SOMATOM go.Open Pro Intelligent 4DCT Imaging White Paper

White paper Intelligent 4DCT imaging Adapting to the patient’s breathing in real time with Direct i4D siemens-healthineers.com/somatom-go-open-pro Michaela Hoesl, PhD Christian Hofmann, PhD Alejandra Pecka, PhD Jannis Dickmann, PhD SIEMENS Healthineers White paper · Intelligent 4DCT imaging Overview The Direct i4D solution For 75% of the time, patients breathe irregularly which can lead to artifacts. 4DCT imaging is the basis for the subsequent tumor contouring needed for treatment planning. Adjusting the target margin (23% of cases) or rescanning the patient (6% of cases) is the approach taken in conventional 4DCT imaging when artifacts obscure the target. Especially for high-precision treatments, such as SBRT, minimal target margins are crucial to limit toxicity and deliver the intended dose. Key features of Direct i4D1 Direct i4D versus Direct i4D was developed to address the issue of irregular conventional 4DCT evaluation2 breathing during scanning and constitutes the world’s Evaluation of Direct i4D by Werner et al. and Skitzak et al. first 4DCT sequence that intelligently adapts to the showed that scoring of image artifacts was significantly patient’s breathing pattern in real time. Direct i4D is a improved for Direct i4D compared to conventional 4DCT sequential 4DCT scan mode consisting of three steps: (p-values < 0.001). 1. Optimal selection of starting parameters based on • Scores were 74% (amplitude-based reconstruction) and the individual patient, 53% (phase-based reconstruction) with artifact-free, 2. Sequential acquisition of a full breathing cycle at versus 13% and 5% for conventional 4DCT, respectively. each couch position using an online gating device • For conventional 4DCT, rescanning was necessary for respiratory information in real time, in 31% (amplitude-based reconstruction) and 37% 3. Retrospective optimal reconstruction based on (phase-based reconstruction), in contrast to only 4% phase or amplitude-based binning. (PB) and 1% (AB) for Direct i4D. • No significant difference in beam-on time was found. 75% of the time, • In this study, acquisition time increased by up to 53% patients breathe irregularly. with Direct i4D, but image quality was improved and the need for rescanning minimized. Courtesy of University Medical Center Hamburg-Eppendorf (UKE), Germany, and Central Alabama Radiation Oncology, Montgomery, USA 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 2 The statements by Siemens Healthineers’ customers described herein are based on results that were achieved in the customer’s unique setting. Because there is no “typical” hospital or laboratory and many variables exist (e.g., hospital size, samples mix, case mix, level of IT and/or automation adoption) there can be no guarantee that other customers will achieve the same results. 2 Intelligent 4DCT imaging · White paper Contents Motion management and 4DCT imaging in radiotherapy 4 Intelligent 4DCT imaging with Direct i4D 6 Technical and clinical evaluation 10 Clinical cases and customer feedback 14 Conclusion 16 3 White paper · Intelligent 4DCT imaging Motion management and 4DCT imaging in radiotherapy Management of respiratory motion in radiation oncology The goal of SBRT is to deliver the prescription dose in a is crucial in modern radiation therapy (RT) techniques. single fraction to increase the biologically effective dose Guidelines were already developed in 2006 by the AAPM (BED). For non-small cell lung cancer (NSCLC) in the early task group 76 and task group 324 is currently updating stages, another treatment strategy is to apply the high the guidelines, showing the ongoing importance of dose in up to five treatment fractions (hypofractionated motion management in RT [1]. treatment)1. Cancers in the chest and abdomen of the human body In hypofractionated or stereotactic treatments, healthy are most affected by respiratory motion, with lung cancer tissue sparing becomes imperative to reduce late the most obvious case. Lung cancer ranks second in terms toxicities as much as possible and, at the same time, of new incidences, with about two million new cases per deliver the curative dose to the target. A clear depiction year worldwide, and suffers from consistently low of the tumor and surrounding organs in motion is average five-year survival rates of < 20% (IARC, World important to achieve the goals of simultaneous healthy Health Organization, 2020). For lung cancer, SBRT has tissue sparing, while treating the tumor, and is crucial for become a common treatment choice, either prior to deciding on the strategy of treatment. While breath-hold surgery or as standalone treatment as an alternative to techniques have been proposed and implemented, they surgery [2–4]. Stereotactic body radiation therapy (SBRT) are not easily tolerated by severely ill patients. Therefore, or stereotactic ablation radiotherapy (SABR) is a high- 4D imaging without the need for breath-hold plays a key precision treatment, e.g., for lung and liver tumors. role in modern RT treatment planning for moving tumors. LALAL Eclipse Doses 27.09.2021 (Respiratory Motion J.00 Sm40 $3 0% Cheat phase mDose cale) 45,0 40.0 20 0 20.0. 