Challenges and opportunities in the development and clinical implementation of artificial intelligence based synthetic computed tomography for magnetic resonance only radiotherapy

Radiotherapy and Oncology, Journal Year: 2024, Volume and Issue: 198, P. 110387 - 110387

Published: June 15, 2024

Language: Английский

---

Comprehensive dose evaluation of a Deep Learning based synthetic Computed Tomography algorithm for pelvic Magnetic Resonance-only radiotherapy

Radiotherapy and Oncology, Journal Year: 2023, Volume and Issue: 184, P. 109692 - 109692

Published: May 6, 2023

Background and PurposeMagnetic Resonance (MR)-only radiotherapy enables the use of MR without uncertainty MR–Computed Tomography (CT) registration. This requires a synthetic CT (sCT) for dose calculations, which can be facilitated by novel Zero Echo Time (ZTE) sequence where bones are visible images acquired in 65 seconds. study evaluated calculation accuracy pelvic sites ZTE-based Deep Learning sCT algorithm developed GE Healthcare.Materials MethodsZTE were 56 patients position. A 2D U-net convolutional neural network was trained using pairs deformably registered ZTE from 36 patients. In remaining 20 dosimetric assessed cylindrical dummy Planning Target Volumes (PTVs) positioned at four different central axial locations, as well clinical treatment plans (for prostate (n = 10), rectum 4) anus 6) cancers). The rigidly registered, plan recalculated doses compared mean differences gamma analysis.ResultsMean to PTV D98% ≤ 0.5% all PTVs (rigid registration). Mean pass rates 1%/1 mm 98.0 ± 0.4% (rigid) 100.0 0.0% (deformable), 96.5 0.8% 99.8 0.1%, 95.4 0.6% 99.4 prostate, respectively.ConclusionsA with high throughout pelvis has been developed. suggests is sufficiently accurate MR-only sites.

Language: Английский

---

Zero-TE MRI: principles and applications in the head and neck

British Journal of Radiology, Journal Year: 2022, Volume and Issue: 95(1136)

Published: May 26, 2022

Zero echo-time (ZTE) MRI is a novel imaging technique that utilizes ultrafast readouts to capture signal from short-T2 tissues. Additional sequence advantages include rapid times, silent scanning, and artifact resistance. A robust application of this technology cortical bone without the use ionizing radiation, thus representing viable alternative CT for both screening "one-stop-shop" MRI. Although ZTE increasingly used in musculoskeletal body imaging, neuroimaging applications have historically been limited by complex anatomy pathology. In article, we review physics including pulse options, practical limitations, image reconstruction. We then discuss optimization settings acquisition, processing, segmentation, synthetic generation, artifacts. Finally, examine clinical utility head neck with examples malformations, trauma, tumors, interventional procedures.

Language: Английский

---

Synthetic CT generation using Zero TE MR for head-and-neck radiotherapy

Radiotherapy and Oncology, Journal Year: 2025, Volume and Issue: unknown, P. 110762 - 110762

Published: Jan. 1, 2025

Language: Английский

---

Clinical Applications of Generative Artificial Intelligence in Radiology: Image Translation, Synthesis and Text Generation

Deleted Journal, Journal Year: 2024, Volume and Issue: 1(1)

Published: Jan. 1, 2024

Abstract Generative artificial intelligence (AI) has enabled tasks in radiology, including tools for improving image quality. Recently, new hotspots have emerged, such as intra- or inter-modal translation, task-specific synthesis, and text generation. Advances generative AI facilitated the move towards low-dose, cost-effective, high-quality radiological acquisition. Large language models can aid radiologists by generating professional answers facilitating patient-physician communications. However, must be aware of potential inaccuracies generated content should only use after rigorous validation their performance.

Language: Английский

---

Evaluating a radiotherapy deep learning synthetic CT algorithm for PET-MR attenuation correction in the pelvis

EJNMMI Physics, Journal Year: 2024, Volume and Issue: 11(1)

Published: Jan. 29, 2024

Abstract Background Positron emission tomography–magnetic resonance (PET-MR) attenuation correction is challenging because the MR signal does not represent tissue density and conventional sequences cannot image bone. A novel zero echo time (ZTE) sequence has been previously developed which generates from cortical bone with images acquired in 65 s. This combined a deep learning model to generate synthetic computed tomography (sCT) for MR-only radiotherapy. study aimed evaluate this algorithm PET-MR pelvis. Methods Ten patients being treated ano-rectal radiotherapy received $$^{18}$$ 18 F-FDG-PET-MR position. Attenuation maps were generated ZTE-based sCT (sCTAC) standard vendor-supplied MRAC. The planning CT scan was rigidly registered cropped gold map (CTAC). PET reconstructed using each compared uptake value (SUV) measurement, automatic thresholded gross tumour volume (GTV) delineation GTV metabolic parameter measurement. last assessed clinical equivalence CTAC two one-sided paired t tests significance level corrected multiple testing of $$p \le 0.05/7 = 0.007$$ p ≤ 0.05 / 7 = 0.007 . Equivalence margins $$\pm 3.5\%$$ ± 3.5 % used. Results Mean whole-image SUV differences −0.02% −3.0% (MRAC), larger regions (−0.5% −16.3%). There no difference GTVs, Dice similarity coefficients $$\ge 0.987$$ ≥ 0.987 However, there parameters. $${\mathrm {SUV}}_{\max}$$ SUV max $$1.0 \pm 0.8\%$$ 1.0 0.8 (± error, sCTAC) $$-4.6 0.9\%$$ - 4.6 0.9 0.7\%$$ 0.7 $$-4.3 4.3 (MRAC) {SUV}}_{\rm mean}$$ mean sCTAC statistically equivalent within margin ( 0.002$$ 0.002 ), whereas MRAC 0.88$$ 0.88 0.83$$ 0.83 ). Conclusion substantially more accurate than current methods only 40 s increase acquisition time. did impact but significantly improve accuracy measurements, clinically CTAC. suggests would enable quantitative be on scanner.

