Journal of Medical Imaging最新文献

{"title":"Digital breast tomosynthesis system concept addressing the needs in breast cancer screening and diagnosis.","authors":"Marcus Radicke, Marcel Beister, Stephan Dwars, Joerg Freudenberger, Pilar B Garcia-Allende, Bernhard Geiger, Katrin Hall, WenMan He, Axel Hebecker, Carina Heimann, Daan Hellingman, Magdalena Herbst, Mathias Hoernig, Thomas Klinnert, Ferdinand Lueck, Ralf Nanke, Ludwig Ritschl, Stefan Schaffert, Sabine Schneider, Daniel Stein, Julia Wicklein, Steffen Kappler","doi":"10.1117/1.JMI.12.S1.S13010","DOIUrl":"10.1117/1.JMI.12.S1.S13010","url":null,"abstract":"<p><strong>Purpose: </strong>Digital breast tomosynthesis (DBT) has been introduced more than a decade ago. Studies have shown higher breast cancer detection rates and lower recall rates, and it has become an established imaging method in diagnostic settings. However, full-field digital mammography (FFDM) remains the most common imaging modality for screening in many countries, as it delivers high-resolution planar images of the breast. To combine the advantages of DBT with the faster acquisition and the unique in-plane resolution capabilities known from FFDM, a system concept was developed for application in screening and diagnosis.</p><p><strong>Approach: </strong>The concept comprises an X-ray tube with adaptive focal spot position based on the flying focal spot (FFS) technology and optimized X-ray spectra. This is combined with innovative algorithmic concepts for tomosynthesis reconstruction and synthetic mammograms (SMs).</p><p><strong>Results: </strong>An X-ray tube with FFS was incorporated into a DBT system that performs 50-deg wide tomosynthesis scans with 25 projections in 4.85 s. Laboratory evaluations demonstrated significant improvements in the effective modular transfer function (eMTF). The improved eMTF as well as the effectiveness of the algorithmic concepts is shown in images from a clinical evaluation study.</p><p><strong>Conclusions: </strong>The DBT system concept enables high spatial resolution at short acquisition times. This leads to improved microcalcification visibility, reduced risk of motion artifacts, and shorter breast compression times. It shifts the in-plane resolution of DBT into the high-resolution range of FFDM. The presented technology leap might be a key contributor to facilitating the paradigm shift of replacing FFDM with DBT plus SM.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"12 Suppl 1","pages":"S13010"},"PeriodicalIF":1.9,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11654754/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142865655","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of patient habitus and acquisition protocol on iodine quantification in dual-source photon-counting computed tomography.","authors":"Leening P Liu, Rizza Pua, Michael Dieckmeyer, Nadav Shapira, Pooyan Sahbaee, Grace J Gang, Harold I Litt, Peter B Noël","doi":"10.1117/1.JMI.11.S1.S12806","DOIUrl":"10.1117/1.JMI.11.S1.S12806","url":null,"abstract":"<p><strong>Purpose: </strong>Evaluation of iodine quantification accuracy with varying iterative reconstruction level, patient habitus, and acquisition mode on a first-generation dual-source photon-counting computed tomography (PCCT) system.</p><p><strong>Approach: </strong>A multi-energy CT phantom with and without its extension ring equipped with various iodine inserts (0.2 to 15.0 mg/ml) was scanned over a range of radiation dose levels ( <math> <mrow> <msub><mrow><mi>CTDI</mi></mrow> <mrow><mi>vol</mi></mrow> </msub> </mrow> </math> 0.5 to 15.0 mGy) using two tube voltages (120, 140 kVp) and two different source modes (single-, dual-source). To assess the agreement between nominal and measured iodine concentrations, iodine density maps at different iterative reconstruction levels were utilized to calculate root mean square error (RMSE) and generate Bland-Altman plots by grouping radiation dose levels (ultra-low: <math><mrow><mo><</mo> <mn>1.5</mn></mrow> </math> ; low: 1.5 to 5; medium: 5 to 15 mGy) and iodine concentrations (low: <math><mrow><mo><</mo> <mn>5</mn></mrow> </math> ; high: 5 to 15 mg/mL).</p><p><strong>Results: </strong>Overall, quantification of iodine concentrations was accurate and reliable even at ultra-low radiation dose levels. RMSE ranged from 0.25 to 0.37, 0.20 to 0.38, and 0.25 to 0.37 mg/ml for ultra-low, low, and medium radiation dose levels, respectively. Similarly, RMSE was stable at 0.31, 0.28, 0.33, and 0.30 mg/ml for tube voltage and source mode combinations. Ultimately, the accuracy of iodine quantification was higher for the phantom without an extension ring (RMSE 0.21 mg/mL) and did not vary across different levels of iterative reconstruction.</p><p><strong>Conclusions: </strong>The first-generation PCCT allows for accurate iodine quantification over a wide range of iodine concentrations and radiation dose levels. Stable accuracy across iterative reconstruction levels may allow further radiation exposure reductions without affecting quantitative results.