150 100 20.0 100 Figure 1: High image quality is crucial to target the tumor and limit toxicity, especially in SBRT treatments. Courtesy of Central Alabama Radiation Oncology, Montgomery, USA https://www.cancer.org/cancer/lung-cancer/treating-non-small-cell/radiation-therapy.html 4 Intelligent 4DCT imaging · White paper Motion management with 4DCT imaging 4DCT imaging gives access to the necessary information for phase-based binning of conventional 4DCT does about the tumor location during the breathing cycle and not adequately match the breathing phases. In addition, plays a crucial role in the preparation and selection of reconstruction of conventional 4DCT does not allow for treatment. For precise contouring, high image quality outliers in breathing amplitudes, which can also cause without large tumor-obscuring artifacts is important. artifacts in the image. Conventional 4DCT images from spiral or cine 4DCT Artifacts can negatively influence the contouring of sequences can suffer from numerous artifacts, especially targets and organs at risk and hence hinder optimal when patients cannot breathe regularly. Wulfhekel et dose calculation and delivery by increasing target al.[5] conducted a retrospective analysis with more than margins. In extreme cases, when images are inadequate 50 patients and found that 75% of the images had for treatment decision, rescanning the patient is artifacts. A fixed respiration frequency is usually estimated necessary without guarantee that the second scan will by the therapist and manually entered prior to the 4DCT be less artifact corrupted. Antony et al.[6] showed that acquisition. The choice of adequate starting parameters in 23% of cases, the medical physicist advised on either depends heavily on the experience of the staff. margin adaptation (17%) or even rescanning (6%) due Artifacts then arise from inappropriately fixed pitch to extensive motion artifacts, neither of which were factors or a predefined fixed time at each couch adequate for subsequent SBRT treatment. position which do not consider irregular breathing. The Especially for hypofractionated treatment schemes or consequence is missing projection data or misassignment SBRT treatments, artifacts become critical. For example, of projection data at neighboring couch positions. Sentker et al.[7] analyzed how 4DCT image quality Missing projection data can lead to interpolation (INT) correlated to clinical outcome in SBRT planning for artifacts, which often manifest over multiple slices. 62 patients with 102 lung and liver metastases. They A misassignment of projection data, for instance in found that 4DCT artifacts negatively influence outcome the case of a change in breathing rate and amplitude, in SBRT of lung and liver metastases, which underlines can lead to stacking artifacts or double structure (DS) the need to address 4DCT artifacts and improve image artifacts (Figure 2). quality. These studies support the hypothesis that high- While amplitude-based binning can handle irregular quality, artifact-reduced 4DCT imaging can improve breathing frequencies to some extent, the reconstruction planning quality and might ultimately influence clinical outcome, especially in SBRT treatments. Irregular breathing frequency Irregular amplitude T Figure 2: Example of artifacts in 4DCT images. Left side: coronal view of an interpolation artifact for an irregular breathing frequency. Right side: stack artifact in sagittal view with breathing curve showing variation in breathing amplitude. Courtesy of University Medical Center Hamburg-Eppendorf (UKE), Germany 5 White paper · Intelligent 4DCT imaging Intelligent 4DCT imaging with Direct i4D1 Siemens Healthineers has developed an intelligent Step 2: Direct i4D intelligent scanning using online 4DCT scanning method called Direct i4D in a clinical respiratory signal information collaboration to solve the problem of 4DCT image artifacts caused by irregular breathing patterns. The Direct i4D acquires the data sequentially, one couch goal was a robust technique that adapts to the patients position after the other, and, at each couch position, breathing in real time, while keeping overall scan time adapts the scan duration to the patient’s breathing pattern within acceptable limits. in real time. An online gating system (either Anzai or The concept and first performance evaluation were RGSC) provides breathing signal information in real time. presented by Werner et al.