Physica Medica-European Journal of Medical Physics最新文献

Quantification of iron in soft tissues through fast kV-switching dual-energy CT imaging: What calibration data are required?

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-04 DOI: 10.1016/j.ejmp.2025.104973

Kostas Perisinakis , Angeliki Ntouli , Thomas G. Maris , Apostolos Karantanas

Purpose

To provide recommendation on the type of calibration data needed for quantification of iron content in soft tissues through fast kV-switching dual-energy CT (DECT).

Methods

Tissue-specific liquid surrogates mimicking human liver, spleen, kidney and muscles with iron concentration of 0–7 mg/ml were prepared and attached circumferentially to a 16-cm polymethylmethacrylate CT phantom. Soft tissue-equivalent gel boluses were employed to create different in size and configuration phantom-vials setups. Each phantom-vials setup was subjected to fast kV-switching DECT imaging with different acquisition protocols. The virtual iron concentration (VIC) in mg/ml was determined for each vial through the iron(water) material density images. VIC-to-true iron concentration (TIC) curves were derived for four phantom-vials setups and three different acquisition protocols. The applicability of derived VIC-to-TIC calibration data was tested in ten DECT examinations from our institution’s archive.

Results

A linear relationship between TIC and VIC values was observed for all tissue-surrogates and phantom-vials setups (R² > 0.94). The VIC-to-TIC regression-lines derived for different tissues were found to differ significantly (p < 0.05). The regression-lines derived for the same tissue type, but different in size phantom-vials setups were also found to differ significantly (p < 0.05). The effects of DECT acquisition protocol and different vials positioning within the phantom-vials setup on derived regression-lines were found to be minor (p > 0.05).

Conclusions

Quantification of iron content through DECT imaging requires tissue- and patient size- specific calibration data. The presented DECT imaging-based method might be useful for monitoring iron levels in patients suspected of iron mis-regulation.

{"title":"Quantification of iron in soft tissues through fast kV-switching dual-energy CT imaging: What calibration data are required?","authors":"Kostas Perisinakis , Angeliki Ntouli , Thomas G. Maris , Apostolos Karantanas","doi":"10.1016/j.ejmp.2025.104973","DOIUrl":"10.1016/j.ejmp.2025.104973","url":null,"abstract":"<div><h3>Purpose</h3><div>To provide recommendation on the type of calibration data needed for quantification of iron content in soft tissues through fast kV-switching dual-energy CT (DECT).</div></div><div><h3>Methods</h3><div>Tissue-specific liquid surrogates mimicking human liver, spleen, kidney and muscles with iron concentration of 0–7 mg/ml were prepared and attached circumferentially to a 16-cm polymethylmethacrylate CT phantom. Soft tissue-equivalent gel boluses were employed to create different in size and configuration phantom-vials setups. Each phantom-vials setup was subjected to fast kV-switching DECT imaging with different acquisition protocols. The virtual iron concentration (VIC) in mg/ml was determined for each vial through the iron(water) material density images. VIC-to-true iron concentration (TIC) curves were derived for four phantom-vials setups and three different acquisition protocols. The applicability of derived VIC-to-TIC calibration data was tested in ten DECT examinations from our institution’s archive.</div></div><div><h3>Results</h3><div>A linear relationship between TIC and VIC values was observed for all tissue-surrogates and phantom-vials setups (<em>R<sup>2</sup></em> > 0.94). The VIC-to-TIC regression-lines derived for different tissues were found to differ significantly (<em>p</em> < 0.05). The regression-lines derived for the same tissue type, but different in size phantom-vials setups were also found to differ significantly (<em>p</em> < 0.05). The effects of DECT acquisition protocol and different vials positioning within the phantom-vials setup on derived regression-lines were found to be minor (<em>p</em> > 0.05).</div></div><div><h3>Conclusions</h3><div>Quantification of iron content through DECT imaging requires tissue- and patient size- specific calibration data. The presented DECT imaging-based method might be useful for monitoring iron levels in patients suspected of iron mis-regulation.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104973"},"PeriodicalIF":3.3,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Deep-learning synthetized 4DCT from 4DMRI of the abdominal site in carbon-ion radiotherapy

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-04 DOI: 10.1016/j.ejmp.2025.104963

Nakas Anestis , Hladchuk Maksym , Parrella Giovanni , Vai Alessandro , Molinelli Silvia , Camagni Francesca , Vitolo Viviana , Barcellini Amelia , Imparato Sara , Pella Andrea , Ciocca Mario , Orlandi Ester , Paganelli Chiara , Baroni Guido

Purpose

To investigate the feasibility of deep-learning-based synthetic 4DCT (4D-sCT) generation from 4DMRI data of abdominal patients undergoing Carbon Ion Radiotherapy (CIRT).