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12806"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11278921/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141789483","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spectral optimization using fast kV switching and filtration for photon counting CT with realistic detector responses: a simulation study.","authors":"Sen Wang, Yirong Yang, Debashish Pal, Zhye Yin, Jonathan S Maltz, Norbert J Pelc, Adam S Wang","doi":"10.1117/1.JMI.11.S1.S12805","DOIUrl":"10.1117/1.JMI.11.S1.S12805","url":null,"abstract":"<p><strong>Purpose: </strong>Photon counting CT (PCCT) provides spectral measurements for material decomposition. However, the image noise (at a fixed dose) depends on the source spectrum. Our study investigates the potential benefits from spectral optimization using fast kV switching and filtration to reduce noise in material decomposition.</p><p><strong>Approach: </strong>The effect of the input spectra on noise performance in both two-basis material decomposition and three-basis material decomposition was compared using Cramer-Rao lower bound analysis in the projection domain and in a digital phantom study in the image domain. The fluences of different spectra were normalized using the CT dose index to maintain constant dose levels. Four detector response models based on Si or CdTe were included in the analysis.</p><p><strong>Results: </strong>For single kV scans, kV selection can be optimized based on the imaging task and object size. Furthermore, our results suggest that noise in material decomposition can be substantially reduced with fast kV switching. For two-material decomposition, fast kV switching reduces the standard deviation (SD) by <math><mrow><mo>∼</mo> <mn>10</mn> <mo>%</mo></mrow> </math> . For three-material decomposition, greater noise reduction in material images was found with fast kV switching (26.2% for calcium and 25.8% for iodine, in terms of SD), which suggests that challenging tasks benefit more from the richer spectral information provided by fast kV switching.</p><p><strong>Conclusions: </strong>The performance of PCCT in material decomposition can be improved by optimizing source spectrum settings. Task-specific tube voltages can be selected for single kV scans. Also, our results demonstrate that utilizing fast kV switching can substantially reduce the noise in material decomposition for both two- and three-material decompositions, and a fixed Gd filter can further enhance such improvements for two-material decomposition.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12805"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11272100/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141789484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning estimation of proton stopping power with photon-counting computed tomography: a virtual study.","authors":"Karin Larsson, Dennis Hein, Ruihan Huang, Daniel Collin, Andrea Scotti, Erik Fredenberg, Jonas Andersson, Mats Persson","doi":"10.1117/1.JMI.11.S1.S12809","DOIUrl":"10.1117/1.JMI.11.S1.S12809","url":null,"abstract":"<p><strong>Purpose: </strong>Proton radiation therapy may achieve precise dose delivery to the tumor while sparing non-cancerous surrounding tissue, owing to the distinct Bragg peaks of protons. Aligning the high-dose region with the tumor requires accurate estimates of the proton stopping power ratio (SPR) of patient tissues, commonly derived from computed tomography (CT) image data. Photon-counting detectors for CT have demonstrated advantages over their energy-integrating counterparts, such as improved quantitative imaging, higher spatial resolution, and filtering of electronic noise. We assessed the potential of photon-counting computed tomography (PCCT) for improving SPR estimation by training a deep neural network on a domain transform from PCCT images to SPR maps.</p><p><strong>Approach: </strong>The XCAT phantom was used to simulate PCCT images of the head with CatSim, as well as to compute corresponding ground truth SPR maps. The tube current was set to 260 mA, tube voltage to 120 kV, and number of view angles to 4000. The CT images and SPR maps were used as input and labels for training a U-Net.</p><p><strong>Results: </strong>Prediction of SPR with the network yielded average root mean square errors (RMSE) of 0.26% to 0.41%, which was an improvement on the RMSE for methods based on physical modeling developed for single-energy CT at 0.40% to 1.30% and dual-energy CT at 0.41% to 3.00%, performed on the simulated PCCT data.</p><p><strong>Conclusions: </strong>These early results show promise for using a combination of PCCT and deep learning for estimating SPR, which in extension demonstrates potential for reducing the beam range uncertainty in proton therapy.