[8],[9]. The implementation The X-ray is switched on just before the patient’s end- featured real-time signal processing to evaluate projection inspiration state using the real-time breathing signal and data coverage during scanning to reduce imaging the information from the reference cycle. Projection data artifacts and also beam-on time. Even more flexibility and breathing signal data are continuously acquired at the was introduced for the user by offering the possibility corresponding table position and analyzed in terms of of phase-based and amplitude-based postprocessing. projection data coverage. The X-ray is switched off once the predefined projection data coverage conditions have Direct i4D is a sequential 4DCT scan mode which follows been fulfilled, just after acquisition of the full breathing the principle of retrospective gating and decoupling of cycle at the end of inspiration. This intelligent acquisition projection data and breathing state definition, in contrast allows for an acquisition of an entire breathing cycle, to a prospective gating technique at predefined breathing even if this cycle varies in length. After the successful states. The Direct i4D technique consists of the following acquisition at one couch position, the couch is moved to three steps. the next position and the steps described are repeated until the desired scanning range is covered. With a scan Step 1: Automatic and intelligent prescan parameter range of up to 160 cm, you can easily accommodate lung, selection (FAST 4D) liver, and esophagus cases. This adaptive scan allows for both irregular breathing frequency and breathing After the topogram, an automatic breathing curve amplitude. The online breathing signal analysis adapts the analysis is performed directly on the CT console, called scan time and beam-on time, e.g., continuing the scan FAST 4D. It monitors the respiratory curve before the if the projection coverage has not yet been fulfilled or scan starts and automatically selects the most suitable waiting until a sufficiently good respiration cycle starts scan parameters in only 20 seconds. This step also serves (Figure 3). This analysis can be parameterized by the user to store a reference cycle for online comparison. The to either prioritize data completeness (image quality), breathing signal is saved in the form of a phase space reduced beam-on time (imaging dose), or a balanced representation for later real-time analysis during scanning. setting (trade-off between quality and dose). FAST 4D selects scan parameters Direct i4D adapts the scan duration to the Direct i4D chooses optimal data a based on breathing prior to scan b patient’s breathing pattern in real time c for 4DCT reconstruction 1 2 Figure 3: Illustration of online adaptation to the patient’s breathing. The orange bands in the breathing curve signal beam-on events. After two regular breathing cycles, the breathing amplitude changed. Direct i4D can react to this instantaneously by waiting until a sufficiently high breathing amplitude returns (1). The subsequent acquisition (2) shows what happens if the breathing amplitude or frequency changes during a beam-on event. The scan is prolonged to secure the necessary projection data coverage. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 6 Intelligent 4DCT imaging · White paper Step 3: Direct i4D1 intelligent reconstruction Retrospectively the 4D image is reconstructed with a For amplitude-based binning, Direct i4D shifts the bins high degree of freedom, e.g., choosing between phase- according to the average breathing cycle and can handle based and amplitude-based binning in steps of five irregular amplitudes such as in Figure 4b. Amplitude- (from 5 to 20). Direct i4D chooses the optimal data based reconstruction is preferred to achieve minimal for reconstruction and can handle datasets with artifacts. In the case of an atypically long breathing cycle, irregularities. For phase-based binning, Direct i4D the maximum scan duration² for one couch position may intelligently sets the reconstruction bins to match the be reached without covering a full inhalation. Direct i4D breathing phase, compared to the average breathing then estimates the missing