Material and methods

A 3-channel conditional Generative Adversarial Network (cGAN) was trained and tested on twenty-six patients, using paired T1-weighted 4DMRI and 4DCT volumes. 4D-sCT data were generated via the cGAN following a 3-channels segmentation approach (air, bone, soft tissue) in two scenarios: (a) 4DCT-based approach (i.e. segmentation relying on 4DCT) and (b) 4DMRI-based approach (i.e. manual segmentation on 4DMRI, to simulate a 4DMRI-only scenario). The network was first validated on a 4D computational phantom, where a ground truth dataset was available. Subsequently, the network was tested on 6 independent held-out-of-training patients. Generated volumes were evaluated with respect to the original 4DCT based on motion analysis, similarity metrics (e.g. Mean Absolute Error (MAE), Normalized Cross Coefficient (NCC)) and dosimetric criteria, by means of recalculating clinically optimized CIRT plans on the 4D-sCT.

Results

For the phantom, similarity metrics were in line with literature results, while dose volume histogram values were below 0.9 %. 4DCT-based patient results demonstrated an accurate representation with respect to the original 4DCT images (MAE: 50.64–51.29 HU), while 4DMRI-only-based results yielded higher values (MAE: 81.15–90.22 HU). Gamma pass rates (3 %/3mm) were ∼ 97 % for the 4DCT-based scenario, showing dosimetric consistency between the compared 4DCT and 4D-sCT dose distributions. D95% values on GTV/CTV were within clinical tolerances for the 4DMRI-only scenario.

Conclusion

Deep learning-based 4D-sCT generation shows potential to support treatment planning in abdominal tumors treated with CIRT.

{"title":"Deep-learning synthetized 4DCT from 4DMRI of the abdominal site in carbon-ion radiotherapy","authors":"Nakas Anestis , Hladchuk Maksym , Parrella Giovanni , Vai Alessandro , Molinelli Silvia , Camagni Francesca , Vitolo Viviana , Barcellini Amelia , Imparato Sara , Pella Andrea , Ciocca Mario , Orlandi Ester , Paganelli Chiara , Baroni Guido","doi":"10.1016/j.ejmp.2025.104963","DOIUrl":"10.1016/j.ejmp.2025.104963","url":null,"abstract":"<div><h3>Purpose</h3><div>To investigate the feasibility of deep-learning-based synthetic 4DCT (4D-sCT) generation from 4DMRI data of abdominal patients undergoing Carbon Ion Radiotherapy (CIRT).</div></div><div><h3>Material and methods</h3><div>A 3-channel conditional Generative Adversarial Network (cGAN) was trained and tested on twenty-six patients, using paired T1-weighted 4DMRI and 4DCT volumes. 4D-sCT data were generated via the cGAN following a 3-channels segmentation approach (air, bone, soft tissue) in two scenarios: (a) 4DCT-based approach (i.e. segmentation relying on 4DCT) and (b) 4DMRI-based approach (i.e. manual segmentation on 4DMRI, to simulate a 4DMRI-only scenario). The network was first validated on a 4D computational phantom, where a ground truth dataset was available. Subsequently, the network was tested on 6 independent held-out-of-training patients. Generated volumes were evaluated with respect to the original 4DCT based on motion analysis, similarity metrics (e.g. Mean Absolute Error (MAE), Normalized Cross Coefficient (NCC)) and dosimetric criteria, by means of recalculating clinically optimized CIRT plans on the 4D-sCT.</div></div><div><h3>Results</h3><div>For the phantom, similarity metrics were in line with literature results, while dose volume histogram values were below 0.9 %. 4DCT-based patient results demonstrated an accurate representation with respect to the original 4DCT images (MAE: 50.64–51.29 HU), while 4DMRI-only-based results yielded higher values (MAE: 81.15–90.22 HU). Gamma pass rates (3 %/3mm) were ∼ 97 % for the 4DCT-based scenario, showing dosimetric consistency between the compared 4DCT and 4D-sCT dose distributions. D95% values on GTV/CTV were within clinical tolerances for the 4DMRI-only scenario.</div></div><div><h3>Conclusion</h3><div>Deep learning-based 4D-sCT generation shows potential to support treatment planning in abdominal tumors treated with CIRT.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104963"},"PeriodicalIF":3.3,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Assessment of CT-to-physical density table for multiple image reconstruction functions with a large-bore scanner for radiotherapy treatment planning

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-04 DOI: 10.1016/j.ejmp.2025.104970

Takuro Okumura , Akito S. Koganezawa , Takeo Nakashima , Yusuke Ochi , Kento Tsubouchi , Yuji Murakami

Purpose

To evaluate the performance of the Aquilion Exceed LB computed tomography (CT) scanner for radiotherapy treatment planning, this study examined the effect of different combinations of the image reconstruction function (IRF) (AiCE and AIDR) and scan parameters on the CT-to-physical density (CT-PD) table and radiation dose in the phantom, and the effect of different object positions on CT values.

Methods

To investigate IRF’s influence on each material, we calculated CT values by varying tube current, pitch, field of view (FOV), and phantom position for each IRF, comparing them with reference values using filtered back projection (FBP). Furthermore, we evaluated changes in depth dose values due to IRF differences using a solid phantom.