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12809"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11576576/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142689215","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Number of energy windows for photon counting detectors: is more actually more?","authors":"Katsuyuki Taguchi","doi":"10.1117/1.JMI.11.S1.S12807","DOIUrl":"10.1117/1.JMI.11.S1.S12807","url":null,"abstract":"<p><strong>Purpose: </strong>It has been debated whether photon counting detectors (PCDs) with moderate numbers of energy windows ( <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> ) perform better than PCDs with higher <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> . A higher <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> results in fewer photons in each energy window, which degrades the signal-to-noise ratio of each datum. Unlike energy-integrating detectors, PCDs add very little electronic noise to measured counts; however, there exists electronic noise on the pulse train, to which multiple energy thresholds are applied to count photons. The noise may increase the uncertainty of counts within energy windows; however, this effect has not been studied in the context of spectral imaging tasks. We aim to investigate the effect of <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> on the quality of the spectral information in the presence of electronic noise.</p><p><strong>Approach: </strong>We obtained the following three types of PCD data with various <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> (= 2 to 24) and noise levels using a Monte Carlo simulation: (A) A PCD with no electronic noise; (B) realistic PCDs with electronic noise added to the pulse train; and (C) hypothetical PCDs with electronic noise added to each energy window's output, similar to energy-integrating detectors. We evaluated the Cramér-Rao lower bound (CRLB) of estimation for the following two spectral imaging tasks: (a) water-bone material decomposition and (b) K-edge imaging.</p><p><strong>Results: </strong>For both the e-noise-free and realistic PCDs, the CRLB improved monotonically with increasing <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> for both tasks. In contrast, a moderate <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> provided the best CRLB for the hypothetical PCDs, and the optimal <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> was smaller when electronic noise was larger. Adding one energy window to the minimum necessary <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> for a given task gained 66.2% to 68.7% of the improvement <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> <mo>=</mo> <mn>24</mn></mrow> </math> provided.</p><p><strong>Conclusion: </strong>For realistic PCDs, the quality of the spectral information monotonically improves with increasing <math> <mrow><msub><mi>N</mi> <mi>E</mi></msub> </mrow> </math> .</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12807"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11413649/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142298697","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Photon Counting: Detectors and Applications.","authors":"Patrick J La Riviere, Mini Das","doi":"10.1117/1.JMI.11.S1.S12801","DOIUrl":"10.1117/1.JMI.11.S1.S12801","url":null,"abstract":"<p><p>The editorial introduces the special issue on photon counting detectors and applications.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12801"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11684477/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142915972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Utility of photon-counting detectors for MV-kV dual-energy computed tomography imaging.","authors":"Giavanna Jadick, Maya Ventura, Patrick J La Rivière","doi":"10.1117/1.JMI.11.S1.S12811","DOIUrl":"10.1117/1.JMI.11.S1.S12811","url":null,"abstract":"<p><strong>Purpose: </strong>High soft-tissue contrast imaging is essential for effective radiotherapy treatment. This could potentially be realized using both megavoltage and kilovoltage x-ray sources available on some therapy treatment systems to perform \"MV-kV\" dual-energy (DE) computed tomography (CT). However, noisy megavoltage images obtained with existing energy-integrating detectors (EIDs) are a limiting factor for clinical translation. We explore the potential for non-spectral photon-counting detectors (PCDs) to improve MV-kV image quality simply by equally weighting all MV photons rather than up-weighting the less informative, lower contrast high-energy photons as in an EID.</p><p><strong>Approach: </strong>Three computational methods were applied to compare non-spectral PCDs with EIDs in MV-kV DE imaging. A single-line integral estimation theory approach was used to calculate the basis material signal-to-noise ratio (SNR) of tissue (1 to 50 cm) and bone (0.1 to 10 cm). CT images of a tissue cylinder with seven bone inserts (densities 1.0 to <math><mrow><mn>2.2</mn> <mtext>  </mtext> <mi>g</mi> <mo>/</mo> <msup><mrow><mi>cm</mi></mrow> <mrow><mn>3</mn></mrow> </msup> </mrow> </math> ) were simulated to assess material decomposition accuracy. Multiple noisy simulations of an anthropomorphic phantom were performed to generate pixel-by-pixel noise profiles.