inhalation bins with bins from cycle, as closely as possible (Figure 4a). The phase-based exhalation at equivalent phase or amplitude and at the binning is preferred if time-related information needs same couch position (Figure 4c). This can potentially to be preserved, e.g., for mid-ventilation information. avoid image artifacts due to the otherwise missing data. a Phase-based binning b Amplitude-based binning 0% 50% 100% Breathing cycle Amplitude Amplitude Time Time Figure 4: Illustration of the intelligent reconstruction for phase-based binning (a) and amplitude-based binning (b). Both reconstructions are supported. (a) The blue bin illustrates the 75% inhale phase for phase-based binning with irregular breathing frequency with Direct i4D. (b) The illustration shows the reconstruction bins shifting to match the amplitude of the average breathing cycle. (c) Illustration of an atypically long breathing cycle where the maximum scan time was reached before inhalation. Missing inhalation bins (dashed petrol) are estimated by bins of the exhalation phase at equivalent phase or amplitude (solid petrol). Supported both for phase-based and amplitude- based binning. c Special case: bin estimation for atypically long cycles Max. scan time Amplitude Time 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. ² Due to dose constraints, the acquisition at one couch position of the sequential scan is interrupted after a maximum time is reached and not repeated. The scan then continues at the next z-position. 7 White paper · Intelligent 4DCT imaging Comparison to conventional 4DCT techniques In Figure 5, Direct i4D1 is illustrated compared to conventional 4DCT techniques, highlighting the adaptive acquisition, which manifests in: 1. Variable beam-on time for each couch position, depending on the breathing frequency, 2. Beam-off times for when there is no acceptable data for artifact-free reconstruction, and 3. Optimal binning for reconstruction, here: in the example of amplitude-based binning. Direct i4D Amplitude-based binning example Conventional 4DCT Phase-based binning example Spiral 4DCT Fixed wait time Fixed scan time Ciné 4DCT Figure 5: Comparison of conventional spiral 4DCT and conventional step and shoot 4DCT with Direct i4D in terms of acquisition and post- processing. For conventional cine 4DCT, the scan duration at every position is fixed (e.g., 5 s = 12 BPM). If the breathing rate drops during the scan to, e.g., 8 BPM (= 7.5 s cycle length), artifacts arise from the acquisition of an incomplete breathing cycle. On the other hand, conventional spiral continuously acquires data. For postprocessing, the conventional 4DCT scan shows phase-based binning, while in this example, amplitude-based binning of Direct i4D can adapt to the change in amplitude. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 8 Intelligent 4DCT imaging · White paper Direct i4D1 scan time estimate The Direct i4D scan time depends on the patient, their the number of couch positions times the time for one breathing frequency and their irregularities, which lead breathing cycle and adding the table movement and to breaks in the beam-on time and longer acquisitions at wait time between. certain couch positions to acquire the optimal data. To get a rough estimate of scan time, let us go through an example with a patient with a regular breathing rate of Exemplary scan time estimation: 12 breaths per minute (BPM) with a desired coverage of N × (time for one full breathing cycle) + 38 cm. The breathing rate equals 5 seconds per breathing (N− 1) × (table movement + wait time) = cycle from inhale to inhale. SOMATOM go.Open Pro 10 × (5 s) + 9 × (1 s + 4 s) = 95 s features a detector size of 64 lines by 0.6 cm. The 38 cm coverage can be acquired in 9.9 rounds, which we will approximate with ten full rounds here. The scan time, For patients with strong breathing irregularities, without the patient-specific additional time to account Direct i4D leads by design to a longer scan and for breathing irregularities, can be estimated by taking beam-on time to account for the irregularities. Breathing phase 0% 100% t1: scanning of one breathing cycle, slices covering 3.84 cm (e.g., 5 s for 12 BPM) 50% t2: table movement to next position (~ 1 s) t3: wait time till next breathing cycle X X x (e.g., 4 s, depending on respiration) t1 t2 t3 Figure 6: Scan time estimate illustration for regular breathing of 12 BPM. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 9 White paper · Intelligent 4DCT imaging Technical and clinical evaluation Technical evaluation: First implementation Technical evaluation: Comprehensive Werner et al.