Language: Английский

---

Deep learning-based pseudo-CT synthesis from zero echo time MR sequences of the pelvis

Insights into Imaging, Journal Year: 2024, Volume and Issue: 15(1)

Published: Aug. 9, 2024

Abstract Objectives To generate pseudo-CT (pCT) images of the pelvis from zero echo time (ZTE) MR sequences and compare them to conventional CT. Methods Ninety-one patients were prospectively scanned with CT MRI including ZTE pelvis. Eleven image volumes excluded due implants severe B1 field inhomogeneity. Out 80 data sets, 60 used train update a deep learning (DL) model for pCT synthesis while remaining 20 cases selected as an evaluation cohort. assessed qualitatively quantitatively by two readers. Results Mean ratings qualitative parameters good perfect (2–3 on 4-point scale). Overall intermodality agreement between was (ICC = 0.88 (95% CI: 0.85–0.90); p < 0.001) excellent interreader agreements 0.91 0.88–0.93); 0.001). Most geometrical measurements did not show any significant difference ( > 0.05) exception transverse pelvic diameter lateral center-edge angle 0.001 0.002, respectively). Image quality tissue differentiation in similar without differences CNRs (all 0.05). Conclusions Using DL-based algorithm, it is possible synthesize sequences. The showed high bone depiction accurate compared Critical relevance statement generated allow accuracy evaluating need radiation exposure. Radiological applications are broad include assessment inflammatory degenerative disease or preoperative planning studies. Key Points pCT, based DL-reconstructed images, may be comparable true images. Overall, pCT. Geometrical Graphical

Language: Английский

---

A unified generation‐registration framework for improved MR‐based CT synthesis in proton therapy

Medical Physics, Journal Year: 2024, Volume and Issue: 51(11), P. 8302 - 8316

Published: Aug. 13, 2024

The use of magnetic resonance (MR) imaging for proton therapy treatment planning is gaining attention as a highly effective method guidance. At the core this approach generation computed tomography (CT) images from MR scans. However, critical issue in process accurately aligning and CT images, task that becomes particularly challenging frequently moving body areas, such head-and-neck. Misalignments these can result blurred synthetic (sCT) adversely affecting precision effectiveness planning.

Language: Английский

---

Utility of zero echo time (ZTE) sequence for assessing bony lesions of skull base and calvarium

Clinical Radiology, Journal Year: 2024, Volume and Issue: 79(12), P. e1504 - e1513

Published: Aug. 30, 2024

Language: Английский

---

Advancements in synthetic CT generation from MRI: A review of techniques, and trends in radiation therapy planning

Journal of Applied Clinical Medical Physics, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 26, 2024

Abstract Background Magnetic resonance imaging (MRI) and Computed tomography (CT) are crucial techniques in both diagnostic radiation therapy. MRI provides excellent soft tissue contrast but lacks the direct electron density data needed to calculate dosage. CT, on other hand, remains gold standard due its accurate information therapy planning (RTP) it exposes patients ionizing radiation. Synthetic CT (sCT) generation from has been a focused study field last few years cost effectiveness as well for objective of minimizing side‐effects using more than one modality treatment simulation. It offers significant time efficiencies, bypassing complexities co‐registration, potentially improving accuracy by registration‐related errors. In an effort navigate quickly developing precision medicine, this paper investigates recent advancements sCT techniques, particularly those machine learning (ML) deep (DL). The review highlights potential these improve efficiency use RTP patient care reducing healthcare costs. intricate web is scrutinized critically, with clinical implications technical underpinnings enhanced revealed. Purpose This aims provide overview most particular focus within RTP, emphasizing performance evaluation, applications, future research trends open challenges field. Methods A thorough search strategy was employed conduct systematic literature across major scientific databases. Focusing past decade's advancements, critically examines emerging approaches introduced 2013 2023 generating MRI, providing comprehensive analysis their methodologies, ultimately fostering further advancement highlighted contributions, identified challenges, provided successes RTP. Classifying approaches, contrasting advantages disadvantages, identifying broad were all part review's synthesis process. Results identifies various consisting atlas‐based, segmentation‐based, multi‐modal fusion, hybrid ML DL‐based techniques. These evaluated image quality, dosimetric accuracy, acceptability. They used MRI‐only treatment, adaptive radiotherapy, MR/PET attenuation correction. also diversity methodologies generation, each own limitations. Emerging incorporate integration advanced modalities including sequences like Dixon sequences, T1‐weighted (T1W), T2‐weighted (T2W), accuracy. Conclusions MRI‐based minimize negative effects acquiring modalities. reviews 2013‐2023 studies methods, aiming revolutionize outcomes. insights researchers practitioners, need standardized validation procedures collaborative efforts refine methods address anticipates continued evolution

Language: Английский

---