</p><p><strong>Results: </strong>PCDs yielded a 15% to 45% improvement in single-line integral SNR for both materials. In CT simulations, similar material decomposition accuracy was achieved, with both EIDs and PCDs slightly underestimating bone density. However, PCDs yield a higher contrast-to-noise ratio and more uniform noise texture than EIDs in virtual monoenergetic images.</p><p><strong>Conclusions: </strong>We demonstrate the potential for improved MV-kV DE CT imaging using non-spectral PCDs and quantify the degree of improvement in a range of object compositions. This work motivates the experimental assessment of PCDs for megavoltage imaging and the potential clinical translation of PCDs to radiotherapy imaging.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12811"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11670364/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142903844","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Contrast-to-noise ratio comparison between X-ray fluorescence emission tomography and computed tomography.","authors":"Hadley DeBrosse, Giavanna Jadick, Ling Jian Meng, Patrick La Rivière","doi":"10.1117/1.JMI.11.S1.S12808","DOIUrl":"https://doi.org/10.1117/1.JMI.11.S1.S12808","url":null,"abstract":"<p><strong>Purpose: </strong>We provide a comparison of X-ray fluorescence emission tomography (XFET) and computed tomography (CT) for detecting low concentrations of gold nanoparticles (GNPs) in soft tissue and characterize the conditions under which XFET outperforms energy-integrating CT (EICT) and photon-counting CT (PCCT).</p><p><strong>Approach: </strong>We compared dose-matched Monte Carlo XFET simulations and analytical fan-beam EICT and PCCT simulations. Each modality was used to image a numerical mouse phantom and contrast-depth phantom containing GNPs ranging from 0.05% to 4% by weight in soft tissue. Contrast-to-noise ratios (CNRs) of gold regions were compared among the three modalities, and XFET's detection limit was quantified based on the Rose criterion. A partial field-of-view (FOV) image was acquired for the phantom region containing 0.05% GNPs.</p><p><strong>Results: </strong>For the mouse phantom, XFET produced superior CNR values ( <math><mrow><mi>CNRs</mi> <mo>=</mo> <mn>24.5</mn></mrow> </math> , 21.6, and 3.4) compared with CT images obtained with both energy-integrating ( <math><mrow><mi>CNR</mi> <mo>=</mo> <mn>4.4</mn></mrow> </math> , 4.6, and 1.5) and photon-counting ( <math><mrow><mi>CNR</mi> <mo>=</mo> <mn>6.5</mn></mrow> </math> , 7.7, and 2.0) detection systems. More generally, XFET outperformed CT for superficial imaging depths ( <math><mrow><mo><</mo> <mn>28.75</mn> <mtext>  </mtext> <mi>mm</mi></mrow> </math> ) for gold concentrations at and above 0.5%. XFET's surface detection limit was quantified as 0.44% for an average phantom dose of 16 mGy compatible with <i>in vivo</i> imaging. XFET's ability to image partial FOVs was demonstrated, and 0.05% gold was easily detected with an estimated dose of <math><mrow><mo>∼</mo> <mn>81.6</mn> <mtext>  </mtext> <mi>cGy</mi></mrow> </math> to a localized region of interest.</p><p><strong>Conclusions: </strong>We demonstrate a proof of XFET's benefit for imaging low concentrations of gold at superficial depths and the feasibility of XFET for <i>in vivo</i> metal mapping in preclinical imaging tasks.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12808"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11478016/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142477763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Iterative clustering material decomposition aided by empirical spectral correction for photon counting detectors in micro-CT.","authors":"J Carlos Rodriguez Luna, Mini Das","doi":"10.1117/1.JMI.11.S1.S12810","DOIUrl":"10.1117/1.JMI.11.S1.S12810","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Purpose: &lt;/strong&gt;Photon counting detectors offer promising advancements in computed tomography (CT) imaging by enabling the quantification and three-dimensional imaging of contrast agents and tissue types through simultaneous multi-energy projections from broad X-ray spectra. However, the accuracy of these decomposition methods hinges on precise composite spectral attenuation values that one must reconstruct from spectral micro-CT. Errors in such estimations could be due to effects such as beam hardening, object scatter, or detector sensor-related spectral distortions such as fluorescence. Even if accurate spectral correction is done, multi-material separation within a volume remains a challenge. Increasing the number of energy bins in material decomposition problems often comes with a significant noise penalty but with minimal decomposition benefits.