[8] introduced the concept of Direct i4D1, phantom validation study and comparison which was rated Best of Physics at the ASTRO 2018 with conventional spiral 4DCT Annual Meeting. A proof-of-concept evaluation with an emphasis on the advantages of Direct i4D compared to Werner et al.[10] recently performed an in-depth conventional cine 4DCT and conventional spiral 4DCT, evaluation and phantom validation study on the by simulation and prototype implementation, was SOMATOM go.Open Pro together with the Varian RGSC presented. gating camera and the Direct i4D scan protocol using the fast gantry rotation time of 0.35 s. A comparison In the concept evaluation, breathing data for 189 patients between conventional spiral 4DCT and Direct i4D at the were analyzed and used for simulation and comparison same scanner in a real-world scenario was performed of Direct i4D, conventional cine 4DCT, and conventional using a CIRS motion phantom, including a dedicated 3D spiral 4DCT. The Direct i4D prototype implementation printed insert. The motion phantom was programmed showed significant projection data coverage, which led with 28 real breathing curves of lung and liver patients to a reduction in coverage failures of 89% (between to cover typical patient breathing patterns. 76% and 82%) compared to spiral scanning (cine 4DCT). Further, in 70 out of the 189 patients, breathing cycles Corresponding measurements were performed using were observed that deviated by more than two seconds Direct i4D and conventional spiral 4DCT with similar compared to their median breathing cycle, highlighting imaging parameters, a pitch factor of 0.07 for the importance of an adaptive method for 4DCT scanning conventional 4DCT, and the fast gantry rotation time as opposed fixed beam on times per couch position of 0.35 s, to enable a direct comparison. Images were in conventional 4DCT. Subsequently Werner et al.[9] reconstructed with ten phases and an isotropic in-plane demonstrated the feasibility of Direct i4D online resolution of 0.64 mm and a slice thickness of 1.5 mm. respiratory-guided 4DCT scanning using a first prototype implementation and phantom study, which confirmed Qualitative, blinded clinical expert rating by the reduction of artifacts due to breathing irregularity Skitzak et al.2 presented in the concept and performance evaluation Images were evaluated based on a blinded clinical (Figure 7). expert rating, consisting of two subgroups with medical physicists and physicians. The image quality was scored from one (bad image quality, need to rescan) to five (artifact-free image). The clinical expert rating showed Direct i4D Conventional 4D 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Very strong artifacts No artifacts Need to rescan 1 2 3 4 5 Artifact-free Figure 7: First evaluation and comparison of conventional 4DCT with Direct i4D, evaluated by Werner, R et al. Intelligent 4DCT Sequence Scanning (i4DCT). Best of Physics, ASTRO 2018, Med. Phys. 46 (8) June 2019. Courtesy of University Medical Center Hamburg-Eppendorf (UKE), Germany 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 2 The statements by Siemens Healthineers’ customers described herein are based on results that were achieved in the customer’s unique setting. Because there is no “typical” hospital or laboratory and many variables exist (e.g., hospital size, samples mix, case mix, level of IT and/or automation adoption) there can be no guarantee that other customers will achieve the same results. 10 Intelligent 4DCT imaging · White paper a significant improvement of Direct i4D1 over Quantitative artifact analysis conventional 4DCT (p-values <0.001). For Direct i4D, overall, 74% (amplitude-based (AB) reconstruction) For a further independent and quantitative evaluation, and 53% (phase-based (PB) reconstruction) were rated the presence and strength of artifacts was evaluated with 4 or 5 (good, artifact-free). In contrast, for automatically. The quantitative analysis found that conventional spiral 4DCT, the rating was 13% (AB) artifact-affected slices were significantly reduced from and 5% (PB). A score of 1 was given for cases that 20% for conventional 4DCT, to 4% for Direct i4D needed rescanning due to loss of important 4D imaging (p < 0.001)(Figure 7). information for clinical decision-making. In this study, the need to rescan was 37% (PB) and 31% (AB) for Beam-on time evaluation conventional spiral 4DCT, but only 4% (PB) and 1% (AB) Further, the beam-on time and scan time of for Direct i4D. The expert ratings were consistent conventional spiral 4DCT and Direct i4D were compared. between the two subgroups