zero. RESULTS: The detectors with the best image quality measurement also had the highest FoM values. The gradient of the inverse relationship between FoMs and threshold gold thickness for the 0.25mm diameter disk was significantly different from zero for calcification clusters (p=0.027), but not for non-calcification lesions (p=0.11). Systems performing just above the minimum image quality level set in the European Guidelines for Quality Assurance in Breast Cancer Screening and Diagnosis resulted in reduced cancer detection rates compared to systems performing at the achievable level. CONCLUSIONS: The clinical effectiveness of mammography for the task of detecting calcification clusters was found to be linked to image quality assessment using the CDMAM phantom. The European Guidelines should be reviewed as the current minimum image quality standards may be too low.

Breast cancer detection rates using four different types of mammography detectors.

OBJECTIVE: To compare the performance of different types of detectors in breast cancer detection. METHODS: A mammography image set containing subtle malignant non-calcification lesions, biopsy-proven benign lesions, simulated malignant calcification clusters and normals was acquired using amorphous-selenium (a-Se) detectors. The images were adapted to simulate four types of detectors at the same radiation dose: digital radiography (DR) detectors with a-Se and caesium iodide (CsI) convertors, and computed radiography (CR) detectors with a powder phosphor (PIP) and a needle phosphor (NIP). Seven observers marked suspicious and benign lesions. Analysis was undertaken using jackknife alternative free-response receiver operating characteristics weighted figure of merit (FoM). The cancer detection fraction (CDF) was estimated for a representative image set from screening. RESULTS: No significant differences in the FoMs between the DR detectors were measured. For calcification clusters and non-calcification lesions, both CR detectors' FoMs were significantly lower than for DR detectors. The calcification cluster's FoM for CR NIP was significantly better than for CR PIP. The estimated CDFs with CR PIP and CR NIP detectors were up to 15% and 22% lower, respectively, than for DR detectors. CONCLUSION: Cancer detection is affected by detector type, and the use of CR in mammography should be reconsidered. KEY POINTS: The type of mammography detector can affect the cancer detection rates. CR detectors performed worse than DR detectors in mammography. Needle phosphor CR performed better than powder phosphor CR. Calcification clusters detection is more sensitive to detector type than other cancers.

Development and experience of quality control methods for digital breast tomosynthesis systems.

OBJECTIVE: To develop tomosynthesis quality control (QC) test methods and use them alongside established two-dimensional (2D) QC tests to measure the performance of digital breast tomosynthesis (DBT) systems used in a comparative trial with 2D mammography. METHODS: DBT QC protocols and associated analysis were developed, incorporating adaptions of some 2D tests as well as some novel tests. The tomosynthesis tests were: mean glandular dose to the standard breast model; contrast-to-noise ratio in reconstructed focal planes; geometric distortion; artefact spread; threshold contrast detail detection in reconstructed focal planes, alignment of the X-ray beam to the reconstructed image and missed tissue; reproducibility of the tomosynthesis exposure; and homogeneity of the reconstructed focal planes. RESULTS: Summaries of results from the tomosynthesis QC tests are presented together with some 2D results for comparison. CONCLUSION: The tomosynthesis QC tests and analysis methods developed were successfully applied. The lessons learnt, which are detailed in the Discussion section, may be helpful to others embarking on DBT QC programmes. ADVANCES IN KNOWLEDGE: DBT performance test equipment and analysis methods have been developed. The experience gained has contributed to the subsequent drafting of DBT QC protocols in the UK and Europe.

A setup for the assessment of the effect of tubular confinement on the acoustic response of microbubbles.

Ultrasound contrast agents are gas filled microbubbles which produced enhanced echoes in ultrasound imaging thus allowing the acquisition of detailed information on the path of blood. It is theoretically known that the size of a vessel affects the behavior of a microbubble, which could potentially be used to discriminate different sized vessels. This information would be useful in the monitoring of neovascularization in tumor growth or treatment. However, currently it is not possible to identify the vessel diameter by any means of signal processing of microbubble echoes. In order to assess microbubble behavior when confined in tubes we compared the acoustic backscatter from biSphere™ microbubbles both free in water and flowing in 200 µm diameter tubes that are similar in size to arterioles. Experimental systems that allow the interrogation of individual microbubbles were designed and modified to allow investigation of both free microbubbles and those in tubes. Unprocessed single microbubble RF data were collected, allowing the calculation of both the fundamental and second harmonic components of the backscattered signal. Microbubbles confined in tubes had lower amplitude response compared to unconfined microbubbles. On consecutive insonations of the same microbubble, free microbubbles produced echoes above noise more often than confined microbubbles. This setup may be used to investigate microbubble behavior in a range of smaller tubes with diameters similar to capillaries thus enabling signal processing design for vessel differentiation.