Results

In the combinations of changes in IRF and scan parameters the change in CT value (

Δ H U

) of each material was within

\pm

10 HU, except for most conditions. The change in physical density (

Δ P D

) was within

\pm

0.02 g/cm³ for all combinations. For changes in phantom position,

Δ H U

was within

\pm

25 HU for changes within the scan FOV, except for Bone 200 mg/cc and 1250 mg/cc. In areas outside the scan FOV with an expanded FOV,

Δ H U

was significantly larger than within the scan FOV. Variations in depth dose for different IRFs using solid phantoms were within

\pm

0.5 %, except at material boundaries.

Conclusion

Our evaluations of the CT values and dose calculations suggested no need to change the CT-PD table, even with multiple IRFs.

{"title":"Assessment of CT-to-physical density table for multiple image reconstruction functions with a large-bore scanner for radiotherapy treatment planning","authors":"Takuro Okumura , Akito S. Koganezawa , Takeo Nakashima , Yusuke Ochi , Kento Tsubouchi , Yuji Murakami","doi":"10.1016/j.ejmp.2025.104970","DOIUrl":"10.1016/j.ejmp.2025.104970","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate the performance of the Aquilion Exceed LB computed tomography (CT) scanner for radiotherapy treatment planning, this study examined the effect of different combinations of the image reconstruction function (IRF) (AiCE and AIDR) and scan parameters on the CT-to-physical density (CT-PD) table and radiation dose in the phantom, and the effect of different object positions on CT values.</div></div><div><h3>Methods</h3><div>To investigate IRF’s influence on each material, we calculated CT values by varying tube current, pitch, field of view (FOV), and phantom position for each IRF, comparing them with reference values using filtered back projection (FBP). Furthermore, we evaluated changes in depth dose values due to IRF differences using a solid phantom.</div></div><div><h3>Results</h3><div>In the combinations of changes in IRF and scan parameters the change in CT value (<span><math><mrow><mi>Δ</mi><mi>H</mi><mi>U</mi></mrow></math></span>) of each material was within <span><math><mo>±</mo></math></span>10 HU, except for most conditions. The change in physical density (<span><math><mrow><mi>Δ</mi><mi>P</mi><mi>D</mi></mrow></math></span>) was within <span><math><mo>±</mo></math></span>0.02 g/cm<sup>3</sup> for all combinations. For changes in phantom position, <span><math><mrow><mi>Δ</mi><mi>H</mi><mi>U</mi></mrow></math></span> was within <span><math><mo>±</mo></math></span>25 HU for changes within the scan FOV, except for Bone 200 mg/cc and 1250 mg/cc. In areas outside the scan FOV with an expanded FOV, <span><math><mrow><mi>Δ</mi><mi>H</mi><mi>U</mi></mrow></math></span> was significantly larger than within the scan FOV. Variations in depth dose for different IRFs using solid phantoms were within <span><math><mo>±</mo></math></span>0.5 %, except at material boundaries.</div></div><div><h3>Conclusion</h3><div>Our evaluations of the CT values and dose calculations suggested no need to change the CT-PD table, even with multiple IRFs.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104970"},"PeriodicalIF":3.3,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FET-UNet: Merging CNN and transformer architectures for superior breast ultrasound image segmentation

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-03 DOI: 10.1016/j.ejmp.2025.104969

Huaikun Zhang , Jing Lian , Yide Ma

Purpose

Breast cancer remains a significant cause of mortality among women globally, highlighting the critical need for accurate diagnosis. Although Convolutional Neural Networks (CNNs) have shown effectiveness in segmenting breast ultrasound images, they often face challenges in capturing long-range dependencies, particularly for lesions with similar intensity distributions, irregular shapes, and blurred boundaries. To overcome these limitations, we introduce FET-UNet, a novel hybrid framework that integrates CNNs and Swin Transformers within a UNet-like architecture.

Methods

FET-UNet features parallel branches for feature extraction: one utilizes ResNet34 blocks, and the other employs Swin Transformer blocks. These branches are fused using an advanced feature aggregation module (AFAM), enabling the network to effectively combine local details and global context. Additionally, we include a multi-scale upsampling mechanism in the decoder to ensure precise segmentation outputs. This design enhances the capture of both local details and long-range dependencies.

Results

Extensive evaluations on the BUSI, UDIAT, and BLUI datasets demonstrate the superior performance of FET-UNet compared to state-of-the-art methods. The model achieves Dice coefficients of 82.9% on BUSI, 88.9% on UDIAT, and 90.1% on BLUI.

Conclusion

FET-UNet shows great potential to advance breast ultrasound image segmentation and support more precise clinical diagnoses. Further research could explore the application of this framework to other medical imaging modalities and its integration into clinical workflows.