with physicians achieving No significant difference in beam-on time was found a score of 1 and 2 for Direct i4D slightly more frequently for the measurements conducted. More specifically, (physicians: Score 1/2 AB: 9%, PB: 14%; medical physicists: compared to Direct i4D, the median beam-on time for Score 1/2 AB: 1% and PB: 9%). spiral 4DCT showed a maximum of + 34% and - 39% for Overall, Direct i4D achieved consistently better ratings individual scans. Overall, the median beam-on time compared to conventional 4DCT. Also, overall, amplitude- was 3% shorter for Direct i4D than for conventional based reconstruction was rated higher quality than spiral 4DCT. The differences were not significant. phase-based reconstruction. Figure 8 highlights the The scan time was measured as increasing by an results of the physicians subgroup. average of 53% for Direct i4D. However, this increase in time is directly related to the additional beam-off time between the couch positions and more importantly to the adjustment to breathing irregularities, which can avoid the need for rescanning. Blinded clinical expert rating Slides affected by artifacts (AB) Direct 30% i4D (PB) 20% 20% (AB) Conv. 10% 4DCT (PB) 4% 0% 20% 40% 60% 80% 100% (AB) (PB) (AB) (PB) Need to rescan 1 2 3 4 5 Artifact-free Direct i4D Conv. 4DCT Phase-based (PB), amplitude-based (AB) binning Figure 8: Comparison of Direct i4D with conventional 4DCT phantom evaluation, as evaluated by Werner et al.[10]. Left: Blinded clinical expert rating of the physician subgroup showing the higher rating of Direct i4D. Right: Double structure and interpolation artifact-affected slices were analyzed automatically and showed a large difference of 20% for conventional 4DCT compared with only 4% for Direct i4D. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 11 White paper · Intelligent 4DCT imaging Clinical evaluation Image quality A direct clinical comparison of Direct i4D1 with Strong artifacts, which are usually caused by breathing conventional 4DCT techniques is difficult. Due to the variations, were not observed with Direct i4D. Due to ALARA principle, the radiation protection guideline, its design it could detect outliers in the breathing curve, according to which doses should be kept as low as such as variations in amplitude, frequency, or both, reasonably achievable, does not allow for unnecessary as shown in Figure 8. Even in cases where there was scanning of the same patient twice. At the same time, a combination of breathing frequency and amplitude it would be difficult to reproduce the exact same level variation, no incomplete data or artifacts were detected. of breathing irregularity even with the same patient. The overall mean projection data coverage was 92% ± 8% Szkitsak et al.[11] therefore presented the first clinical (median 93%) for inhalation and 93% ± 7% (median 94%) evaluation of Direct i4D, evaluating only Direct i4D image for exhalation. For the challenging cases, projection quality in a patient cohort of 129 patients with thoracic coverage remained high but showed a small decrease tumors. A subgroup of 30 challenging patients was (p = 0.02) to 89% ± 9% for inhalation (median 90%) and identified based on expert analysis of the breathing 90% ± 7% (exhalation) median 93%. curves by two physicists specialized in 4DCT artifacts. This subgroup presented irregularities, e.g., in the form of breathing pauses, frequency or amplitude variations, and baseline drifts. a Variation in amplitude b Variation in frequency c Combination of a) and b) Figure 9: Breathing irregularities as shown in Szkitsak et al.[11]. Direct i4D beam-on is visualized and shows the adaptation to each of the variations in breathing frequency and amplitude. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 12 Intelligent 4DCT imaging · White paper Expert rating Beam-on time In Figure 10, the expert rating study from Szkitsak et Comparing the entire cohort with the challenging al.