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Approach: &lt;/strong&gt;We begin with an empirical spectral correction method executed in the tomographic domain that accounts for distortions in estimated spectral attenuation for each voxel. This is followed by our proposed iterative clustering material decomposition (ICMD) where clustering of voxels is used to reduce the number of basis materials to be resolved for each cluster. Using a larger number of energy bins for the clustering step shows distinct advantages in excellent classification to a larger number of clusters with accurate cluster centers when compared with the National Institute of Standards and Technology attenuation values. The decomposition step is applied to each cluster separately where each cluster has fewer basis materials compared with the entire volume. This is shown to reduce the need for the number of energy bins required in each decomposition step for the clusters. This approach significantly increases the total number of materials that can be decomposed within the volume with high accuracy and with excellent noise properties.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Utilizing a (cadmium telluride 1-mm-thick sensor) Medipix detector with a &lt;math&gt;&lt;mrow&gt;&lt;mn&gt;55&lt;/mn&gt; &lt;mtext&gt;-&lt;/mtext&gt; &lt;mi&gt;μ&lt;/mi&gt; &lt;mi&gt;m&lt;/mi&gt;&lt;/mrow&gt; &lt;/math&gt; pitch, we demonstrate the quantitatively accurate decomposition of several materials in a phantom study, where the sample includes mixtures of soft materials such as water and poly-methyl methacrylate along with contrast-enhancing materials. We show improved accuracy and lower noise when all five energy bins were used to yield effective classification of voxels into multiple accurate fundamental clusters which was followed by the decomposition step applied to each cluster using just two energy bins. We also show an example of biological sample imaging and separating three distinct types of tissue in mice: muscle, fat, and bone. Our experimental results show that the combination of effective and practical spectral correction and high-dimensional data clustering enhances decomposition accuracy and reduces noise in micro-CT.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 Suppl 1","pages":"S12810"},"PeriodicalIF":1.9,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11676343/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142903840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-supervised learning for chest computed tomography: training strategies and effect on downstream applications.","authors":"Amara Tariq, Gokul Ramasamy, Bhavik Patel, Imon Banerjee","doi":"10.1117/1.JMI.11.6.064003","DOIUrl":"https://doi.org/10.1117/1.JMI.11.6.064003","url":null,"abstract":"<p><strong>Purpose: </strong>Self-supervised pre-training can reduce the amount of labeled training data needed by pre-learning fundamental visual characteristics of the medical imaging data. We investigate several self-supervised training strategies for chest computed tomography exams and their effects on downstream applications.</p><p><strong>Approach: </strong>We benchmark five well-known self-supervision strategies (masked image region prediction, next slice prediction, rotation prediction, flip prediction, and denoising) on 15 M chest computed tomography (CT) slices collected from four sites of the Mayo Clinic enterprise, United States. These models were evaluated for two downstream tasks on public datasets: pulmonary embolism (PE) detection (classification) and lung nodule segmentation. Image embeddings generated by these models were also evaluated for prediction of patient age, race, and gender to study inherent biases in models' understanding of chest CT exams.</p><p><strong>Results: </strong>The use of pre-training weights especially masked region prediction-based weights, improved performance, and reduced computational effort needed for downstream tasks compared with task-specific state-of-the-art (SOTA) models. Performance improvement for PE detection was observed for training dataset sizes as large as <math><mrow><mo>∼</mo> <mn>380</mn> <mtext>  </mtext> <mi>K</mi></mrow> </math> with a maximum gain of 5% over SOTA. The segmentation model initialized with pre-training weights learned twice as fast as the randomly initialized model. While gender and age predictors built using self-supervised training weights showed no performance improvement over randomly initialized predictors, the race predictor experienced a 10% performance boost when using self-supervised training weights.</p><p><strong>Conclusion: </strong>We released self-supervised models and weights under an open-source academic license. These models can then be fine-tuned with limited task-specific annotated data for a variety of downstream imaging tasks, thus accelerating research in biomedical imaging informatics.</p>","PeriodicalId":47707,"journal":{"name":"Journal of Medical Imaging","volume":"11 6","pages":"064003"},"PeriodicalIF":1.9,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11550486/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142630349","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}