{"title":"FET-UNet: Merging CNN and transformer architectures for superior breast ultrasound image segmentation","authors":"Huaikun Zhang , Jing Lian , Yide Ma","doi":"10.1016/j.ejmp.2025.104969","DOIUrl":"10.1016/j.ejmp.2025.104969","url":null,"abstract":"<div><h3>Purpose</h3><div>Breast cancer remains a significant cause of mortality among women globally, highlighting the critical need for accurate diagnosis. Although Convolutional Neural Networks (CNNs) have shown effectiveness in segmenting breast ultrasound images, they often face challenges in capturing long-range dependencies, particularly for lesions with similar intensity distributions, irregular shapes, and blurred boundaries. To overcome these limitations, we introduce FET-UNet, a novel hybrid framework that integrates CNNs and Swin Transformers within a UNet-like architecture.</div></div><div><h3>Methods</h3><div>FET-UNet features parallel branches for feature extraction: one utilizes ResNet34 blocks, and the other employs Swin Transformer blocks. These branches are fused using an advanced feature aggregation module (AFAM), enabling the network to effectively combine local details and global context. Additionally, we include a multi-scale upsampling mechanism in the decoder to ensure precise segmentation outputs. This design enhances the capture of both local details and long-range dependencies.</div></div><div><h3>Results</h3><div>Extensive evaluations on the BUSI, UDIAT, and BLUI datasets demonstrate the superior performance of FET-UNet compared to state-of-the-art methods. The model achieves Dice coefficients of 82.9% on BUSI, 88.9% on UDIAT, and 90.1% on BLUI.</div></div><div><h3>Conclusion</h3><div>FET-UNet shows great potential to advance breast ultrasound image segmentation and support more precise clinical diagnoses. Further research could explore the application of this framework to other medical imaging modalities and its integration into clinical workflows.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104969"},"PeriodicalIF":3.3,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761057","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Accuracy of contour propagation from planning computed tomography to iterative cone-beam computed tomography using a deformable image registration algorithm for assisting head and neck radiotherapy

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-03 DOI: 10.1016/j.ejmp.2025.104972

Hayate Washio , Seiya Murata , Yoshihiro Ueda , Yasuhiko Yamane , Koji Konishi

Background

The deteriorated image quality of cone-beam computed tomography (CBCT) reduces the accuracy of contour propagation. We investigated the accuracy of contour propagation from planning CT (pCT) to iterative CBCT (iCBCT) using deformable image registration and compared it with that of replanning CT (reCT) and Feldkamp–Davis–Kress algorithm-based CBCT (FDK-CBCT). No report exists regarding iCBCT improving the accuracy of this technique for images of the head and neck region.

Methods

We included 29 patients who underwent radiotherapy for head and neck cancer. ReCT and CBCT were performed on the same day. The gross tumor volume (GTV) and organs at risk, including the brain stem, spinal cord, mandible, parotid glands, submandibular glands, and larynx, were manually contoured by radiation oncologists on pCT and reCT images. Contour propagation was performed using MIM software. Manually delineated contours on reCT images and deformably generated contours on reCT, FDK-CBCT, and iCBCT images were compared to determine the accuracy of contour propagation using the Dice similarity coefficient (DSC), mean distance to agreement (MDA), and Hausdorff distance (HD).

Results

The mean DSC values for all contoured organs were 0.84 across reCT, FDK-CBCT and iCBCT. A mean DSC value of >0.8 was observed for all organs, except for the larynx and the GTV. The MDA was <1.5 mm for all organs and images, whereas the HD value showed a variation of >3 mm.

Conclusion

The results demonstrated no statistically significant difference in contour propagation from pCT to iCBCT compared to pCT to reCT.

{"title":"Accuracy of contour propagation from planning computed tomography to iterative cone-beam computed tomography using a deformable image registration algorithm for assisting head and neck radiotherapy","authors":"Hayate Washio , Seiya Murata , Yoshihiro Ueda , Yasuhiko Yamane , Koji Konishi","doi":"10.1016/j.ejmp.2025.104972","DOIUrl":"10.1016/j.ejmp.2025.104972","url":null,"abstract":"<div><h3>Background</h3><div>The deteriorated image quality of cone-beam computed tomography (CBCT) reduces the accuracy of contour propagation. We investigated the accuracy of contour propagation from planning CT (pCT) to iterative CBCT (iCBCT) using deformable image registration and compared it with that of replanning CT (reCT) and Feldkamp–Davis–Kress algorithm-based CBCT (FDK-CBCT). No report exists regarding iCBCT improving the accuracy of this technique for images of the head and neck region.</div></div><div><h3>Methods</h3><div>We included 29 patients who underwent radiotherapy for head and neck cancer. ReCT and CBCT were performed on the same day. The gross tumor volume (GTV) and organs at risk, including the brain stem, spinal cord, mandible, parotid glands, submandibular glands, and larynx, were manually contoured by radiation oncologists on pCT and reCT images. Contour propagation was performed using MIM software. Manually delineated contours on reCT images and deformably generated contours on reCT, FDK-CBCT, and iCBCT images were compared to determine the accuracy of contour propagation using the Dice similarity coefficient (DSC), mean distance to agreement (MDA), and Hausdorff distance (HD).</div></div><div><h3>Results</h3><div>The mean DSC values for all contoured organs were 0.84 across reCT, FDK-CBCT and iCBCT. A mean DSC value of >0.8 was observed for all organs, except for the larynx and the GTV. The MDA was <1.5 mm for all organs and images, whereas the HD value showed a variation of >3 mm.</div></div><div><h3>Conclusion</h3><div>The results demonstrated no statistically significant difference in contour propagation from pCT to iCBCT compared to pCT to reCT.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104972"},"PeriodicalIF":3.3,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Experimental investigation of discrete range modulation proton radiography with a focus on edge-blurring improvement