[11] is presented. No image received a score of 1, subgroup, no significant difference in beam-on time indicating the need to rescan. For amplitude-based was observed (p). Overall, the average value per reconstruction, 78% of the images were artifact free couch position was 4.9 ± 1.6 s (median 4.5 s). For large (rating 4 and 5), for phase-based reconstruction 63% variations in the breathing curves, the time increased of the images were artifact free, despite strong breathing to 5.1 ± 1.7 s, (median 4.6 s). For the remaining patients, irregularities. Strong artifacts leading to a low rating the differences were not significant (p = 0.64). of 2 were observed in 2% (AB) and 9% (PB) of the study. On average, physicians gave a lower rating compared to physicists (Ø AB 4.3 vs. 4.0; PB 4.0 vs. 3.7, p < 0.001). In keeping with previous studies, AB resulted in significantly higher image quality compared to PB (Ø AB: 4.1, PB: 3.8, p < 0.001). Cohort with challenging subgroup Direct i4D clinical expert rating MD, (PB) Phys., (PB) MD, (AB) Phys., (AB) Regular breathing 0% 20% 40% 60% 80% 100% Need to rescan 1 2 3 4 5 Artifact free Irregular breathing Phase-based (PB), amplitude-based (AB) binning 78% (AB) vs. 63% (PB) were artifact free. Medical doctor (MD), physicist (phys.) Figure 10: Result of clinical expert rating of patient cohort with 129 patients containing a subgroup of 30 challenging patients. Breathing curves from two Direct i4D1 cases can be seen as an example for a case with regular breathing and a challenging case with irregular breathing. On the right, the Direct i4D image quality rating by physicists (phys.) and physicians (abbreviated as MD) can be seen (taken from the study by Szkitsak et al.[11]). Note that no image received a rating of 1, indicating need to rescan. A total of 78% of the images reconstructed with amplitude-based binning were artifact free. Images courtesy of University Medical Center Hamburg-Eppendorf (UKE), Germany 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 13 White paper · Intelligent 4DCT imaging Clinical cases and customer feedback Case 1: Two shallow breathing cycles skipped during acquisition Direct i4D1 analyzed the breathing pattern and waited to acquire a new z-position during two shallow breathing cycles. Figure 11: 120 kV image, scan length 260 mm, slice thickness 1.5 mm, reconstruction kernel Qr40 SAFIRE3. Two shallow breathing cycles were skipped by Direct i4D. Courtesy of the Department of Radiation Oncology, Universitätsklinikum Erlangen, Germany “We are close to planning 100% of our lung cancer patients with Direct i4D. It provides excellent images with barely any artifacts. It also provides superb images in the abdominal area to assess movement of liver lesions or adrenal glands. We have found it to facilitate our workflow massively,” says Jordi Saez, PhD, Hospital Clínic de Barcelona, Spain. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 14 Intelligent 4DCT imaging · White paper Case 2: Intelligent setting of reconstruction bins The intelligent setting of reconstruction bins (amplitude-based binning) resulted in artifact-free 4DCT imaging. Figure 12: 120 kV image, scan length 492 mm, slice thickness 2 mm, rotation time 0.35 s, scan time 66 s, CTDIvol: 23.98 mGy, DLP 1158 mGy*cm. Courtesy of University Medical Center Groningen, Netherlands “The Direct i4D feature1 is one fundamental “Direct i4D simplifies the 4DCT workflow and change of the SOMATOM go.Open Pro. We produces excellent results, even for patients have much fewer motion artifacts and the with irregular breathing patterns. Therefore, quality of the 4DCT sequence which then it helps to determine the best possible enters treatment planning is improved,” individual dose plan for RT,” says Prof. Dr. Christoph Bert, Head of Medical Physics, Department says Peter Albeck Qvistgaard, Head of Radiography Department, of Radiation Oncology, Universitätsklinikum Erlangen, Germany. Aarhus University Hospital, Aarhus, Denmark. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 15 White paper · Intelligent 4DCT imaging Conclusion 75% of RT patients breathe irregularly, which can lead to artifacts in conventional 4D CT scans.[5] Direct i4D1 is a new 4DCT scan mode designed to support radiation therapy planning. It offers an intelligent acquisition that adapts to individual breathing patterns in real time and features an intelligent reconstruction which can significantly reduce artifacts. The implementation aims at an acceptable trade-off between scan time and image quality and relies on an external surrogate for online breathing information. Completely artifact-free images might not be produced for every patient, e.g., in the case of severe breathing irregularity. Driven by intelligence and automation, Direct i4D simplifies the acquisition of high-quality 4DCT images, a key component to support accurate tumor and OAR delineation, as well as treatments including motion management. 1 Optional. Requires Varian RGSC or Anzai as online gating device. Available for SOMATOM go.Open Pro, SOMATOM X.ceed, and SOMATOM X.cite. 16 Intelligent 4DCT imaging · White paper References [1] Keall PJ, et al. The management of respiratory [7] Sentker T, et al. 4D CT image artifacts affect motion in radiation oncology report of AAPM Task local control in SBRT of lung and liver metastases. Group 76. 