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-03 DOI: 10.1016/j.ejmp.2025.104961

I-Chun Cho , Yu-Hsin Cheng , Yi-Ting Liu , Meng-Wei Ho , Yu-Ming Wang , Tsukasa Aso , Sheng-Wen Hsiao , Yi-Chun Lin , Kang-Hsing Fan , Tsi-Chian Chao

Purpose

Proton radiography’s reliability in range verification is hindered by image quality due to multiple Coulomb scattering. This study evaluated the feasibility of using discrete range modulation (DRM) proton radiography for proton range estimation with Monte Carlo simulations and experimental approaches.

Methods

The PTSim Monte Carlo code analyzed the image resolution of the DRM method for various checkerboard phantom configurations and identified the source of edge-blurring using the geometric trigger feature. Additionally, an in-house MATLAB code was developed to deconvolute the energy dose curve and reduce edge-blurring effects in DRM images caused by multiple Coulomb scattering (MCS). In the experimental part, a CIRS phantom and a human-like Alderson Radiation Therapy phantom were used to acquire the first DRM image and apply a commercial 2D detector under clinical conditions.

Results

The simulation results from the checkerboard phantom revealed that the image resolution of the DRM image can reach 1 mm in both 5 cm and 9 cm phantom. The geometric trigger feature in the simulation helped remove the edge-blurring effect caused by MCS from the DRM image. Experiments with the CIRS phantom showed a maximum water-equivalent path length (WEPL) prediction error of approximately 1 mm for various materials. The experiment with the human-like phantom demonstrated that DRM can image complex structures, including soft tissue and skeletal regions.

Conclusions

In conclusion, the DRM method showed potential for clinical use, producing high-quality images, providing accurate WEPL prediction, correcting edge-blurring caused by MCS, and imaging complex structures.

{"title":"Experimental investigation of discrete range modulation proton radiography with a focus on edge-blurring improvement","authors":"I-Chun Cho , Yu-Hsin Cheng , Yi-Ting Liu , Meng-Wei Ho , Yu-Ming Wang , Tsukasa Aso , Sheng-Wen Hsiao , Yi-Chun Lin , Kang-Hsing Fan , Tsi-Chian Chao","doi":"10.1016/j.ejmp.2025.104961","DOIUrl":"10.1016/j.ejmp.2025.104961","url":null,"abstract":"<div><h3>Purpose</h3><div>Proton radiography’s reliability in range verification is hindered by image quality due to multiple Coulomb scattering. This study evaluated the feasibility of using discrete range modulation (DRM) proton radiography for proton range estimation with Monte Carlo simulations and experimental approaches.</div></div><div><h3>Methods</h3><div>The PTSim Monte Carlo code analyzed the image resolution of the DRM method for various checkerboard phantom configurations and identified the source of edge-blurring using the geometric trigger feature. Additionally, an in-house MATLAB code was developed to deconvolute the energy dose curve and reduce edge-blurring effects in DRM images caused by multiple Coulomb scattering (MCS). In the experimental part, a CIRS phantom and a human-like Alderson Radiation Therapy phantom were used to acquire the first DRM image and apply a commercial 2D detector under clinical conditions.</div></div><div><h3>Results</h3><div>The simulation results from the checkerboard phantom revealed that the image resolution of the DRM image can reach 1 mm in both 5 cm and 9 cm phantom. The geometric trigger feature in the simulation helped remove the edge-blurring effect caused by MCS from the DRM image. Experiments with the CIRS phantom showed a maximum water-equivalent path length (WEPL) prediction error of approximately 1 mm for various materials. The experiment with the human-like phantom demonstrated that DRM can image complex structures, including soft tissue and skeletal regions.</div></div><div><h3>Conclusions</h3><div>In conclusion, the DRM method showed potential for clinical use, producing high-quality images, providing accurate WEPL prediction, correcting edge-blurring caused by MCS, and imaging complex structures.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104961"},"PeriodicalIF":3.3,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A French survey on practices of identifying and monitoring patients undergoing interventional radiology procedures

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-02 DOI: 10.1016/j.ejmp.2025.104967

Joël Greffier , Marjorie Ferré

Purpose

To take stock of practices in France in terms of the criteria and methods for detecting and monitoring patients at-risk who have undergone interventional procedures.

Materials and methods

A questionnaire was sent to all members of the Imaging-Section at the SFPM. The questions concerned the type and value of alert thresholds reached during interventional procedures requiring the expertise of a medical physicist, thresholds leading to systematic patient follow-up, established follow-up procedures and also organizational questions. Data were collected for 4 interventional specialties, independently.

Results

Twenty French centers participated in the study for one or more interventional specialties. Only 2 centers used different thresholds for different interventional specialties. The dosimetric indicators most often used for alert thresholds were a combination of Dose-Area-Product (DAP) plus Air-Kerma (AK; 35 %) and AK alone (30 %). For DAP, the predominant alert threshold values were 300 Gy.cm² (23 %) and 500 Gy.cm² (18 %), and 3 Gy (41 %) and 5 Gy (23 %) for AK. The most commonly-used dosimetric indicator for patient follow-up was peak skin dose (80 %) alone or in combination with another dosimetric indicator. The most common follow-up threshold values were 3 Gy (50 %) and 5 Gy (33 %). Previous examinations in the same anatomical region were taken into account over periods of 2 months (30 %) and 3 months (30 %) and the most common type of follow-up was the patient consultation alone or combined with self-monitoring.

Conclusion

This national survey study showed that at-risk patients were monitored in all centers, that there were heterogeneities between centers in the way patients are identified and monitored.

{"title":"A French survey on practices of identifying and monitoring patients undergoing interventional radiology procedures","authors":"Joël Greffier , Marjorie Ferré","doi":"10.1016/j.ejmp.2025.104967","DOIUrl":"10.1016/j.ejmp.2025.104967","url":null,"abstract":"<div><h3>Purpose</h3><div>To take stock of practices in France in terms of the criteria and methods for detecting and monitoring patients at-risk who have undergone interventional procedures.</div></div><div><h3>Materials and methods</h3><div>A questionnaire was sent to all members of the Imaging-Section at the SFPM. The questions concerned the type and value of alert thresholds reached during interventional procedures requiring the expertise of a medical physicist, thresholds leading to systematic patient follow-up, established follow-up procedures and also organizational questions. Data were collected for 4 interventional specialties, independently.</div></div><div><h3>Results</h3><div>Twenty French centers participated in the study for one or more interventional specialties. Only 2 centers used different thresholds for different interventional specialties. The dosimetric indicators most often used for alert thresholds were a combination of Dose-Area-Product (DAP) plus Air-Kerma (AK; 35 %) and AK alone (30 %). For DAP, the predominant alert threshold values were 300 Gy.cm<sup>2</sup> (23 %) and 500 Gy.cm<sup>2</sup> (18 %), and 3 Gy (41 %) and 5 Gy (23 %) for AK. The most commonly-used dosimetric indicator for patient follow-up was peak skin dose (80 %) alone or in combination with another dosimetric indicator. The most common follow-up threshold values were 3 Gy (50 %) and 5 Gy (33 %). Previous examinations in the same anatomical region were taken into account over periods of 2 months (30 %) and 3 months (30 %) and the most common type of follow-up was the patient consultation alone or combined with self-monitoring.</div></div><div><h3>Conclusion</h3><div>This national survey study showed that at-risk patients were monitored in all centers, that there were heterogeneities between centers in the way patients are identified and monitored.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104967"},"PeriodicalIF":3.3,"publicationDate":"2025-04-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143747352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Feasibility study of using fast low-dose pencil beam proton and helium radiographs for intrafractional motion management

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-01 DOI: 10.1016/j.ejmp.2025.104959

Alexander A. Pryanichnikov , Jennifer J. Hardt , Ethan A. DeJongh , Lukas Martin , Don F. DeJongh , Oliver Jäkel , Niklas Wahl , Joao Seco

Purpose

This study aims to evaluate the feasibility of using fast, low-dose proton (pRad) and helium (HeRad) radiography for intrafractional motion management. This approach uses pencil ion beam delivery systems, modern particle imaging detectors and fast image reconstruction.

Methods

A plastic respiratory phantom underwent four-dimensional computed tomography (4DCT) using a commercial X-ray scanner, experimental pRad with a continuous proton beam from a clinical serial cyclotron, and experimental pRad and HeRad with pulsed proton and helium beams from a synchrotron-based ion therapy facility. Open-source patient 4DCT data were used in a Monte Carlo simulation study to evaluate pRad and HeRad in a realistic patient geometry. Treatment plans involving mixed carbon-helium beams were calculated using matRad and simulated in TOPAS.

Results

The experimental pRad achieved a temporal resolution of 8 fps for the cyclotron-based facility, while both pRad and HeRad achieved 2 fps for the synchrotron-based facility within a 10 cm × 10 cm region of interest. pRad reconstructed the respiratory phantom motion pattern with a dose of less than 2 µGy per image. In simulations of mixed carbon-helium beams, HeRad, both integral and single iso-energy, detected water equivalent thickness differences with sub-millimeter accuracy across different phases of the patient’s 4DCT data.

Conclusion

This study demonstrates that low-dose small-field proton and helium radiography, utilizing pencil beam scanning, can effectively monitor intrafractional anatomical displacements with millimeter-level spatial accuracy and sub-second temporal resolution. Current particle imaging and beam delivery technologies have the potential to enable real-time patient monitoring in promising mixed ion beam therapy.