2006; 33(10); 2006. Radiother. Oncol. 2020 Jul; 148: 229–234, [2] Chang JY, et al. Stereotactic ablative radiotherapy doi: 10.1016/j.radonc.2020.04.006. versus lobectomy for operable stage I non-small-cell [8] Werner R, Sentker T, Madesta F, Gauer T, and lung cancer: a pooled analysis of two randomised Hofmann C. Intelligent 4D CT sequence scanning trials. Lancet Oncol. 2015; 16(6): 630–637, (i4DCT): Concept and performance evaluation. doi: 10.1016/S1470-2045(15)70168-3. Med. Phys. 2019; 46 (8): 3462–3474, [3] Ricco A, et al. Lung metastases treated with doi: 10.1002/mp.13632. stereotactic body radiotherapy: the RSSearch® [9] Werner R, et al. Intelligent 4D CT sequence patient Registry’s experience. 2017: 4–11; scanning (i4DCT): First scanner prototype doi: 10.1186/s13014-017-0773-4. implementation and phantom measurements [4] Nguyen KNB, Hause DJ, Novak J, Monjazeb AM, of automated breathing signal-guided 4D CT. and Daly ME. Tumor Control and Toxicity after Med. Phys. 2020; 47(6): 2408–2412, SBRT for Ultracentral, Central, and Paramediastinal doi: 10.1002/mp.14106. Lung Tumors. Pract. Radiat. Oncol. 2019; 9(2): [10] Werner R, et al. Comparison of intelligent e196–e202, 2019, 4D CT sequence scanning and conventional doi: https://doi.org/10.1016/j.prro.2018.11.005. spiral 4D CT: a first comprehensive phantom [5] Wulfhekel E, Grohmann C, Gauer T, and Werner R. study. Phys. Med. Biol. 2021; 66(1): 1500, Compilation of a database for illustration and doi: 10.1088/1361-6560/abc93a. automated detection of 4DCT motion artifacts. [11] Szkitsak J, et al. First clinical evaluation of Radiother. Oncol. 2014 Jan; 111: S266, breathing controlled four-dimensional computed doi: 10.1016/S0167-8140(15)31861-2. tomography imaging. Phys. Imaging Radiat. [6] Antony R, et al. Independent review of 4DCT scans Oncol. 2021; 20: 56–61, used for SABR treatment planning, doi: https://doi.org/10.1016/j.phro.2021.09.005. doi: 10.1002/acm2.12825. 17 At Siemens Healthineers, we pioneer breakthroughs in On account of certain regional limitations of sales rights healthcare. For everyone. Everywhere. By constantly and service availability, we cannot guarantee that all bringing breakthrough innovations to market, we products included in this brochure are available through support healthcare professionals to deliver high-quality the Siemens Healthineers sales organization worldwide. care, leading to the best possible outcome for patients. Availability and packaging may vary by country and are Our portfolio, spanning from in-vitro and in-vivo subject to change without prior notice. Some/All of the diagnostics to image-guided therapy and innovative features and products described herein may not be cancer care, is crucial for clinical decision-making available in the United States. and treatment pathways. With our strengths in patient twinning, precision therapy, as well as digital, data, The information in this document contains general and artificial intelligence (AI), we are well positioned technical descriptions of specifications and options as to take on the biggest challenges in healthcare. We will well as standard and optional features that do not always continue to build on these strengths to help fight the have to be present in individual cases. world’s most threatening diseases, improving the quality Siemens Healthineers reserves the right to modify the of outcomes, and enabling access to care. design, packaging, specifications, and options described We are a team of 66,000 highly dedicated employees herein without prior notice. Please contact your local across more than 70 countries passionately pushing Siemens Healthineers sales representative for the most the boundaries of what’s possible in healthcare to help current information. improve people’s lives around the world. Note: Any technical data contained in this document may vary within defined tolerances. Original images always lose a certain amount of detail when reproduced. The statements by Siemens’ Healthineers customers described herein are based on results that were achieved in the customer’s unique setting. Because there is no “typical” hospital or laboratory and many variables exist (e.g., hospital size, samples mix, case mix, level of IT and/or automation adoption) there can be no guarantee that other customers will achieve the same results. Siemens Healthineers Headquarters Siemens Healthineers AG Siemensstr. 3 91301 Forchheim, Germany Phone: +49 9191 18-0 siemens-healthineers.com Published by Siemens Healthineers AG · online · 14515 0224 · ©Siemens Healthineers AG, 2024