{"title":"Feasibility study of using fast low-dose pencil beam proton and helium radiographs for intrafractional motion management","authors":"Alexander A. Pryanichnikov , Jennifer J. Hardt , Ethan A. DeJongh , Lukas Martin , Don F. DeJongh , Oliver Jäkel , Niklas Wahl , Joao Seco","doi":"10.1016/j.ejmp.2025.104959","DOIUrl":"10.1016/j.ejmp.2025.104959","url":null,"abstract":"<div><h3>Purpose</h3><div>This study aims to evaluate the feasibility of using fast, low-dose proton (pRad) and helium (HeRad) radiography for intrafractional motion management. This approach uses pencil ion beam delivery systems, modern particle imaging detectors and fast image reconstruction.</div></div><div><h3>Methods</h3><div>A plastic respiratory phantom underwent four-dimensional computed tomography (4DCT) using a commercial X-ray scanner, experimental pRad with a continuous proton beam from a clinical serial cyclotron, and experimental pRad and HeRad with pulsed proton and helium beams from a synchrotron-based ion therapy facility. Open-source patient 4DCT data were used in a Monte Carlo simulation study to evaluate pRad and HeRad in a realistic patient geometry. Treatment plans involving mixed carbon-helium beams were calculated using matRad and simulated in TOPAS.</div></div><div><h3>Results</h3><div>The experimental pRad achieved a temporal resolution of 8 fps for the cyclotron-based facility, while both pRad and HeRad achieved 2 fps for the synchrotron-based facility within a 10 cm × 10 cm region of interest. pRad reconstructed the respiratory phantom motion pattern with a dose of less than 2 µGy per image. In simulations of mixed carbon-helium beams, HeRad, both integral and single <em>iso</em>-energy, detected water equivalent thickness differences with sub-millimeter accuracy across different phases of the patient’s 4DCT data.</div></div><div><h3>Conclusion</h3><div>This study demonstrates that low-dose small-field proton and helium radiography, utilizing pencil beam scanning, can effectively monitor intrafractional anatomical displacements with millimeter-level spatial accuracy and sub-second temporal resolution. Current particle imaging and beam delivery technologies have the potential to enable real-time patient monitoring in promising mixed ion beam therapy.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104959"},"PeriodicalIF":3.3,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143747351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Clinically applicable semi-supervised learning framework for multiple organs at risk and tumor delineation in lung cancer brachytherapy

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-04-01 DOI: 10.1016/j.ejmp.2025.104968

Guobin Zhang , Daguang Zhang , Qiang Cao , Shubin Yang , Yijun Xiao , Zhenzhong Liu

Purpose

The generalization ability of deep learning-based automatic segmentation techniques for lung cancer in practical clinical applications remains under-validated. We reported an investigation that validated a robust semi-supervised conditional nnU-Net (SSC-nnUNet) model in multiple organs at risk (OARs) and tumor segmentation in lung cancer brachytherapy, also explored its potential in robot-assisted puncture diagnosis and treatment.

Materials and methods

Six hundred seventy-four patients with CT data from four partially labeled datasets were divided into training and validation sets at a ratio of 4:1, 181 patients from multiple centers (private dataset) with fully annotated data provided by 3 experienced radiation experts were used for testing comparison. Six experienced experts from multiple centers were asked to correct model-generated contours, and 8 junior oncologists were assigned to delineate contours based on model supporting. To verify the feasibility of the contouring model in robot-assisted surgical operations, an equivalent human model experiment was designed specifically for lung cancer puncture treatment.

Results

In model-based experienced expert assessment, the mean revision degree achieved a competitive score of 1.38 % by 6 multicenter experts. In model-based junior oncologist assessment, they acquired a mean revision degree and efficiency improvement of −1.82 % and 83.4 %, respectively. Guided by the segmentation results of OARs and tumors, an average puncture error of 0.78 mm was achieved across 10 puncture experiments.

Conclusion

The SSC-nnUNet model showed a significant improvement in the segmentation quality and efficiency especially in junior oncologist delineation. Specifically, robot-assisted experiments illustrated that the model has great application potential in clinical treatment.

{"title":"Clinically applicable semi-supervised learning framework for multiple organs at risk and tumor delineation in lung cancer brachytherapy","authors":"Guobin Zhang , Daguang Zhang , Qiang Cao , Shubin Yang , Yijun Xiao , Zhenzhong Liu","doi":"10.1016/j.ejmp.2025.104968","DOIUrl":"10.1016/j.ejmp.2025.104968","url":null,"abstract":"<div><h3>Purpose</h3><div>The generalization ability of deep learning-based automatic segmentation techniques for lung cancer in practical clinical applications remains under-validated. We reported an investigation that validated a robust semi-supervised conditional nnU-Net (SSC-nnUNet) model in multiple organs at risk (OARs) and tumor segmentation in lung cancer brachytherapy, also explored its potential in robot-assisted puncture diagnosis and treatment.</div></div><div><h3>Materials and methods</h3><div>Six hundred seventy-four patients with CT data from four partially labeled datasets were divided into training and validation sets at a ratio of 4:1, 181 patients from multiple centers (private dataset) with fully annotated data provided by 3 experienced radiation experts were used for testing comparison. Six experienced experts from multiple centers were asked to correct model-generated contours, and 8 junior oncologists were assigned to delineate contours based on model supporting. To verify the feasibility of the contouring model in robot-assisted surgical operations, an equivalent human model experiment was designed specifically for lung cancer puncture treatment.</div></div><div><h3>Results</h3><div>In model-based experienced expert assessment, the mean revision degree achieved a competitive score of 1.38 % by 6 multicenter experts. In model-based junior oncologist assessment, they acquired a mean revision degree and efficiency improvement of −1.82 % and 83.4 %, respectively. Guided by the segmentation results of OARs and tumors, an average puncture error of 0.78 mm was achieved across 10 puncture experiments.</div></div><div><h3>Conclusion</h3><div>The SSC-nnUNet model showed a significant improvement in the segmentation quality and efficiency especially in junior oncologist delineation. Specifically, robot-assisted experiments illustrated that the model has great application potential in clinical treatment.</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104968"},"PeriodicalIF":3.3,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143738335","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Plan complexity and dosiomics signatures for gamma passing rate classification in volumetric modulated arc therapy: External validation across different LINACs

IF 3.3 3区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Physica Medica-European Journal of Medical Physics

Pub Date : 2025-03-29 DOI: 10.1016/j.ejmp.2025.104962

Chao Li , Zhuo Su , Bing Li , Wenzheng Sun , Dang Wu , Yizhe Zhang , Xia Li , Zejun Xie , Jing Huang , Qichun Wei

Purpose

This study aims to enhance gamma passing rate (GPR) classification by integrating plan complexity signature, dosiomics signature, and comprehensive plan parameters, and to validate this method using data from different linear accelerators (LINACs).

Methods

This study included 235 volumetric modulated arc therapy (VMAT) treatment plans delivered using the TrueBeam LINAC as the primary dataset, along with 47 plans from the VitalBeam LINAC for external validation. The primary dataset was split into training (N = 166) and test (N = 69) subsets. Extracted features included 47 plan complexity metrics, 851 dosiomics features, and 20 plan parameters. Plan complexity score (PCscore) and dosiomics score (Doscore) were derived using the least absolute shrinkage and selection operator (LASSO). Four classification models were developed by combining PCscore, Doscore, and plan parameters according to a gamma criterion of 2 %/2 mm (γ2%/2 mm). A nomogram was constructed to combine these signatures with plan parameters. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

Results

The combined model incorporating PCscore, Doscore, and plan parameters exhibited high discriminative power, with areas under the curve (AUC) of 0.894, 0.899, and 0.904 for the training, test, and external datasets, respectively. At γ3%/2 mm, the model maintained robust performance with AUCs of 0.842 and 0.833 in the test and external datasets. Calibration curves and DCA validated the model’s effectiveness.

Conclusions

Integrating plan complexity and dosiomics signatures with key plan parameters significantly improves GPR classification for VMAT treatment plans, offering a robust approach for patient-specific quality assurance (PSQA).

{"title":"Plan complexity and dosiomics signatures for gamma passing rate classification in volumetric modulated arc therapy: External validation across different LINACs","authors":"Chao Li , Zhuo Su , Bing Li , Wenzheng Sun , Dang Wu , Yizhe Zhang , Xia Li , Zejun Xie , Jing Huang , Qichun Wei","doi":"10.1016/j.ejmp.2025.104962","DOIUrl":"10.1016/j.ejmp.2025.104962","url":null,"abstract":"<div><h3>Purpose</h3><div>This study aims to enhance gamma passing rate (GPR) classification by integrating plan complexity signature, dosiomics signature, and comprehensive plan parameters, and to validate this method using data from different linear accelerators (LINACs).</div></div><div><h3>Methods</h3><div>This study included 235 volumetric modulated arc therapy (VMAT) treatment plans delivered using the TrueBeam LINAC as the primary dataset, along with 47 plans from the VitalBeam LINAC for external validation. The primary dataset was split into training (N = 166) and test (N = 69) subsets. Extracted features included 47 plan complexity metrics, 851 dosiomics features, and 20 plan parameters. Plan complexity score (PCscore) and dosiomics score (Doscore) were derived using the least absolute shrinkage and selection operator (LASSO). Four classification models were developed by combining PCscore, Doscore, and plan parameters according to a gamma criterion of 2 %/2 mm (γ2%/2 mm). A nomogram was constructed to combine these signatures with plan parameters. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).</div></div><div><h3>Results</h3><div>The combined model incorporating PCscore, Doscore, and plan parameters exhibited high discriminative power, with areas under the curve (AUC) of 0.894, 0.899, and 0.904 for the training, test, and external datasets, respectively. At γ3%/2 mm, the model maintained robust performance with AUCs of 0.842 and 0.833 in the test and external datasets. Calibration curves and DCA validated the model’s effectiveness.</div></div><div><h3>Conclusions</h3><div>Integrating plan complexity and dosiomics signatures with key plan parameters significantly improves GPR classification for VMAT treatment plans, offering a robust approach for patient-specific quality assurance (PSQA).</div></div>","PeriodicalId":56092,"journal":{"name":"Physica Medica-European Journal of Medical Physics","volume":"133 ","pages":"Article 104962"},"PeriodicalIF":3.3,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143725505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0